LoopSuit · Operator's Guide / 2026

Field manual · rebuilt 2026-09-21 · the ~150 newest Saraev + Herk + Hormozi videos (Jul–Sep 2026)

Everything you do to get this agency past $10k/mo.

One page, current as of September 2026, deprecations removed. Hover any framed block for a copy icon — it copies as Markdown. Full-context MD files for AI project knowledge live in Md docs/agency-guide-2026/.

5×2 = 10 inboxes · 20/mailbox
one lane, three channels in parallel
gate: ≥5 meetings / 1,000 sends
verdicts at ~3 months, not 3 weeks

⚠️ September 2026 audit — six things on this page were wrong until today

Rebuilt against the ~150 videos published since the July pull. If you read this guide before 21 Sep, re-read these:

WasNowWhy it matters
30–40 sends/day per mailbox · 900/day fleet20/day, ceiling 25 · 600/day fleetESPs tightened in 2026. Running at 30 is the most likely cause of a campaign decaying from 7% to 1% in a fortnight. Grow by adding inboxes, never by raising volume.
Paid Clarity Session sold from coldFree 15-min session, arriving with work already doneAudits convert on warm traffic only. From cold, asking a stranger for an hour to explain their business to someone who obviously wants to sell them something is the ask that kills the reply.
Any generic database as the main lead sourceBuying signals — hiring a designer, or freshly funded15–25% reply vs a 3.43% baseline. And a verifier was never the fix for the 68%/60% bounce — it deletes guessed addresses, it cannot create good ones. Fix the source.
"Under 5% bounce is fine"Under 2% · verified-only email finder → MillionVerifierGoogle treats sustained 2%+ as a spam signal across the whole domain, not just the campaign.
Judge a reply by how it readsQualify on the call, never on the replyTerse lowercase "sure" becomes thousands of dollars. You cannot tell from four words.
Daily content from day oneProof content now, treadmill after $10–20k/moCase studies and a clean GitHub are sales collateral. Daily posting on two platforms is a job you cannot afford yet — and the people telling you otherwise are the survivors.

New tabs, with the detail: 📬 Inbox Engine (what to buy, in what order — start here if you are stuck on mailboxes) · 💰 Pricing (value-based pricing calculator + the road to $10–15k) · ✍️ Copy Lab (the four-question message architecture).

01North Star & daily rhythm

The whole strategy in one breath — everything else on this page serves it.

The one-paragraph strategy

Run one cold-email lane — AI products designed and shipped in 2–4 weeks, sold to funded seed/Series-A founders — on a Zapmail→Instantly fleet of 5 domains × 2 mailboxes = 10 inboxes at 20 sends/day (200/day), growing to 30 inboxes only after 30 days under 2% bounce. Leads come from buying signals, not a database — companies hiring a designer, or freshly funded (15–25% reply vs a 3.43% baseline) — found via Sales Nav Spotlights and job-board scrapers, emailed via a pay-only-for-verified finder, gated by MillionVerifier, valid only, bounce under 2%. From cold the working session is free and you arrive having already done work — never a paid audit from cold. Projects price at 10–20% of the annual value of the problem, floor $3,000, 50% up front, objective 30-day milestones → support retainer → one fractional anchor. Lane gate: ≥5 booked meetings per 1,000 sends. Cold email runs in parallel with Upwork at 3 applications/day and 10 founder DMs, because those have no warmup clock and produce the first money. Templates are human-approved once; AI only fills variables. Promotion before building, every single day.

WhenWhatTime
Daily Mon–SatReply queue — respond to all outstanding leads (approve drafts, never compose)~30 min
Daily Mon–SatLinkedIn — 20–30 connection requests / DMs to lane ICPs~30 min
Daily Mon–SatInstantly glance (bounce <5%) + one campaign improvement, however small~10 min
SundayDeterminations: funnel numbers → call variant winner → pick ONE next test → load next batch~45 min
Monthly1 public case study + pricing review (raise if closing >1 in 3)~2 h

Rule  Hormozi's Rule of 100 daily — 100 outreach touches or 100 minutes on the funnel. FSH hours don't count. Building doesn't count. Founder day shape: promotion first, delivery second, building last (4-4-4, scaled to your ~90 min).

02Cold outreach

The 4-part formula survives. Around it, five 2026 rules changed — including who writes the copy.

the email skeleton1. PERSONALIZATION — hook attention, prove a human wrote it 2. WHO AM I — credibility in ONE sentence, outcome-first 3. THE OFFER — specific, time-bound, risk-reversed 4. CTA — one action, lowest friction ("X or Y work for you?") 5–8 sentences. Plain text. No links, no HTML, no images.

ReversalAI doesn't write emails

Saraev's walk-back: heavy AI copy now pattern-matches as spam ("people that use AI really intensely tend to suck"). Humans write template banks once; AI only fills casualization variables ("Vancouver" → "East Van").

NewGive-before-ask

"Quick question" openers are dead — they extract value from a stranger. Order: credential/social proof → guarantee → low-friction time ask.

NewVoluntary disclosure

One deliberately off-topic personal line per email ("I'm from [city] originally — anyway…") breaks the AI-spam pattern and builds closeness. Yours: TD Bank, FSH, 2am builds. Sprinkle, don't wreck.

RuleThe visibility hierarchy

Fix in order: sender name → subject → preview text → offer. "If you get the email clicked, that's 90% of the battle." Never waste the preview on "Hi [Name],".

RuleCompression

Replace every term with a higher-info-density term; shave 20–30% length. Skimmable is a pattern interrupt.

DoTwo-step sequence

Day 0 + one bump (day 3–4). Follow-ups are C-tier with diminishing returns. Send 7am–7pm recipient time, Mon–Sat; Sunday is for numbers.

give-before-ask opener shape[personal/credential line] — I spent 3 years shipping TD Bank's mobile app; these days I build revenue-recovery systems for [vertical] like yours. [voluntary disclosure] I'm from [city] originally — anyway. [offer + guarantee] I'll [named outcome] in [timeframe] — and if we don't hit [KPI], you don't pay the balance. [CTA] Worth 15 minutes? Tue 3:30 or Wed 10 work?

Lists — the 2026 sourcing chain

  1. Buying signals — the primary pathSignal-triggered outreach replies at 15–25% against a 3.43% baseline. For this offer the strongest signal is a company actively hiring a product designer: budget allocated, pain stated publicly, deadline real, hiring manager named — and your offer is a direct substitute for the hire. Source it two ways in parallel: Sales Nav Spotlight filters (Job Opportunities + Recent funding + Headcount growth), exported with a pay-per-result Apify actor at ~$0.01/lead; and job-board actors (Wellfound, YC, Greenhouse/Lever/Ashby) at ~$3/1K with no LinkedIn exposure. Full setup in the Inbox Engine tab.
  2. The 100-row QA gateBefore buying a single enrichment credit: paste 100 rows into Claude — "how many of these are actually my ICP?" ≥50 = ship it. Under 50 = fix the filter and re-pull. Sixty seconds. Skipping this killed two lanes.
  3. Apify actor — budget/ammunition only~$1.50–2 per 1,000 email-attached, ≈$10/1k after verification. Free tiers return pattern-guessed addresses — a real domain with an invented mailbox — which is what bounced UnitOS at 68% and QSR at 60%. Use it only for disposable volume on throwaway inboxes.
  4. Verified-only finder → MillionVerifierFind the address with a tool that charges only when the email is verified (Anymail Finder, Findymail) — risky and not-found cost nothing, which fixes the bounce problem at source rather than downstream. Then MillionVerifier as the final gate: send valid only. Target bounce is under 2%, not under 5% — above 2% sustained, Google penalises the whole domain. Bouncer is an optional catch-all recovery pass; skip it until you are discarding 30%+ of a good list.
  5. Dedup + your 7-stage pipelineClean → enrich → casualize → 19-col Instantly CSV. Personalization = your approved bank + variable fill.

Tiers  S: private lists (conference attendees, associations — his only 15–20% reply campaigns) · A: scarily specific personalization · B: custom-scraped > databases · C: follow-ups. Geography arbitrage is real: a Czech clone of a 0.5%-reply US campaign hit ~10% positive. Test one non-US lane.

Infrastructure numbers (September 2026)

ThingCurrent answer
Mailbox math20/day/mailbox — hard ceiling 25 · start 5 domains × 2 = 10 inboxes = 200/day · ceiling 10 domains × 3 × 20 = 600/day (30–40/day is deprecated — ESPs tightened in 2026)
Buying inboxesZapmail ~$3/mo (Saraev's named pick, Sep 2026) or Instantly airmail. Never dedicated SMTP pools. Check the AUP explicitly permits cold outbound before buying — that is what killed the Zoho fleet.
Warmup14–21 days on your own domains — or pre-warmed and send day one. These are different jobs, not alternatives: self-warm the brand inboxes; buy 3–5 pre-warmed as the bridge and as ammunition you are willing to burn. Re-bid the stack every 2–3 months.
A/B testing500–1,000 sends per variant, judged on reply rate, ONE variable per week. Big forks first, micro-iterations later.
BenchmarksPlan at 0.25% booked-meeting rate (1 per 400 new-lead emails — remember ~half your sends are follow-ups); 0.5% is good and is Saraev's own current number; 0.06% means the top of the funnel is broken. When a campaign bombs: ~50–60% copy, ~20–30% list/mailbox, ~10% luck.
Lane gate≥5 booked meetings per 1,000 sends — or the lane dies. Verdicts at ~3 months (60-day market lag).

03Proposals & closing

A proposal is a test: "does this person understand my problem better than I can articulate it?"

DoctrineProblem-first, ≥50% diagnosis

Half the doc quantifies their pain in dollars before any solution appears. Price last, framed as investment. A 10-page diagnosis once turned a $5k deal into $100k+.

MathThree cost layers

Hard costs + opportunity cost + operational chaos. Reframe savings as revenue: "10 staff-hours = $5,000/mo of selling time recovered," never "$500 of wages saved."

2026The single-artifact close

Proposal + contract + Stripe + signature in one document, with a personal video walkthrough. Payment auto-books onboarding. Every extra step leaks ~10% of leads.

System3 decisions, 72 hours

Traditional close (SMB default): the call earns decision #1 ("worth solving?"); the proposal carries #2–3 inside a 72-hour price window. Automated follow-up until an explicit yes/no.

pricing rules· Value-based: charge 10–30% of the quantified annual pain (~20% mid) · Fixed-price first. NEVER retainers cold (100 retainer pitches ≈ 2 yes; 100 fixed-price ≈ 15 yes, 3 → retainers) · Never time-based build fees — flat $X for the outcome, 4–6 wk scope · Pay-per-meeting when asked: Rate = LTV × CVR(halved) × 15% e.g. $25k × 0.10 × 0.15 = $375/booked meeting · Guarantee ≈ 3× top line at 2–5% refund cost → charge 100% upfront · Price cap for builds: "total will not exceed $X" (pad 50%) · Raise prices when closing >1 in 3 · Never discount — cut scope · Time estimates: +30% buffer, multiples of 4 business days from Monday

EV  Every customized proposal in flight ≈ $4,000 expected value (5 × 20% close × $20k LTV). Track proposals-in-flight, not tasks done. — wiki: problem-first-proposal, dont-sell-retainers-cold

04Sales calls & objections

The working session is a constraint-and-KPI diagnosis (Herk: "builders are pharmacists, consultants are doctors"). Consult, don't close. ⚠️ From cold it is free — a paid audit sold to a stranger is the fastest way to get no reply at all; paid diagnostics convert on warm traffic only. Open with "if you 10×'d tomorrow, what would break first?" — then say nothing.

RACE discovery spine — ask theseREACH "How are you currently reaching new customers?" ACQUIRE "What happens in the first 60 seconds after a lead comes in?" CONVERT "How do leads become paying customers? Where do they stall?" EXPAND "When did you last reactivate past customers?" Only sell systems that move one of the four. Anything else: smaller market, less urgency, lower price. Selling "savings" is the trap.

HormoziCLOSER script

Clarify why they came · Label the problem · Overview past attempts · Sell the vacation (never the flight — no process talk) · Explain away concerns · Reinforce the decision.

HormoziBANT before the pitch

Budget · Authority ("anyone else in this decision?") · Need ("what happens if you don't solve it?") · Timing. Unqualified in 5 min → downsell, not a 45-min pitch.

Map7 pains → 3 masters

Leads · speed-to-lead · closing · onboarding bottleneck · payroll bloat · bad software · owner chaos — all reduce to make money / save time / kill chaos. Quantify the bleeding one, in dollars, on the call.

FlowClarity Session

Map the customer journey → find the constraint (what caps revenue, not what annoys) → dollar-quantify → attach a KPI → close to a fixed-price pilot. Never a cold retainer.

The worked scenario (all scripts below use it)

Suppose you're selling…

Rebound — the lapsed-client reactivation system — to Maya, owner of a 6-chair med spa in Austin. Her numbers (from the Revenue Leak Calculator or the session itself): ~2,400 client records, ~700 lapsed 6+ months, average ticket $180 → she's leaking roughly $8–10k/month. Your asks: a free 15-minute working session (from cold — you arrive having already pulled her lapsed-client math) then the $1,997 pilot (post-session), guaranteed: "reactivation system live in 14 days and 30 rebooked clients in 60 days, or the balance is refunded." Swap the nouns for the QSR lane (lapsed clients → Swiggy/Zomato invisibility, tickets → covers) and every script below still works.

Mock scripts — the famous objections

Blended from Miner's NEPQ word-tracks, Hormozi, and Saraev — adapted to a paid-diagnostic motion. Hover the underlined phrases for why each line works.

Objection 1"How much does it cost?" — asked way too early
Maya"Before we book anything — what does this cost?"
You"Yeah, that's not a problem. Builds like this run anywhere from $1.5k to $15k — it depends entirely on what we find, and that's exactly what the session pins down. The session itself is $397, and if you don't feel you got $397 of value out of the hour, I refund it — independent of anything else."
You"And honestly — the system that got this email into your inbox is the same kind of system we'd build for you."
Objection 2"That's too expensive." — after the pilot pitch
Maya"Two grand… that's a lot right now."
You(leaning in, concerned) "That could genuinely be a problem for you, then. Let me ask — if the funds weren't the issue, is this something you feel would work for you?"
Maya"I mean… yeah, probably."
You"Hm — why do you feel like it would, though?"
Maya"Because nobody's chasing those 700 people. They liked us, they just… drifted."
You"Right — and that drift is the $8k a month we measured. So the real question is how you'd free up the $1,997 once, so you can stop losing the $8k every month you told me about. Best case: the system's live in two weeks and 30 of those clients are back on your calendar in 60 days. Worst case: they're not, and the balance comes back to you. The only guaranteed outcome is that doing nothing keeps the leak open. What do you want to do?"
Objection 3"I need to think it over."
Maya"This sounds good, but let me think about it."
You"Yeah, that's not a problem. What's your timeframe for getting back to me — tomorrow, day after? Let's just put ten minutes on the calendar now so you're not chasing me down and I'm not chasing you."
Maya"Sure, Thursday morning works."
You"Booked. Oh — before I let you go — what specifically were you wanting to think through? Just so I know what questions you'll have Thursday."
Maya"Honestly… I'd want my manager Priya to be okay with it, she runs the front desk."
You"That makes sense — should Priya just join us Thursday, then? Fifteen minutes, I'll walk her through exactly what lands on her desk — which is almost nothing, that's the point."
Objection 4"I need to talk to my business partner."
Maya"I'd have to run this by my business partner first."
You"Of course. How does he feel about the lapsed-client situation — the $8k a month side of it?"
Maya"He's the one who keeps bringing it up, honestly."
You"Then let's not play telephone — would it help if the three of us did a 20-minute call and I walked him through the same numbers? And so I'm useful in that call — what do you think he's going to say?"
Maya"Probably that we tried a marketing agency before and it went nowhere."
You"Good — then I'll bring the two things an agency couldn't give you: a fixed price with an end date, and the refund clause if the 30 bookings don't show up."
Objection 5"Just send me some information." — the email classic
Maya"Can you just email me some info about it?"
You"Happy to. Just so I send the right thing and not a brochure dump — what's the main thing you'd want it to answer?"
Maya"Mostly what it'd cost and how much work it is for my staff."
You"Perfect — I'll send the med-spa one-pager: pricing, the 14-day timeline, and the part your front desk touches (about ten minutes a week). The pilot pricing in it holds for 72 hours — and let's say it covers what you need: what would you want the next step to be?"
Objection 6"We already work with an agency / have software."
Maya"We already pay a marketing agency, and the booking software has reminders built in."
You"Keep both — seriously. Honestly, I'm not even sure we could add anything yet — it depends what they're doing about the 700 people who already stopped coming. The agency brings new faces in the door; reminders catch people who already booked. Who's calling the ones who drifted — and what did that change in booked revenue last quarter?"
Maya"…Nobody, really. The reminders only go to people with appointments."
You"That's the gap. Reactivation isn't marketing and it isn't reminders — it's its own lane, and it's the cheapest revenue you have because these people already chose you once."
Objection 7"I'm slammed — call me back next month."
Maya"This month is crazy — try me again in a few weeks?"
You"Yeah, not a problem. Rather than us playing tag in three weeks, grab your calendar for a second — let's lock ten minutes on the 4th and I'll call you then."
Maya"Fine — the 4th at 10."
You"Done. And — small thing to sit with until then: the reason this works for busy months is that the system does the chasing, not you or your front desk. 'Slammed' is the problem it's for."
Objection 8"Won't AI messages feel robotic to my clients?"
Maya"My clients pay for a personal experience. AI texts feel… cheap."
You"You're right — most of it does, and most of it is. That's exactly why this isn't a chatbot blasting your list. Every message is one you approved, in your front desk's name and your voice — the system just picks who gets it, when, based on their visit history. Nothing goes out that you haven't read. Want to see the actual messages other spas approved?"

NEPQ (Jeremy Miner) — the verdict

Researched July 2026: his videos, book (The New Model of Selling), and the credible critiques. Short version: it's SPIN-with-feelings plus the best objection word-tracks on the market — steal three things, skip three things.

the NEPQ question ladder → mapped onto the $397 clarity sessionCONNECTION (min 0–3) "What made you grab this session, of everything else you could've spent the morning on?" SITUATION (3–10) "Walk me through what happens after a client's third visit — who follows up, and when?" PROBLEM (10–20) "How long has the follow-up been falling through?" → "Has that had an impact on you?" → "In what way?" → echo her emotional word back: "…Frustrating?" SOLUTION (—) "If those 700 people started rebooking on their own — what does the calendar look like in March?" CONSEQUENCE (~20) "And if nothing changes and the drift continues — where's the client base in two years?" ── FLIP: paid diagnostic means you now PRESCRIBE ────────────────── FINDINGS (20–35) Your numbers, her leak, the three fixes, the KPI. COMMITMENT (35–45) "Does this feel like it'd work for you?" → "Why do you feel like it would?" → "What would you want the next step to be?"

StealThe question ladder + gap

"The bigger the gap between where they are and where they want to be, the cheaper your solution seems." Run it for the first 15–20 minutes; her stated pain funds the whole pitch. The probing stack — "tell me more / walk me through / give me a specific example / how long has that been going on / has that had an impact on you / in what way?" — is gold with owner-operators, whose business pain is personal.

StealThe five tones

Curious (the workhorse — all discovery), confused ("hold on — how do you mean?" makes them expand), concerned (pain probes, "off the record…"), challenging (consequence questions, only after trust, last third of the call), playful (disarms). Slow the second half of every important question. Keep volume and speed flat; vary only pause and pitch.

StealThe objection loop

Agree ("yeah, that's not a problem") → book the specific next step (calendar commitment) → "before I go — what were you wanting to go over in your mind?" → the real objection surfaces → handle THAT one with questions, tied to her stated outcome. This loop is Objections 3, 4, and 7 above.

SkipStatus theater

"Just to see if I'd be available for you," waiting 3 minutes to call back, scripted stutters. Maya paid you — the authority is real, the games read as games. Keep the substance (specific booked next steps), drop the posture.

SkipFive-rung pain excavation

Miner's depth is calibrated to B2C life decisions (insurance, biz-opp). Spa and restaurant owners buy on time-back and revenue math — go one or two rungs down the pain ladder, then bring the numbers. Your consequence question is a dollar figure, not a feeling: "22 missed rebookings a month at $180 is $47k a year."

SkipFaked detachment

His sellers perform "I'm not sure we can help you." Yours is true: "If the numbers say this isn't worth building, I'll tell you that in the session — that's what the $397 buys." The honest version of his best frame is stronger than the theatrical one. Put it in the offer copy.

NEPQ signature phrasings — verified, adapted to loopsuit"I'm not sure we could even help you yet — it depends what [current provider] is doing about [the leak]." ← master disarm "Would you be opposed to…?" ← no-oriented ask (cold) "Would it be appropriate to…?" / "Would that help you?" "Yeah, that's not a problem." ← objection opener, always "How do you mean?" ← confused tone, they expand "What was it about [the calculator / the email] that caught your eye?" "Is there anything you'd change about how [follow-up] works right now, if you could?" ← the dissatisfaction-finder "How long has that been going on?" → "Has that had an impact on you?" → "In what way?" ← the pain ladder "What happens if you don't do anything about this and it gets worse?" "Why now, though?" ← they argue for urgency "Money aside — why do you feel like it would work, though?" "Before I go — what were you wanting to go over in your mind?" Say: "funds," "agreement," "feel." Never: "budget," "contract," "think."

Tone  Constant volume and speed (135–185 wpm); vary only pauses and pitch. Ask the closing question with raised pitch — then a massive silence. Damaging admissions first ("here's what sucks about this") — trust beats promise. And BANT's authority question early ("anyone else in this decision?") prevents Objection 4 from ever appearing.

Reminder: scripts are for internalizing, not reciting. The tooltips carry the mechanisms — learn those, and the words rearrange themselves naturally on a live call.

05Offers & positioning

The Hormozi layer — including the one warning that reshapes your ladder.

⚠ The Dead Zone (Hormozi, 2026)

The $1,500–3,000/mo services retainer for SMBs is a dead zone: volatile clients, 4–6 month stick, and they blame you for their own sales problems. Escape to either pole — never price in the middle.

Go DOWN · <$400/mo

Fully automated, near-zero delivery cost: monitoring, audits, self-serve tools. Entry-tier Rank Radar (India / small fleets) lives here.

dead zone
$1.5–3k/mo
Go UP · high-ticket

Big one-time builds for clients with a working sales process. Big Head ($2–7k+ fixed) + Long Tail ($250–500/mo consumable).

The exception — software is not a services retainer

The Dead Zone is about MANAGED SERVICES: monthly labor, ambiguous attribution, client blames you for their sales. Software with near-zero marginal delivery cost is exempt — it has no labor tail and the value is directly observable. Rank Radar at NA pricing ($1,500–2,500/mo per fleet) sits ABOVE the dead zone, not below it, because the buyer can attribute a rank drop to lost revenue without your help.

the one pricing architectureCLARITY SESSION — $1,500 (was $397) Paid diagnostic. Qualifies the buyer. Refunded if it does not surface three actionable opportunities. BLUEPRINT — from $4,000 Credited into the build. CUSTOM BUILD — from $12,000 (was $1,997) Fixed price, capped in writing, KPI guarantee. AI OPS RETAINER — from $2,500/mo (replaces $297 keep-alive) Monitoring + monthly improvements + one new workflow per quarter. EMBEDDED PARTNER — from $6,000/mo Published, not "by invitation." RANK RADAR — India ₹18–60k/mo · North America $1,500–2,500/mo Same product, different market. India for case studies and iteration speed; NA for revenue. FLEET AUDIT — $600 / ₹45,000 one-time The near-term revenue product. Credited against month one.
Dream outcome × Likelihood of achievement
Time delay × Effort required
= VALUE

Levers  Likelihood is a solo operator's cheapest lever — done-for-you delivery makes belief "go through the roof." And "latency beats magnitude 7 days a week" — "live in 7 days" out-sells "20% better eventually." Priority tier = half the time, add a zero.

DoObstacle bonus stack

Every objection → build the mini-solution once → it becomes a permanent named bonus: "I have a bonus specifically for that." Each bonus must feel worth the core offer.

DoName the result

"Empty-Chair Recovery Sprint" beats "AI automation package." Never change the deliverable; re-wrap the name quarterly to cure fatigue. A/B test 3–4 names.

DoGuarantees that print

"If you don't get X by Y, I will Z" — power lives in the Z. Clarity Session: unconditional value guarantee ("not $500 of value → refund" — claimed twice in years). Builds: conditional on access/data/SLAs.

DoLead magnets that deprive

Type: "reveal a problem" (calculator, audit ✓). Rule: fully solve problem A in a way that exposes problem B — the paid offer. Free content teaches WHAT/WHY; the offer sells HOW.

2026Proof is the AI-era moat

AI floods the market with identical advice; the differentiator is demonstrated results — live dashboards, documented ops, real case studies. (Why un-gating links-hub matters.)

2026Sell a department

"Bring your own agent": decompose a role into ~5 tasks, automate task-by-task, pitch "I'm your entire reactivation department." Department-shaped offers command department money.

Price diagnosis  Closing >80% = underpriced 3–4× · 60–80% = 2–3× · 30–40% = right · <30% = not a price problem (wrong avatar or broken pre-call motion). A 10% raise ≈ +25% net profit. Anchor first, anchor high; charge upfront; "10 months, get 2 free."

06The 2026 AI stack

Both creators independently dropped visual no-code in late 2025. The model is the framework now.

StackClaude Code is the harness

Routines = the n8n scheduler replacement; Managed Agents = the hosted tier. New automations native; working flows stay — tokens cost more than compute.

RuleNo wrappers

"Every additional framework is inversely correlated with money made." Skills only for repeated tasks (proposals, onboarding); voice-dump the rest with OCCD: Objective, Context, Constraints, Definition of Done.

CostManager + cheap workers

Frontier model plans/reviews (~5–15% of work); Sonnet/GLM-class execute at 1/5–1/10 price. Extract the expensive model's process into skills — "you can't keep the intelligence; you can keep the process."

HygieneContext discipline

Quality degrades ~250k tokens (Opus-class) — clear and hand off via session docs. Work in prompt-cache sprints; cap thinking effort by default; grep, don't full-read.

DangerVerify or it didn't happen

Real incidents: an agent blasted a discount code to a 150–200k list; another "sent" 25% of claimed emails. Scoped keys, hooks, human approval on anything outward. Worker ≠ judge.

MemoryMarkdown wiki, not vectors

Karpathy-style: raw/ → wiki/ + index + CLAUDE.md routing. Herk runs his whole business at this level. Ingest evergreen only; route to live sources for volatile data.

The metrics tracker (build first, feeds everything)

caveman spreadsheet — per lane, per daysends · replies · positive replies · meetings booked · meetings had · show-ups · proposals out · closes · $ collected (all different numbers — track each) computed: reply rate · close-to-reply · revenue-per-proposal · DoD growth jobs: Sunday determinations · day-21 lane verdict · bottleneck diagnosis (default bottleneck: top of funnel — the icebreaker)

07Deprecated — do NOT do these

Everything below was good advice once — including in your own wiki and April playbook. It isn't anymore.

DeadDo instead
Per-lead AI-written copy / icebreakersHuman-approved template banks; AI fills variables only
"Quick question" subjects & openersGive-before-ask; ambiguous friendly subjects
Judging A/B at 50–100 sends500–1,000/variant, reply-rate, one variable/week
Apollo-first sourcing · free contact scrapers as the main sourceBuying signals: Sales Nav Spotlights (Job Opportunities + funding + headcount growth) and job-board actors, exported pay-per-result at ~$0.01/lead
Generic database lists · "under 5% bounce is fine"Buying signals (hiring / funded) · verified-only finder → MillionVerifier · valid only · under 2%
Google Workspace inboxes · 30-day warmup · SMTP poolsZapmail ~$3/mo · 14–21 days or pre-warmed · re-bid quarterly
⚠️ 30–40 sends/day per mailbox · 900/day fleet20/day per mailbox (ceiling 25) · 600/day fleet. ESPs tightened in 2026 — this is the most load-bearing correction on the page
⚠️ Paid Clarity Session sold from coldFree 15–20 min working session from cold, arriving with work already done. Paid diagnostics sell to warm audiences only — inbound, referral, someone who already knows you
Asking a stranger to explain their business to youAssume the problem. State it as true; let reply data tell you which ones are real
Judging a reply by how it is writtenQualify on the call, never on the reply. Terse lowercase "sure" becomes thousands of dollars
Fully autonomous reply/send botsDraft + one-click human approve, always
n8n/Make for new builds · agent wrappers · heavy skill librariesClaude Routines/scripts · few skills for repeated tasks
$1.5–3k/mo SMB management retainersBarbell: <$400/mo automated OR high-ticket Big-Head-Long-Tail
Hourly / monthly time-based build pricingFixed price + guarantee + price cap
Vector DBs / RAG rigs for business memoryMarkdown wiki + routing (level 2)
Personal branding from day oneShadow outreach + 2-hour minimum presence; content after customer #1
Selling "savings" automationsRACE revenue systems; reframe hours as revenue via opportunity cost

08The build queue

In priority order. Everything feeds, measures, or closes the funnel — nothing here is a new product.

  1. Funnel metrics trackerThe caveman spreadsheet, per lane, in System 0. The instrument that runs Sundays and the day-21 lane verdict.
  2. Single-artifact proposal machineProposal + contract + Stripe + signature in one doc, video walkthrough, payment auto-books onboarding, automated follow-up until yes/no.
  3. Reply queue (light)Five-bucket triage → voice-matched drafts from the approved bank → one-click approve. Possibly a Claude Routine. Never auto-send.
  4. Lead-pull scriptApify actor + MillionVerifier chained into the existing pipeline. Softens the Apollo bottleneck.
  5. Copy factory (reframed)Fable co-writes per-lane template banks — sequences, disclosure lines, subjects, objection replies, DM scripts. You approve once; the pipeline fills variables.

Shelf (build after customer #1)

Daily client check-in bot · onboarding/welcome routine · reactivation campaign ("we owe you $175" → case study 3 days later) · client reporting. Never build: sending infra, a CRM (spreadsheet first), dashboards nobody opens, framework wrappers.

LoopSuit AI · Operator's Guide 2026 · compiled 2026-07-11 by Fable · sources: Saraev main + Daily Updates (newest ~55 videos), Herk (newest 20), Hormozi (310-note vault + 2026), LoopsuitBrain wiki · full-context MD files: Md docs/agency-guide-2026/ · strategy: the Postmortem & War Plan artifact (v3.1) · re-verify against new videos every 4–6 weeks

Mentor mode · built around IST, FSH, Axiom, the gym, and a 2am hard stop

The operating day.

Same energy you already have — pointed in the right order. The rule underneath everything: promotion → delivery → building. Outreach happens before FSH; building happens only after both.

D1The honest read (your mentor's notes)

KeepThe crunch — move it earlier

Your one-day deep-shift style produces work Sandeep calls the best he's seen. Don't fight it — re-time it. New rule: the day a deliverable lands, open the file for 30 minutes (starts kill procrastination), then schedule the crunch within 48h — not at the deadline. Same quality, plus a buffer day, and "reliably early" is the exact evidence Sandeep needs for the $5k → $10–11k conversation after your 45 days.

AnchorSleep is a schedule, not a virtue

Sandeep already told you: end at 2am IST — that's still 4:30pm EST, nobody in Canada thinks you left early. The anchor is the wake time, fixed at 11:00, seven days. Everything else flexes; that doesn't. The 16-hour binges feel productive and cost you the next day — you've been paying 2 days for 1.

TruthThe Maker School gap is sends, not skill

The people posting $10–20k MRR at month 4–6 in Nick's community are not better designers, better builders, or better positioned — most can't design a button. The single variable between you and them is emails out the door. You felt relief when Leaptools ended because you already knew this is the right game. Now play it daily.

FixThe pedestal problem

Kinetic-typography videos that took days and got 3 likes — that's polish spent where volume was needed. New law: post more, polish less. A screenshot + a real number + 15 minutes beats a produced video. Proof accumulates; virality is irrelevant. (The LinkedIn tab has the posts pre-written so this costs you minutes.)

AxiomCapped, not killed

Axiom is real work with a real partner — it gets one deep day (Saturday, ~8h cap) plus one weekday evening. What it never gets: hours before the reply queue is done, or "one more test" past 2am. It's gated from the income goal by your own rule — the calendar should say so too.

JoyBuilding is the dessert

You don't need discipline to build — you need permission to. So the plan gives it to you: every day ends with a build block after promotion and delivery are done. Building as reward is sustainable; building as avoidance was the rut. And Cortex/portfolio work lives on Friday, scheduled, guilt-free.

D2Unblock this week (in order — #1 is today)

  1. Reactivate Instantly (today, 10 min)The card failed 4 days ago — every paused day is warmup decay. Fix billing first; nothing else works without it.
  2. Fix the Zoho app passwords (today, ~1h)Per mailbox: Zoho → Security → App Passwords → generate new → reconnect in Instantly → confirm warmup restarts. Warmup died, so assume the Zoho fleet needs 2–3 weeks of re-warm before real volume.
  3. Bridge with pre-warmed inboxes (today, 20 min)Don't wait 3 weeks to send. Buy 3–6 Instantly airmail / pre-warmed mailboxes (~$3–5/mo each) — the 2026 doctrine: send day one at ~1.35× true cost. Lane A starts on these while Zoho re-warms.
  4. The [VERIFY] pass (30 min)Open the copy banks, search [VERIFY-PERSONAL] and [VERIFY-PRICE], approve or edit. This is the only thing between you and loaded campaigns.
  5. Pull the lead list (~1h)Signal first: companies hiring a product designer, or funded in the last 12 months. Sales Nav Spotlights and job-board actors, as two comparable lanes → 100-row QA gate through Claude (≥50 or fix the filter) → verified-only email finder → triage → MillionVerifier. Load valid only. Not a free contact scraper — pattern-guessed addresses are what bounced UnitOS at 68%. Full chain in the Inbox Engine tab.
  6. Load campaign, schedule sends (~1h)Two-step sequence, 20/day/mailbox (ceiling 25), provider matching ON, open tracking OFF, "sent via Instantly" OFF. The moment sends are scheduled, LoopSuit has a live funnel for the first time ever.
  7. Start the caveman sheet (20 min)A Google Sheet beats waiting for the System-0 build: per lane per day — sends, replies, positive, meetings booked/had, proposals, closes. Fill it nightly at shutdown.

D3The day (IST) — illustrated

■ LoopSuit■ FSH■ Health / life■ Axiom
11:00
☀️Wake — fixed, 7 days a week—

Phone allowed for ONE thing from bed (Nick's habit): scan overnight replies, flag the hot ones. No building, no feeds. Flag generously — a four-word lowercase "sure" is not a signal of anything; you qualify on the call, never on the reply.

11:30
🏋️Gym60 min

Before the laptop opens, or it doesn't happen — evenings always get eaten. This hour is also where call energy comes from; Clarity Sessions are performed, not attended.

12:30
🍳Food · shower · reset45 min

Eat like someone with an 8pm sales call.

13:15
🎯THE POWER BLOCK — promotion first, always90 min

1. Reply queue — every outstanding lead, and book the ugly replies too (20m) → 2. Three Upwork applications, each with a Loom or a doc attached (25m) → 3. 10–20 founder DMs / connects (25m) → 4. Instantly + sheet glance, bounce under 2% (10m) → 5. ONE campaign improvement, however small (10m).

The Upwork line is new and it is the most important addition to this day. It is the only item here with no warmup clock in front of it — pre-qualified buyers who have already said they want to pay. Justin Lob ran exactly this streak alongside a 9-to-5 and closed $5,000 on day ~20. Do not break the chain.

14:45
🎨FSH deep block4.5–5 h

Canada is asleep — zero pings, pure make-time. Deliverable rule: same-day 30-minute open, crunch within 48h of assignment, ship a day early. Early is the raise argument.

19:30
🍛Dinner · people · off-screen60 min

Real break. The evening block works because this one exists.

20:30
📞US window opens — calls & flex block3 h

20:30 IST = 11am ET: Clarity Sessions, partner calls, and hot-reply conversations book HERE. No calls booked? This becomes the build block — Mon/Wed: LoopSuit build queue (tracker → proposal machine → reply routine). Tue: Axiom evening. Fri: content — but capped hard at 30 minutes and aimed at proof, not volume (one case study or one build receipt beats five posts; Saraev and Herk both put the daily-content threshold at $10–20k/mo, which you are below) + Cortex/portfolio time, guilt-free.

23:30
🤝Canada overlap — FSH sync & sends2 h

2pm ET: Sandeep, standups, feedback rounds, anything collaborative. Light work — reviews and replies, not deep design. US lanes are mid-morning; watch live replies land.

01:30
📊Shutdown ritual20 min

Fill the caveman sheet → note tomorrow's ONE campaign improvement → close every tab. The numbers you log here are the Sunday determination and the day-21 lane verdict.

02:00
🛑HARD STOP—

Sandeep's words: "we love the work — prioritize your sleep." 2am IST is 4:30pm ET; nothing you do after this hour survives contact with tomorrow. Lights out by 2:45.

D4The week

DayPower blockFSHEvening (20:30–23:30)
Monoutreach 90mdeep blockcalls / build queue
Tueoutreach 90mdeep blockAxiom evening (3h cap)
Wedoutreach 90mdeep blockcalls / build queue
Thuoutreach 90mdeep blockcalls / proposals out
Frioutreach 90mdeep blockcontent 30m Cortex / fun build
Satoutreach 60m—AXIOM DEEP DAY (8h cap, ends 23:00)
Sunoff till evening—determinations 45m nice time — actually off

Rule  The power block happens 6 days a week even on Axiom day, even mid-FSH-crunch, even on a Canada travel day — it's 60–90 minutes and it is the business. Miss everything else before you miss it.

D5The $10k CAD math (why this plan gets you there)

The gap is smaller than it feels

Floor: FSH $5,000 CAD. Gap to goal: $5,000 CAD ≈ $3,600 USD/mo. Two paths close it — one is capacity-capped, one isn't.

Path A — Services (capacity-capped). At revised pricing, ONE $12,000 build per quarter plus two AI Ops retainers at $2,500 closes the gap — versus 2–4 builds per month at old pricing. Same revenue, roughly a third of the delivery load.

Path B — Rank Radar (uncapped, the actual ceiling-lifter). 8–13 North American fleets at $1,500–2,500/mo = $12,000–32,500/mo recurring, with near-zero marginal delivery cost per additional customer. This is the only path in this guide where the tenth customer costs roughly the same to serve as the first.

The honest sequencing. Path A funds the runway. Path B is the business. Fleet Audits ($600) are the wedge into Path B — they are sellable today with only rank capture working, they generate the case studies, and they qualify buyers who can approve recurring spend. You don't need a miracle; you need the power block to happen ~78 times in a row.

Sept  Canada trip (Sept–Jan, T4 window): build the phone-only version of the power block NOW — reply queue and DMs work from anywhere; sends run themselves. The system should not notice you changed continents. That's also the test that it's a business, not a laptop habit.

Caveat  This holds for the FUNNEL, not for delivery. Sends, reply queue, and DMs are location-independent. Builds are not. Before September, either (a) have Rank Radar subscriptions carrying enough revenue that build volume can drop, or (b) plan the trip around a delivery gap. The system travels; the build capacity does not.

Proof posts, not thought leadership

LinkedIn: the builder who ships.

The positioning: never say you're one of the best product designers — post work that makes people conclude it. Real products, real numbers, zero AI-guru vocabulary. One post a week, 30 minutes, after the power block. Companies should discover you've been building in silence.

L1The rules

Cadence1/week, timeboxed

Friday evening block, 30-minute cap, post lands 6:30–7:30pm IST (= morning ET, your audience's coffee scroll). Content is channel #6 — it never displaces outreach.

FormatThe first line is the post

LinkedIn truncates at ~2 lines — the hook must survive the cut. One visual (real screenshot > produced video — the kinetic-typography lesson). Short paragraphs. No hashtag salad (0–3 max). Write to ONE reader: a founder or owner, not "my network."

BannedThe guru voice

No "🚀 excited to announce," no "AI won't replace you, people using AI will," no lessons-listicles about tools you haven't shipped with. If a sentence could appear on anyone's feed, it doesn't belong on yours.

EngineComments beat posts

10 minutes after posting: reply to every comment. Daily: 3–5 substantive comments on posts by design/AI builders your buyers follow — that's where profile visits actually come from.

L2The post bank (ready — personalize the [brackets], attach the screenshot, ship)

These are raw material, not scripts. The guide's own rule is post more, polish less — thirty pre-written posts become thirty posts you perfect instead of thirty days of shipping rough ones. Take one, spend fifteen minutes on it, ship it in whatever state it's in.

post 1 · rank radar (the product story)Restaurant brands live and die by where they rank on Swiggy and Zomato. Almost none of them can see it. So I built Rank Radar: it checks every outlet's rank on both apps, daily, by locality — against the 3 nearest competitors — and flags the moment something slips. Rating dip, broken listing, delisted bestseller: HQ knows the same day, not three weeks later through soft sales. Some things I learned building it: → Rank isn't a leaderboard, it's a tap. It drips orders when you're up and silently stops when you're down. → The painful cases are never the bad outlets. It's the good outlet that slipped to page 2 while everyone watched the sales report. → Ops teams run mystery diners for kitchens — and nothing for listings. Built with a Playwright polling engine, a fleet-view dashboard, and an owner-level overview page. Designed and shipped solo, end to end. [screenshot: fleet view] If you run a multi-outlet brand and want your outlets audited against your competitors — free, report either way — my DMs are open.
post 2 · the outbound engine (honest framing — it's yours)I spent the last months quietly building an outbound engine for my studio: ~30 mailboxes across 10 domains, a 7-stage lead pipeline that cleans, enriches, and personalizes at scale, a dedup engine, and a reply system where AI drafts and a human approves every send. Capacity: ~900 emails a day. The same category of system I build for clients who need pipeline. The counterintuitive lesson from the research phase: in 2026, LESS AI in cold email works better. AI-managed inboxes pattern-match hyper-optimized copy as spam instantly. So the system's rule is human-approved templates, AI fills only small variables — a machine that sounds like a person because a person wrote it. Infrastructure is a commodity. Taste in what to send isn't. [screenshot: pipeline dashboard, numbers blurred] If your business depends on outbound and it's all manual, this is exactly the kind of system I build. DMs open.
post 3 · cortex (the saas + waitlist cta)I gave myself one sprint to answer a question: can one designer, working alone with AI agents, ship a real SaaS — not a demo? The result is Cortex: an Obsidian-style knowledge graph in the browser. Your notes become nodes, connections become visible, and an AI layer helps you find what you actually meant to remember. What "designed and built solo" means now: → Design system, interactions, graph physics: mine → Backend, auth, data layer: built alongside AI agents → Time from first sketch to working product: [X days] The part that changed how I work: I stopped treating AI as autocomplete and started treating it as a build partner with infinite patience and zero taste. Taste stays my job. [screenshot/gif: the graph view] It's live behind a waitlist → [cortex-join link]. Referrals move you up the queue. I'd genuinely love to hear what you'd want a visual second brain to do.
post 4 · the silence post (proof dump — highest ceiling)I shipped 7 products in the last year and posted about almost none of them. Fixing that today. While working full-time as a product designer, I built — design AND code, end to end: → Rank Radar — Swiggy/Zomato rank monitoring for restaurant chains → Rebound — lapsed-client reactivation for med spas → Cortex — a visual second-brain SaaS (live waitlist) → A 600/day cold outbound engine with AI-drafted, human-approved replies → A legal case-tracking automation (Python + Claude + Telegram) → An internal collaboration hub for a design agency → An interactive proposal engine that turned client docs into experiences None of this went on LinkedIn because I kept telling myself "after one more feature." That was the wrong instinct. Proof compounds only when it's visible. So: one build, posted properly, every week or two. Starting now. [image: 7-product collage / grid] Which one should I break down first? 👇
post 5 · the ai-era designer workflow [VERIFY with Sandeep before naming FSH work]A client deck went through 40+ revisions across a team. Nobody could answer "what changed since Tuesday?" So instead of writing a status email, I built a tool: a visual diff for slide decks. Every version compared side by side, every change logged, all front-end, shipped in an evening. This is the part of "AI changed design work" nobody posts about. Not generating images — compressing the boring 20% of every project (status, changelogs, handoffs) into tools that take an evening to build and outlive the project. My rule now: if I explain the same thing twice, I build the thing that explains it. [screenshot: the diff tool, client work blurred] Designers: what's the status-update you're most tired of writing?
post 6 · rebound (the med spa problem)Most businesses think they have a lead problem. Usually they have a leak problem. A med spa I looked at had ~2,400 client records. Around 700 of those people came two or three times, liked the service, and quietly stopped coming. Nobody's job was to bring them back. At a $180 average ticket and a conservative 7% win-back rate, that's roughly $26,000 a year — sitting in the database, from people who already trusted them once. So I built Rebound: it segments the drifted clients by treatment cycle and spend, then re-engages them at the right moment — in the front desk's name and voice, with every message owner-approved before it sends. Nothing robotic goes out. Reactivation is the cheapest revenue most businesses have, and almost none of them work it. [screenshot: the segments view] If you run an aesthetics practice and want your actual number, I run it in a short working session. DMs open.
post 7 · the career arc (personal — highest engagement)Three years ago I was designing mobile banking screens for TD Bank, used by millions of people, and watching a lot of good design die in the handoff to engineering. I kept thinking: what if there was no handoff? What if the person who designed it just... built it? So I learned to ship the whole thing. Today I design AND build production products solo — Rank Radar, Rebound, Cortex, and internal tools for the agency I work with — with AI as the build partner and my taste as the thing that doesn't get outsourced. I've had roles end. The honest truth is that the last time it happened I felt relief — because I already knew I wanted to build things I control. Still designing full-time. But building in the open now. This is the start of that.
post 8 · court tracker (automation case study)A lawyer I built for was losing hours a week refreshing case portals to catch updates. So I built a system that watches the case sources, uses AI to read and summarize every change, and sends a plain-English update straight to Telegram — "Case X moved, here's what changed and what it means" — no portal, no dashboard, no app to open. Built with Python, the Claude API, and a Telegram bot. The pattern I keep coming back to: the best automation isn't a new tool you have to check. It's the work quietly showing up where you already are. What's the recurring thing in your week that a system should just... handle?
post 9 · interactive proposals (the craft flex)Most proposals are a PDF nobody finishes reading. For a recent engagement I built the proposal as an interactive experience instead — the client's problem quantified in their own numbers, the plan as something you move through, the pricing framed against the cost of doing nothing. Designed and built as a real web experience, not a slide export. Here's what I've learned: a proposal is a test. The client is silently asking "does this person understand my problem better than I can explain it myself?" Spend 80% of the document proving you do, and the price stops being the conversation. [screenshot: the interactive proposal]
post 10 · the "less AI" lesson (thought-provoking, not guru)Everyone's racing to put more AI into their outreach. After building a cold email system this year, I think that's backwards. Inboxes are AI-managed now. They pattern-match "perfectly optimized, every sentence is value" copy as spam almost instantly — because that's exactly what spam looks like in 2026. The emails that actually land are the ones a human obviously wrote. A little rough. One oddly personal line. No polish. So the system I built uses AI for the unglamorous parts — finding leads, enriching, routing replies — and keeps a human on the actual words. AI as the infrastructure, not the voice. The counterintuitive rule: the more AI floods a channel, the more the human touch is worth.
post 11 · building an AI operating system (the systems flex)I spent a while building myself a second brain. Scraped ~800 transcripts from the best operators I could find, distilled them into 786 linked, searchable notes, and wired it so I can query my own library while I work — "what does the data actually say about X" — and get a straight answer with sources. The lesson wasn't about note-taking. It was this: the bottleneck was never information. It was getting the right piece of it into my hands at the moment I needed it. That's what most "AI" should be doing for businesses — not generating more, but retrieving the right thing at the right time.
post 12 · what solo-building 7 products taught me (listicle done right)I shipped 7 products solo in the last year while working full-time. Not a flex — most people never post the lessons, so here are mine: 1. Building was never the bottleneck. Distribution was. I could build anything; I just wasn't telling anyone. 2. Taste is the moat. AI made building cheap, which means the thing that's scarce is knowing what's worth building and what "good" looks like. 3. Ship ugly, then polish. My most-polished project got 3 likes and took a week. Proof compounds; virality doesn't. 4. The boring 20% — status updates, changelogs, handoffs — is where the best small tools hide. 5. One document beats three. Every extra step between intent and money loses people. Which of these do you want the full story on?

Order  Post 4 (the silence post) goes first — it retro-frames everything and starts the series. Then roughly weekly, alternating a product post (Rank Radar, Rebound, Cortex, Court Tracker) with a lesson/personal post (the career arc, the "less AI" lesson, the 7-products list) so the feed never reads as one long ad. The med-spa math post (in linkedin/profile-and-pinned-post.md) stays pinned throughout. Anything touching FSH client work: clear the framing with Sandeep first.

The long game · 20–30 min/day cap · never displaces outreach

𝕏: build in public, quietly relentless.

Goal: in 12 months, be a recognized design-led builder in the Canada/US/India AI space. Not a growth hacker — the person whose ship-logs people check. X is where builders actually watch each other work; LinkedIn is where their companies find out.

X1The system

BioIdentity first

"Product designer who ships the whole thing. Building Cortex (visual second brain) + AI systems for real businesses in public. ex-TD Bank." Pin the Cortex thread once it exists.

CadenceSmall, daily, honest

1 ship-log post/day (3–5/week is the floor) — screenshot + one true sentence, 5 minutes, from the Friday/evening build blocks. 5 replies/day to builders in the design×AI space — replies are where followers actually come from at zero. 1 longer thread every 2 weeks (a real build breakdown).

PillarsFour, rotate

① Ship logs (what got built today, with the screen) ② Before/after UI (designer advantage — nobody else's logs look good) ③ Receipts (real numbers: sends, ranks, waitlist count — including the bad ones) ④ One-lesson posts (a specific thing learned building, never generic advice).

BannedThe BIP clichés

No "Day 1 of building 🚀," no follower-count posts, no engagement bait ("agree?"), no faked traction — the Cortex rule (no fake counters) applies to the feed too. Bad numbers posted honestly build more trust than good numbers implied.

FlagshipCortex is the show

Cortex is the perfect BIP vehicle: visual, consumer-facing, waitlist to point at. The agency work appears as anonymized systems posts (the outbound engine, the reply queue). Axiom stays off the feed until it's profitable — your own rule.

TruthHonest expectations

X compounds over 6–12 months and pays in relationships and inbound credibility — not clients this quarter. That's why it's capped at 30 min/day and lives AFTER the power block. 300 real builder-followers who trust your logs beat 10k passive ones.

X2The post bank — 17 ready posts

These are raw material, not scripts. The guide's own rule is post more, polish less — thirty pre-written posts become thirty posts you perfect instead of thirty days of shipping rough ones. Take one, spend fifteen minutes on it, ship it in whatever state it's in.

Paste as-is; swap only the [brackets] (a number, a link, or the media to attach). Lowercase-casual is intentional — it's the native X builder voice. Space them out: one a day, never two of the same pillar back to back.

01 · the opener (pin this)product designer for 7 years. the last 12 months I quietly shipped 7 products — design AND code, solo, around a full-time job. i'm going to post the build logs from here on out. real screens, real numbers, the failures included. first up: cortex, an obsidian-style knowledge graph SaaS, and the 900-emails-a-day outbound engine that runs my studio. no courses, no "AI will replace you" takes. just the work.
02 · cortex ship log — the physicscortex ship log. spent today making the knowledge graph *feel* right. the nodes were jittering — springs fighting the collision solver. fix wasn't more force, it was damping. now dragging a note and watching its connections follow feels like moving something physical. this is the 20% of the work nobody sees and the whole reason a graph tool lives or dies. [attach: 15-sec screen recording of the graph]
03 · cortex — the "why a graph" postevery notes app shows you a list. but your notes aren't a list. they're a graph pretending to be one — this idea connects to that one, this project touches three others, and you only remember the connection when you stumble back into it. cortex shows you the graph. notes become nodes, connections become visible, and an AI layer surfaces the link you forgot you made. building it in public. live behind a waitlist → [link]
04 · cortex — the build thread (every 2 weeks, longer)i built an obsidian-style knowledge graph SaaS in [X] days, solo, as a designer who codes. how it actually went, with receipts: 🧵 1/ premise: notes are a graph. every tool flattens them to a list. i wanted to build the graph you actually think in. [hero gif] 2/ design first. day one was pure feel — node physics, hover states, the little "aha" when a note snaps into a cluster. product taste is the moat now; the code is rentable. [before/after] 3/ the workflow: one expensive model orchestrates, cheaper agents build, i review like a design director with an infinitely patient team that has zero taste. taste stays my job. [screenshot of the loop] 4/ what broke: [honest failure + the fix] 5/ what's next: [next feature]. it's live behind a waitlist → [link], referrals jump the queue. logs continue tomorrow.
05 · rank radar — the alert that is the productthis one alert is the entire product: "[outlet name] dropped from #4 to #19 on zomato for 'biryani near me' overnight." restaurant brands lose orders like this constantly — a rank slip in one locality, at one time of day — and find out three weeks later through soft sales. rank radar watches every outlet on swiggy + zomato, daily, and catches it the same night. built the polling engine with playwright. [attach: screenshot of the alert]
06 · rank radar — fleet view designdesigning dashboards for people who manage 30 locations taught me one thing: the overview IS the product. built "fleet view" for rank radar — every outlet's aggregator rank on one screen, sorted by who's slipping, so an ops head sees the problem outlet in two seconds instead of opening 30 tabs. the hard part was never the data. it was deciding what NOT to show. [attach: fleet view screenshot]
07 · outbound engine — the reply queuemy inbox reply system now drafts responses in my voice from an approved bank, and i just tap ✓ or edit. handled 27 cold-email replies in 11 minutes this morning. every send was human-approved — nothing auto-fires — but the drafting is done for me. this is the boring infrastructure that actually makes outbound survivable for one person. [attach: reply queue screenshot]
08 · outbound engine — the counterintuitive lessoncounterintuitive thing i learned building a cold outreach system in 2026: use LESS AI in the actual emails, not more. inboxes are AI-managed now, and they pattern-match "perfectly optimized, every line is value" copy as spam instantly. the emails that land are the ones a human clearly wrote — a little rough, one oddly personal line, no polish. so my system's rule: humans write the templates, AI only fills small variables. a machine that sounds like a person because a person wrote it.
09 · the thesis post (pin-worthy)the designers who win the next 5 years won't be the ones who prompt the best. they'll be the ones who can hold taste AND ship the whole thing — design, frontend, backend, the deploy — without losing the vision in a handoff. i spent 3 years at a bank watching beautiful designs die in engineering translation. now i build the design and the product, end to end, and nothing gets lost. taste is the one thing AI can't rent you. everything else, it can.
10 · rebound — the reactivation mathevery med spa is sitting on money it can't see. ~700 clients who came 2-3 times, liked it, and quietly drifted. at a $180 ticket and a conservative 7% win-back, that's ~$26k/year — from people who already chose them once. nobody's job is to bring them back. so i built the system that does: segments the drifted list, messages them at the right point in their cycle, in the front desk's voice, every message owner-approved. reactivation is the cheapest revenue a business has. [attach: dashboard]
11 · proposal-as-one-document (a lesson)shipped a small thing today that i think about a lot: a proposal that is ALSO the contract, the payment page, and the signature — one document, one link. reason: every extra step between "yes" and "paid" leaks ~10% of buyers. separate proposal → separate contract → separate invoice = you lose a third of the people who already said yes. collapse the steps. the friction between intent and money is where deals quietly die. [attach: screen]
12 · court tracker — python + claude + telegrambuilt a legal case-tracking system for a client that lives entirely in telegram. python watches the case sources, claude reads and summarizes the updates, and the lawyer gets a plain-english ping — "case [X] moved, here's what changed and what it means" — instead of refreshing a portal. favorite kind of build: no app to open, no dashboard to check. the work just shows up where you already are.
13 · the visual diff tool (an evening build)a client deck went through 40+ revisions across a team and nobody could answer "what actually changed since tuesday?" so instead of writing a status email, i built a visual diff for slide decks — every version side by side, every change logged. all front-end, shipped in one evening. my rule now: if i explain the same thing twice, i build the thing that explains it. [attach: the diff tool]
14 · an honest failure post (post these — they build trust)spent 4 hours today chasing a bug that turned out to be me. [the outbound pipeline was silently dropping leads whenever a company name had an emoji in it — one unhandled character killed the whole row]. classic: i was debugging the clever part while the dumb part quietly failed. shipped the fix. lesson re-learned for the hundredth time: validate the boring inputs first. [attach: the fix, optional]
15 · cold campaign — week-1 receipts (post the real numbers)building in public means posting the numbers before they're good. med spa cold campaign, week 1: [X] sent, [X]% reply rate, [X] meetings booked. under my target of a meeting per ~150 sends. not panicking — you don't judge a variant under 1,000 sends. changing ONE thing (the subject line), sending 1,000 more, then reading the data. that's the whole game: volume, then judgment. [attach: sheet]
16 · building a second brain from 800 transcriptsi scraped ~800 youtube transcripts from the best operators i could find and fed them into a system that distills them into a linked, searchable knowledge base — 786 notes, cross-referenced, that i can query while i work. the surprising part: the value wasn't the notes. it was that i could now ask my own library "what does the data say about X" and get a straight answer with sources. a second brain that actually answers.
17 · the orchestration workflow (a "how i build" post)how i actually build now, as one person: i don't write most of the code. i orchestrate. an expensive model plans and reviews; cheaper agents do the heavy lifting in parallel; i sit in the middle as the director — deciding what's good, what ships, what gets torn up. it feels less like coding and more like running a studio where the staff never sleep and have no ego. my whole job is taste and judgment. that's the part that doesn't get automated.

Engage  The daily 5 replies matter more than the posts at zero followers: pick ~10 accounts at the intersection of design × AI × indie (peers slightly ahead of you, not mega-accounts), turn on notifications for 5, and reply with something only a builder would say. Almost nobody posting about AI can design, and almost nobody who can design ships — that gap is your whole edge.

Gamified · one block = one Power Block, run for real

The road to $10k, one run at a time.

Every block below is a real projection, not decoration — built from the same funnel assumptions as the rest of this guide. Drag the slider to any goal. Click a block to mark every run up to it done (click the frontier block to step back). Hover any block for what it's worth.

✓
Daily quest · 6× per week
The Power Block
⏱ 90 minutes · before FSH, always
🏆 Reward: +1 block on the grid · streak stays alive

🎯Set your target

Target monthly income (CAD, FSH + LoopSuit combined) $10,000

The funnel — one typical month at goal pace
How this is calculated (the honest assumptions)
This is a stage-by-stage funnel simulation, not one blended number. Per send: 85% deliverability (post-MillionVerifier, plain text) → 1.5% of delivered reply (realistic public-list rate) → 30% of replies are positive → 55% of positives book a call → 70% show up → 15% of held calls close a $12,000 build (was 25% at $1,997 — higher price, longer cycle, fewer closes; the tradeoff is explicit); 30–45% of held calls also buy a $1,500 Design Sprint add-on. Note (Sep 2026): this is no longer modelled as a paid session sold from cold — paid diagnostics only convert on warm traffic. From cold the session is free and the paid line item moves after the call. That chains to ≈0.2% booked-call rate (the guide's target band is 0.5–1%; 0.03% is "bad but viable") and ≈1 build per ~4,500 sends at maturity.

The real-life frictions are modeled too: copy maturity ramps 40% → 70% → 100% over the first ~9 weeks (you don't have winning copy on day one); sends ramp from ~60/day to the scenario's ceiling as your restarted warmup completes (full 600/day = 10 domains × 3 mailboxes × 20/day — the 2026 per-mailbox limit; 30/day is deprecated); there's a ~8-week pipeline lag from send to cash (reply → booking → close window — higher ticket stretches this vs the old ~4-week $1,997 cycle); AI Ops retainers churn ~6%/mo; and — the binding constraint everyone ignores — solo delivery capacity is capped at ~2–4 builds/month alongside FSH (~20–25h of real work per build against ~10–12h/wk of delivery time). That cap is why a pure services path isn't infinite. Above it, builds flatten and growth comes from the AI Ops tail stacking ~2 retainers/month — which is why very high services-only targets take so long and eventually go unreachable.

Rank Radar at NA pricing ($1,500–2,500/mo per fleet) is the primary ceiling-lifter — near-zero marginal delivery cost per additional customer, so it is not bound by the solo-build cap. Raising prices and templatizing delivery (the Rebound recipe → 4–6/mo) help Path A; Path B is what actually removes the ceiling. The model still refuses to assume you'll have those Rank Radar fleets before you've sold them.

Two different numbers, both shown: "monthly run-rate" includes project cash flow (builds + sessions at current pace — it stops if you stop sending); "durable recurring" is only FSH + AI Ops MRR (it keeps paying if you take a month off). The run-rate curve is more pessimistic than the old $1,997/25% model on close count — higher prices mean longer cycles and fewer closes — while each close carries more revenue and roughly a third of the old delivery load. USD→CAD ≈1.39. A projection, not a promise: the point is the shape of the curve, not the exact digit.

Operating plan · 15 Sep 2026 → Feb 2027 · tick a quest, earn the XP

The Mission.

The Road to $10K tab models where the money comes from. This tab is what you actually do on which day to get there — India sprint, the Vancouver routine, the mail infrastructure, and the four-week calendar. Every box you tick is saved on this device. Nothing here needs a new product built.

1
Operator
0 XP · 300 to Sender
0Quests
0%Mission
0🔥 Streak
0Sent

01The one idea

Every plan you have had treated your channels as five separate jobs — a job hunt, a cold-email agency, a personal brand, the Shree partnership, the residency file. Five launches for one person around a full-time job, so each one gets a fifth of your courage and none of them ships. Collapse them into one asset with five buyers.

The single asset

A public, weekly, face-on record of you designing and shipping AI products.

One production line. You already ship this work — the only new act is publishing it. Everything below is a buyer for the same asset.

👤 Founders

Proof → fractional and fixed-scope contracts at $6–9k/mo. Design engineers who ship are the scarcest hire in 2026.

📧 Cold-email replies

Prospects Google you after Day-0. Today they find a ghost. A shipping record is the difference between a reply and silence.

🏢 Sandeep / FSH

A visibly senior public figure is worth a title, an assignment letter, and a credible "leads our India delivery" story.

🤝 Shree's network

METAVRASH's brand and tech arm gets a face. His clients see who branded them — he said so himself on 12 Sep.

🇨🇦 The residency file

A documented senior professional with real Canadian employer ties beats an invisible remote worker if anyone ever asks.

The second idea: stop looking for a job

FSH is not a salary — it is the T4 employer that lets days abroad count toward your PR residency obligation. That means you keep it, and "get a higher-paid job" is permanently off the table. Good: it removes the distraction and forces the wealthier path. The only way up is contracts and products stacked on top of FSH, priced like the design engineer you are. Two people to see in Vancouver make that safe — an immigration lawyer and a cross-border accountant.

02The prime rule

Sent > Built.

The audit found one failure mode repeated across eight channels: the hard, skilled, ambiguous work is always finished, and the undone step is always the one where a stranger can say no. 400 emails ever. 809 applications prepared, zero sent. 133 posts written, zero published.

So the metric changes. One number, daily: things sent to humans who can say no — emails, DMs, posts, proposals, invoices, meeting requests. A day with a shipped feature and a zero in this column is a zero day.

The Send Block runs before Cursor opens. Thirty minutes, first thing after the gym. Nothing gets built until it is done. Three Upwork applications are inside it, every single day — that is the one channel that can pay you before the mailboxes finish warming.

Sent today
0
—

03Phase 1 — the India sprint

15 Sep → 1 Oct. Seventeen days, FSH still running. Everything here fits in about 90 minutes a day plus one Sunday. Order matters — the top item is worth more than everything under it combined.

0%
India sprint15 Sep – 1 Oct · 17 days
Active now

Tue 16 – Wed 17 · Money and infrastructure

✓
Send Shree the priced scope letter
150 XP
Three paragraphs: GARUDIEN scope with a fixed number and payment schedule; METAVRASH rebrand plus the two dummy vertical packs as a separately priced line; one sentence that future METAVRASH equity should reflect an execution-partner stake. Ask a ₹2.5L/month retainer floor. Why first: he said "fee" on 10 Sep and the scope has grown every week since with no number attached. This single letter roughly doubles your monthly income and costs under an hour.
✓
Buy Zapmail · connect 3 domains · start warmup
100 XP
Full walkthrough in section 06 below. Why now: warmup is a 14–21 day clock that only runs once you click start. Buy it today and the fleet is ready the week you land in Vancouver. Every day you wait is a day added to the end.
✓
Publish LinkedIn p64 · schedule the next ten
80 XP
Unedited, no face clip — the clip is the excuse. Then use LinkedIn's own native scheduler to queue p65–p70. Why: 77 posts are written and the calendar is already seven days late. Scheduling removes the daily decision, which is the thing that actually fails.
✓
Start the Rank Radar week-sweep
60 XP
Lunch 13:00 and dinner 19:00 IST, full fleet, per WEEK-SWEEP-PROTOCOL.md. Why: Shree's gate for the Raymond meeting is one week of live data. Your data is from 2 June. This runs by itself while you do everything else — it is the cheapest unblocking on the list.

Thu 18 – Sat 20 · Proof a stranger can find

✓
Three case studies live on loopsuitai.com
120 XP
VeritasGuard, Rank Radar, and one brand system. The copy already exists as planner posts p69–p71 — this is a publishing job, not a writing job. One page each, real numbers, your photo at the bottom. Why: every buyer in section 01 Googles you. Right now they find a studio site with no case studies and no face.
✓
Two GitHub repos public and clean
70 XP
Cortex, plus one Next.js/Supabase starter. README with a screenshot at the top. Why: every founder considering a design engineer opens GitHub. An empty profile reads as "designer who says he codes." Two clean repos read as proof and cost an afternoon.
✓
Cal.com booking link live and in every bio
40 XP
Free. One 30-minute "Working session" event type. Why: this is how meetings get booked while you sleep — your reply bank already assumes a {booking_link} exists. Two concrete slots plus a link beats a bare link.
✓
Open the X account · pin the opener · queue 3 logs
60 XP
Follow 02-x-build-in-public/account-setup.md; schedule through Buffer or Typefully free tier. Why: X is where founders scout design engineers in 2026, and a new account needs volume before taste shows. 56 ship-logs are already written.

Sun 21 · The Sunday system — repeats every week from here

✓
Run the first Sunday batch (2 hours)
80 XP
Schedule 14 LinkedIn posts and 7 X logs · shoot 3 face clips on the phone from 03-media/shoot-list.md · read the scoreboard and write one line about what changes next week · recalculate the 0.75% trading risk unit (your own Monday rule). Why: this single session is what makes the other six days automatic. It is the whole answer to "I can't afford a VA."

Mon 22 – Sat 27 · Pipeline and the one named buyer

✓
Raymond meeting, with a live week of fleet data
130 XP
Shree in the room. Ask for the sales export — sales by store, lunch versus dinner, order source. Why: Raymond is the only buyer you have with a name and a decision. Do it in person while you are still in India; after 2 Oct it becomes a video call with a time-zone problem.
✓
Build the lead pipeline end-to-end for ONE lane
110 XP
Apollo pull → lead_triage.py → MillionVerifier → personalization → campaign loaded in Instantly and left paused, waiting on warmup. Why: build the whole chain once, now, so that in October the action is "unpause" rather than "build a pipeline." The lane is in section 05.
✓
Founder DMs begin — 10 a day, by hand
90 XP
YC W26/S26, Series A AI, fintech-AI. One specific line about their product, one line about what you would ship in two weeks, your Cal link. Why: this is the only channel with a same-day feedback loop, and it cannot be automated without risking the account. It is the one manual habit in the whole system.
✓
Five referral messages
50 XP
Sandeep, Shree, Vivek at Vital Wires, Himanshu, the Court Tracker client. One sentence: "who else should see this?" Why: referral is the highest-converting channel that exists and there is no record of you ever asking. Five messages, maybe twenty minutes.
✓
Book the lawyer and the accountant for Vancouver week one
70 XP
Immigration lawyer — "Does this FSH structure satisfy R61, and what must the assignment letter say?" Cross-border accountant — "BC PR, ~8 months a year in India, USD contract income plus a T4 — where should contract income land?" Why: the residency structure is the load-bearing wall of this entire plan and it is currently an informal understanding. The tax answer is worth 20–30% of everything you earn.
✓
Draft the FSH ask (section 07)
40 XP
Written now, delivered in person in Vancouver. Why: you will not improvise this well in the room, and everything in it is free for Sandeep to give.

Sun 28 – Wed 1 Oct · Pack

✓
Second Sunday batch · everything runs without you
60 XP
Posts scheduled through mid-October, warmup running, sweep running, campaign paused-and-loaded, DMs queued. Why: the flight and the first week of jet lag are exactly when a fragile system dies. Make the machine survive a week of your absence.

What wheels-up looks like

Shree priced · proof public · nine inboxes warming · one campaign loaded · DMs flowing · lawyer and accountant booked · Raymond met. Zero new products built. That is the whole test of this phase.

04Phase 2 — the Vancouver routine

2 Oct → February. Four to five months, and you want to come out of it in the best shape of your life and your career. Those are the same project: the gym is the mechanism. It puts a hard, physical, non-negotiable event before the Send Block, so the day starts with two wins banked before FSH even opens.

06:45
Up. Coffee, music.

The thirty minutes of messing about is in the plan, deliberately, and it lives here rather than leaking into the afternoon.

07:15
Gym · 1 hour

Three lifts, two conditioning, one long walk per week. Fixed slots so there is nothing to decide. Target: 90 sessions by February — the body follows attendance, not programmes.

08:15
Shower · Shree call window

08:15 PT is 20:45 IST — the only slot that works for both of you. Keep it free on the days he needs it.

08:30
The Send Block · 30 min · Cursor stays closed

Approve the drafted replies (book the ugly ones too) → 3 Upwork applications, each with a Loom or doc attached → 10 founder DMs → glance at Instantly (bounce under 2%, one small improvement) → log the scoreboard. This is the load-bearing habit of the entire plan, and the Upwork line is the only part of it with no warmup clock in front of it.

09:00
FSH deep work · 3.5 hours

Your job, and your residency. Protect it — this is the asset everything else is built on top of.

12:30
Lunch + 10 minutes of LinkedIn comments

Three to five substantive comments on posts your buyers already read. On a small account, comments out-reach posts — this is the cheapest distribution on the list.

13:00
FSH + Shree delivery, interleaved

One laptop, one IDE, many tabs — this is where the multitasking is legitimate, because both are paid work.

16:00
LoopSuit block · 90 min

Founder calls (16:00 PT suits US founders), proposals, Looms, one case study a month. This is the only slot where new building is allowed.

17:30
Yours.

Tuesday or Thursday is a meetup (section 08). Otherwise: fun, people, the joint, whatever you like. Evenings are not negotiable in either direction.

Sunday
The Sunday system · 2 hours

Batch the week's posts, shoot 3 face clips, read the numbers, reset the trading risk unit. The only weekend work there is.

One discipline rule, not ten

No joint before the gym and never before 17:30. That is the whole enforcement mechanism — one rule you either keep or don't, rather than a code of conduct you negotiate with daily. Phone out of the bedroom, screens off by 23:00, so that 06:45 is real rather than aspirational.

05The four-week calendar

15 Sep → 12 Oct, the two phases side by side. Click any day to tick it. The rhythm is the point: Monday to Saturday is the Send Block and the day's post; Sunday is the batch. The dated milestones sit on top of that rhythm, never instead of it.

MonTueWedThuFriSatSun

🔁 Every Mon–Sat

Send Block 08:30 · 3 Upwork applications · 10 founder DMs · 3–5 LinkedIn comments · log the scoreboard · gym. (Posting drops to 2–3 a week of real proof rather than one a day — content earns its hours above $10–20k/mo, and you are below it.)

🗓 Every Sunday

Batch 14 LinkedIn + 7 X posts · shoot 3 face clips · read the week's numbers · reset the trading risk unit.

📅 Monthly

One public case study · one long Loom walkthrough · pricing review (raise if you are closing more than one in three).

06Zapmail → Instantly, the right way

You own 10 Hostinger domains with the DNS wiped clean after Zoho killed 29 accounts. Start with 5 domains × 2 mailboxes = 10 inboxes, not all ten domains — same $39 Starter plan, one more mailbox than a 3×3 split, and the risk spread across five domains instead of three. The reserve five join once bounce has held under 2% for 30 days. Sending domains: loopsuitstudio · loopsuitlabs · loopsuitbuilds · buildwithloopsuit · shipwithloopsuit. loopsuitai.com never carries outbound risk — it is where buyers land after they Google you. Roughly $100–110 for month one, with Sales Nav free on the trial.

  1. Connect the domains in Zapmail, not Hostinger

    Zapmail → Connect Existing Domain → pick 3 of your Hostinger domains. Zapmail gives you nameservers; in Hostinger, replace the existing nameservers with theirs. Propagation takes 24–48 hours.

  2. Then never touch DNS again

    From that moment Zapmail owns SPF, DKIM, DMARC and MX and manages them automatically. Editing or deleting those records by hand is the single most common way people break this — leave them alone.

  3. Create 3 mailboxes per domain

    Real-person patterns: jastej@, j.sehra@, jastej.s@. Never hello@ or info@ — role accounts get filtered. Pick Google mailboxes; Gmail-to-Gmail placement matters for your ICP.

  4. Verify green twice

    Wait for green in Zapmail, then check independently at mxtoolbox.com. Zapmail's dashboard sometimes shows verified before propagation has finished, and connecting early is what poisons a fleet.

  5. Warm up 14–21 days · this is the step you keep postponing

    Warm inside Zapmail until the warmup score is above 80. It is one click to start and then it runs itself. Starting it on 16 Sep means the fleet is live the week you land. Do not connect to Instantly before the score clears 80.

  6. Export into Instantly

    Zapmail has a one-click export that authorises into your Instantly workspace and the mailboxes appear on their own. If it fails: Instantly → Email Accounts → Add → Custom SMTP/IMAP — SMTP port 587 STARTTLS (or 465 SSL if Zapmail specifies), IMAP 993 SSL for reply tracking. A port mismatch fails silently, so test both before you suspect the password.

  7. Cap the volume yourself

    Instantly does not enforce warmup limits on custom SMTP accounts. Set 20/day per inbox. The hard ceiling is 25, not 30 — ESPs tightened in 2026 and 15–20/day is now the normal operating range. Grow the fleet by adding inboxes, never by raising per-inbox volume. Keep Instantly's own warmup running alongside sending. 10 × 20 = 200/day.

  8. Campaign hygiene

    All 10 accounts assigned with inbox rotation on · 3-step sequence (Day 0 / +3 / +4) · open tracking off so replies and bounces stay the only honest metrics · plain text · no link in Day-0, Cal link in replies only · never load an address that has not come back valid from MillionVerifier. Target bounce is under 2% — above that sustained, Google penalises the whole domain; Instantly hard-kills at 5%; and you have seen what 60% does.

The timeline this produces

Buy now → DNS live +2 days → warmup ~14–21 days → 10 inboxes live at 200/day → add the reserve five domains once bounce holds under 2% for 30 days → 600/day (10 domains, 30 inboxes) by December — note that step also forces Zapmail Growth (+$60) and Instantly Hypergrowth (+$50), so it waits for proof. Do not raise per-inbox volume to get there — add inboxes. First statistically readable data (n ≥ 1,000 delivered) lands mid-November. Plan income from this channel in Q1, not Q4 — and judge nothing before then.

And do not wait for any of it: buy 3–5 pre-warmed inboxes today and start sending at ~60/day this week. Pre-warmed boxes are rented on the provider's generic domains — you don't own them, which is exactly why they are the right thing to burn while you establish a feedback loop. The Zapmail fleet is the brand asset; the pre-warmed ones are ammunition.

07One lane, and why the other four are dead

The lane: "An AI product designed and shipped in 2–4 weeks" sold to funded seed and Series-A founders and fintech. Apollo filter: raised in the last 12 months, 5–40 employees, US/CA/UK, no designer on the team page.

✕ NA QSR / Rank Radar

Sold Rank Radar to DoorDash restaurants. Your scraper only reads Swiggy and Zomato — the lane sold a product that does not exist. Dead.

✕ UnitOS franchise

No seats, no multi-tenancy, nothing to sell on a call. Dead until the product exists.

✕ Rebound / med spa

$297/mo DFY with high delivery cost. One close a month barely registers against your salary. Dead as a cold lane.

✕ LocalOps home services

Same economics, no product, no proof. Dead.

Why the new lane instead

$8–15k a close, so a single close is a real month rather than pocket money. You deliver it alone — it is literally the offer already written in your experience doc. Same ICP as the founder DMs, so email and DM run on one list and reinforce each other instead of splitting your attention. And the proof you are publishing in section 03 is exactly the proof this buyer wants to see.

The reply layer — the automation that actually matters

Sending was never your bottleneck; replying was. Build this once and the channel survives you being busy: Instantly categorises replies → interested ones land in Gmail → a Claude routine drafts the response from your approved reply bank into a Gmail draft → you read it and tap send from your phone during the Send Block. Nothing auto-sends, ever. Cal.com link in every reply so meetings book themselves. instantly_pull.py runs on a schedule and pushes you the weekly numbers instead of you remembering to check.

Sales flow — Saraev's three-decision rule

Working session → proposal inside 72 hours as one artifact (proposal + agreement + Stripe + signature in a single document) → paid. Three yes/no decisions, which is the Goldilocks zone; every extra step leaks roughly 10% of leads. Never sell a retainer cold — sell the fixed 2-week build first, and let maintenance follow it.

08The FSH conversation

Sandeep cannot pay more right now, and that is a real constraint you should accept gracefully. But everything below costs him nothing, and he has told you his team depends on you. Have it in person, in Vancouver, in week one — not over Slack.

① Title: Head of Product / Lead Design Engineer

Free for him. It roughly doubles your contract rate elsewhere, and a senior person "assigned abroad to lead India delivery" is a far more credible residency story than a designer working from home.

② A written assignment letter

India posting, defined term, an expectation of return to a Canadian position. Drafted with the lawyer's wording. This is your residency paper — without it the arrangement is an informal understanding.

③ Written permission for outside contract work

He already knows about LoopSuit. Getting it on paper means the T4 can never be disputed later, which protects the thing the whole plan rests on.

④ 10–15% finder's commission on clients you bring FSH

Then actually bring one Indian client in under your name. This is what makes "FSH India operations" a real thing rather than a description — which is precisely what the assignment letter needs.

⑤ A salary review with a trigger, not a feeling

"Next two contracts signed → $X." Turns "when we can afford it" into a date you can both point at.

⑥ Rev share if Creative Hub is ever white-labelled

A long shot that costs nothing to ask for while you are asking for the rest.

The risk nobody has priced yet

Regulation 61 lets days abroad count only when a Canadian business assigns you to a position outside Canada, with an expectation of return. PRs who were in substance working remotely from their home country by choice have been refused. The hole today is that FSH has no Indian operations — which is why item ④ matters as much as the letter. Spend $500–1,500 on one immigration lawyer consult before you fly. If you are building your whole career around this structure, you do not get to guess. Note too that this path does nothing for citizenship, which needs 3 physically-present years in 5 — a separate decision on a separate clock.

09Becoming the obvious hire

Saraev's own curriculum is about ten hours of skill-building and then 99% customer acquisition. You have the skill a hundred times over and have done roughly 1% of the acquisition. Neither he nor Hormozi is better than you at building — they are seen. Here is the smallest system that fixes that.

Daily · LinkedIn post

Scheduled on Sunday, so no daily decision exists. Buyers, Sandeep and Shree's world all see you every day. 77 posts already written.

Daily · X ship-log

Lowercase, specific, one screenshot. Founders scout design engineers here, and it warms up every DM you send.

Weekly · one 60–90s "build receipt"

Screen plus your face, one cut, no kinetic typography. One shoot, four surfaces — LinkedIn native, X, YouTube Shorts, Reels. Face is the trust text cannot buy.

Monthly · case study + long Loom

A 10–15 minute "how I shipped X." This is the asset founders watch before they book, and it compounds in search.

Biweekly in Vancouver · speak at a meetup

The underused lever. A 10-minute lightning talk — "I designed and shipped a banking-grade AI product alone in 14 days." Organisers are always short of speakers. It puts you in front of 50 founders at once and produces a clip. Nobody in your position does this.

Once · one lead magnet, capped at a day

A "2-week build scoper": founder pastes an idea, gets a scoped spec and a price band, leaves an email. It is the {tool_link} your reply bank already assumes exists.

The positioning line — use it everywhere, unchanged

"Design engineer. I design and ship AI products end-to-end — ex-TD Bank mobile, LoopSuit AI."
Not "Senior Product Designer" (prices you as an agency designer). Not "AI studio founder" (prices you as an agency). The category you belong to did not exist as a hiring line three years ago and is now the scarcest seat at AI companies: everyone has designers who cannot ship and engineers who cannot design. You are both, with a bank on your résumé. AI tooling note: Higgsfield Soul for product b-roll only; your real face for anything about trust; a HeyGen avatar never speaks as you.

10The money ladder

All figures USD per month. The point of this ladder is that the base roughly doubles from one letter — before a single cold email is sent or a single contract is signed.

FSH net$3,900 CAD · unchanged, and that is fine
~$2,850
Shree retainer₹2.5L floor · ask this week
~$3,000
Base, Octoberalready 2× today, no new clients
~$5,850
First contract clienttarget 15 Nov
+$6–9k
Second client or cold closeDec–Jan
+$6–9k
Rank Radar tenant via Shreehe sells, you build, split
+$0.5–2k
January target≈ $20–27k CAD
$15–20k

The engine nobody has been counting: geo-arbitrage

You earn in CAD and USD and live in India roughly eight months a year. At the January target, against Indian cost of living of $1.5–2.5k, the savings rate is $12–16k a month — $150–190k a year. Held in a properly structured Canadian corporation at small-business rates (the accountant's call, not mine), three years of that plus one product at modest MRR is seven figures with no equity event required. Shree's equity and the trading engine are upside on top of that, never inside the base.

11Decisions already made

These are settled so they stop consuming attention. Reopen one only with a reason, not a mood.

Renewal Radar — shelved

Its own odds document says the number-one lever is ten in-person broker demos a week in Vancouver. You leave that market in February and live in India eight months a year. Sharp market read, wrong operator. Sunk cost: one day. Revisit only if a broker network approaches you.

Rank Radar — Shree's to sell

You build, he sells into his network, you split. Stop owning its distribution; his relationships are worth more than your outbound.

ApplyOS — repurposed, not deleted

You cannot take another job without breaking the residency structure. So it becomes a contract scout: filter to contract, fractional and founder-led companies, output DM targets. The 809 prepared applications get archived, not sent.

HJ Axiom — contained

Your own two-engine plan already has the rules: 0.75% risk, weekly recalculation, ratchet down at −15%, nothing withdrawn for six months. It stays automated, gets one touch a week, and never enters the income base until it has actually paid out.

No VA until $8k/mo blended

Until then, scheduling tools are the VA — LinkedIn native scheduler, Buffer, Claude routines. Revisit the month you cross it, not before.

No new products

Nothing new gets started until one existing thing has taken money from a stranger. Your own July operating principle, and the one you break most often — ten products, zero paying customers.

12Trophies

These unlock on their own as you tick quests and log sends. They are the milestones that actually change the shape of the year.

Phone pack · Vivek Singh / Vital Wires · synced to 18-slide peak-first deck

VW meeting script

Story first, roadmap second, proof always at hand. Don’t read slides. Close with a starting stage — not a price. Max 3 spoken sentences per slide — he interrupts.

Deck 18 slides Intro Shree → LoopSuit AI Ask Map → Propose → 48hr

Open the deck on the other screen. Speak from here.

Open deck ↗

Mindset

Vivek is ~50, consulting veteran, low retention, will interrupt — that’s engagement, not rejection. When he talks, stop presenting and converse; re-enter the deck at the nearest chapter.

You are the specialist he’s evaluating, not a vendor pitching up. Story first, roadmap second, proof always at hand.

Opener

One line of context — then straight into slide 01. Do not open with a bio.

“Shree and I have worked together on consulting engagements for a while — he’s seen my work up close.”

Then open slide 01. No bio dump. Frame: “I’ll start with one finished story — then we bring it back to Vital Wires.”

Room timing · ~20–25 min

~1–2 min — opener (Shree line) → slide 01
~7–9 min — Before → Rebuild → Guidelines → Hub (let him poke)
~2 min — The gap (“one method — here’s why it matters”) — stop if he objects; handle live
~4–5 min — Plan + skim Create / Launch / Control / Sell (don’t lecture every bullet)
~1 min — Channels: click website mock + SAP RISE Roundtable only
~3–4 min — Lead engine (+ Franchisee Shift / selected work if energy is high)
~3 min — How we start → How we work → The Ask → silence

If he hijacks

Stop presenting. Converse. Re-enter at the nearest chapter (Gap, Plan, or Ask). Full recovery lines live in VW Cheat → If he interrupts.

Slide speak lines

Aligned to the live 18-slide peak-first deck. Each slide: say → deeper ammo → if he asks. Don’t read — skim and speak. Max 3 spoken sentences.

01 Hero

Say

  • “Before I explain anything, let me just show you a piece of work — because it says more than a deck ever will.”
02 Blenz · Before

Say (~60–75s)

  • “This is Blenz — a Canadian beverage brand. History, stores, franchise footprint — the delivery was there.”
  • “The surfaces weren’t. This is what the market was seeing.” (Pause — let him look at both frames.)
  • “Four things they needed fixed: no scalable system — every asset reinvented; no brand guidelines — teams guessing; clinical look — safe, dated, not Gen Z; and the ask wasn’t a logo tweak — it was a full rebrand. That’s when I came on board.”
  • Bridge: “Twelve weeks later — this is what we shipped.”

Deeper ammo

  • Don’t trash the old brand — diagnose. Delivery was there; surfaces hadn’t caught up (mirrors Vital Wires).
  • Uncropped frames matter — he has to see clinical before the rebuild.

If he asks

“Why show a coffee brand to an enterprise firm?”

“You’re not buying coffee branding. You’re seeing how deep a brand system goes when one senior owns it end-to-end. Same discipline — different buyer.”

03 Blenz · Rebuild

Say (~45–60s + let media breathe)

  • “Twelve weeks. A complete rebuild — not a refresh.”
  • “One system end-to-end: packaging, merch, store interior, and the rules that hold it together.”
  • “This is the creative quality bar I’d hold Vital Wires to — enterprise substance deserves this craft on its surfaces.”

Deeper ammo

  • Every asset from one system — nothing designed twice.
  • Don’t narrate every frame — let the paper-bag / system media play.

If he asks

“You did ALL of this?”

“Design and direction, all mine. Production scales through my team and my tools — that’s coming next.”

04 Blenz guidelines

Say

  • “Not a logo drop — a complete brand book you can scroll.”
  • “Identity, packaging, merch, interiors — and the rules that keep it consistent.”
  • “Click inside and scroll if you want — depth of system, not a moodboard.”

Deeper ammo

  • This is what makes brand ONE-TIME to establish and ONGOING to enforce — the distinction Vivek’s world already thinks in.
05 Creative Hub

Say

  • “Then I built the tool their whole leadership works in.”
  • “Briefs, reviews, approvals — one owned place, not chats and folders. I designed it AND shipped it.”
  • Bridge: “Same pattern I’d install for Vital Wires brand work. Now — why this matters for you.”

Deeper ammo

  • Bespoke, not another Kanban. The agency owns it outright.
  • Feedback cycles: days → hours.

If he asks

“Who is this for?”

“A Canadian design agency I work with — their CEO and full leadership are active users.”

06 The gap

Enterprise delivery. Weak market signal. · 01 Referrals · 02 Surface · 03 Proof

Say

  • “Everything you just saw is one method — diagnose the surface, rebuild the system, install how the team runs it.”
  • “Here’s why it matters for Vital Wires: your delivery is enterprise-grade; your market signal doesn’t show it. Buyers meet your website and materials before they meet your team — and right now those surfaces don’t show the ROI you already deliver.”

Slide 06 pushbacks below — diagnose, don’t insult. If he goes full referrals → Objections #6.

If he asks · surfaces / website / “how do you know?”

“How did you come up with that? Are you saying our website is bad? We have case studies.”

“I’m not saying the site is broken — I’m saying the buyer journey starts there, before they meet your people. I haven’t done a full audit of Vital Wires yet; this is the pattern I see in enterprise consulting: strong delivery, packaging that doesn’t carry the same weight. Case studies help only if a buyer can find them fast, they sound like how deals actually close, and sales actually uses them in every proposal. If yours already do that, this card gets lighter — the presence sweep would show it side by side.”

“Why quieter? We post / we’re active on LinkedIn.”

“Quieter relative to the delivery — not ‘you don’t post.’ A firm can post every week and still look quieter than its wins deserve if the posts don’t carry proof, outcomes, and a clear why-Vital-Wires. Premium work behind quieter packaging usually means: generic capability language, thin or buried proof, decks that don’t feel like the caliber of the implementation. That’s the hypothesis. Happy to be wrong after a short side-by-side look.”

Hover / “shortlisted late or not at all” — “Says who?”

“That’s the buyer behavior I’m naming — they rush the site and materials first. Late shortlist is what happens when packaging doesn’t match delivery: they don’t feel the ROI in thirty seconds, so you enter late or not at all. I’m not claiming a secret look at your pipeline — I’m naming the risk that pattern creates. The sweep is how we check whether Vital Wires is in that pattern or not.”

If he asks · “What ideas do you have?” (don’t back off — give direction + Map)

01 Pipeline capped by referrals — “So what would you do?”

“Not replace referrals — arm them and add a second engine. Concretely: a diligence kit so every warm intro lands harder (site, one-pager, proof that matches the deal); then presence-plus-pipeline for verticals you already own — when SAP ships something or a regulation drops, that’s a campaign, not a hope. Exact mix depends on how deals close today — that’s Map, then a scoped proposal in forty-eight hours. I won’t invent a full GTM plan on this slide.”

02 Surface undersells substance — “Okay, what would you change?”

“First move is the sweep — website, LinkedIn, decks, collateral — working vs underselling, side by side. Directionally: sharpen positioning so a buyer feels the ROI in thirty seconds; make proof impossible to miss; make sales materials sound like how deals actually close. If the homepage is already strong, we leave it and fix what’s around it. I won’t redesign Vital Wires in the air — I’ll write the scoped fixes after Map.”

03 Proof stuck in memory — “What ideas for that?”

“Build a reusable proof library sales can grab without hunting Slack. Directionally: outcome one-pagers by vertical — not just logos; a small set of case frames that paste into every proposal; a rule that every closed win produces one reusable asset within a week. That’s how wins stop living only in people’s heads. Which formats matter most for Vital Wires — I’ll pin that in Map from how your deals actually close, then scope it in the proposal.”

Deeper ammo · slide 06

  • Frame = hypothesis + pattern, not “I audited you last night.”
  • Agree with what’s true (“you post,” “you have case studies”) then reframe quality/findability/use.
  • On “what ideas?”: always 2–3 concrete directions + “scoped after Map / 48hr proposal” — never only “I need to research.”
  • Bridge out: “That’s the gap. Next slide is the road — Create → Launch → Control → Sell.”
07 The plan

Say

  • “So the road has four stages. Create the foundation. Launch it so it’s seen. Control it so everything stays on brand. Sell — turn presence into pipeline. First two are one-time projects; last two are ongoing. You choose where we start.”
08 Create

Say

  • “Positioning, identity, website, proof. Nothing launches until the foundation is right — that’s the Blenz discipline you just saw.”
09 Launch

Say

  • “Shipped quietly is the same as not shipped. Launch is the announcement, the films, LinkedIn and email — the new brand actually landing in the market.”
10 Control

Say

  • “You know how consulting decks get made — cut-paste from three old decks at eleven pm before the client meeting. Control is the system that catches that: team produces, we check, on-brand ships. Every proposal, every deck, every post.”

Keep the cut-paste line verbatim — advisor anecdote; lands hard with an ERP consulting CEO.

11 Sell

Say

  • “And this is where it pays for itself. When SAP releases something or a regulation drops in a vertical you own, that’s a campaign waiting to run. Presence becomes conversations, conversations become pipeline — measured, so we double down on what converts.”
12 Channels

Say

  • “Day to day it looks like this — everywhere the firm communicates, kept consistent and alive.”

THEN CLICK EXACTLY TWO: website mock + events (SAP RISE Roundtable). Skip the rest unless asked.

13 Lead engine

Say

  • “And pipeline isn’t theory for me — this is my own outbound system. Nine hundred emails a day capacity, every lead researched and personalized by AI, every send human-approved. I run this for my own studio; it’s the same class of system I’d stand up for you.”

If he asks

“What’s the conversion / reply rate?”

“I’ll give you the honest number — my first test batch of 350 got a 4% reply rate, and I’m scaling volume now. I’d rather show you a working system and real early numbers than promise you a rate I haven’t earned yet.”

14 Franchisee Shift

Say

  • “One more, quickly — sometimes the marketing problem is a product problem. I led the full product design of this franchise-operations platform: HQ to franchisee, announcements, brand standards, one system instead of WhatsApp chaos.”
15 Selected work

Say

  • “The broader track record — TD Bank’s mobile app among them, seven years of product design.”

Flip-through. Don’t dwell.

16 How we start

Say

  • “One path, no package menu. We map priorities in a conversation, I send a scoped proposal within forty-eight hours, you approve before anything builds.”
17 How we work

Say

  • “Start small, expand only if it’s working. One senior partner accountable. And everything transfers — files, brand system, any tools we build. No hostage retainers.”
18 Close

Say

  • The Ask — then SILENCE.
“So — of the two starting points, brand foundation or presence-plus-pipeline: which feels like the right first ninety days for Vital Wires?”

Ask → silence. Don’t fill the pause.

The ask

Say this cleanly. Then stop talking.

“So — of the two starting points, brand foundation or presence-plus-pipeline: which feels like the right first ninety days for Vital Wires?”

End state

He saw real work · Map → Propose → Build understood · starting stage picked (or Map booked) · proposal within 48 hours · no price invented in the room · silence after the ask.

Objections

The eight answers live in one place — open that tab when he pushes back.

Open VW Objections ↗

Q&A — strong answers

Internalize. Don’t recite. Tone: calm, precise, peer-to-peer. Pricing / ROI / referrals / team size → VW Objections tab (once).

01“Are you selling me software or marketing?”
Vivek“Are you selling software or marketing?”
You“Marketing first — brand and presence for Vital Wires. Software when marketing needs a product underneath. Creative Hub and Franchisee Shift exist because I’ve hit problems you can’t solve with a Canva file. For Vital Wires I’d still start Create / Launch. Tools only if they clearly accelerate the marketing outcome.”
02“How do you work with our internal team / sales?”
Vivek“How do you work with our team?”
You“I don’t replace your delivery or your sales leadership. I make them easier to buy. Practically: focused loops with whoever owns brand and content; short loops with sales so proof matches how deals actually close; one review path so assets don’t die in chat. You stay the face of Vital Wires — I make the market materials and systems match that.”
03“Show me something live — not slides.”
Vivek“Show me something live.”
You“Yes.” (Blenz guidelines / Creative Hub / Lead engine.) “This is production work, not a pitch mock. When you’re done looking, I’ll bring it back to where Vital Wires should start.”
04“How is this different from Metavrash / Shree?”
Vivek“Isn’t this Metavrash?”
You“Shree introduced us — I’m grateful for that. This engagement is independent LoopSuit AI work: brand, presence, and pipeline systems for Vital Wires. Different scope, different accountability. If anything overlaps with Metavrash delivery, we keep lanes clean so you’re not paying twice for the same outcome.”
05“We’re enterprise / SAP-world. Do you get that buyer?”
Vivek“Do you understand enterprise buyers?”
You“Yes — and that’s why ‘pretty brand’ isn’t enough. Enterprise buyers judge you in the first thirty seconds of the site and the first page of a capability deck. I’ve shipped product design inside TD Bank’s mobile world — regulated, multi-stakeholder, high bar. Vital Wires already has the substance. We make the surfaces match.”
06“What about BeatRoute?”
Vivek“Should marketing cover BeatRoute too?”
You“Only if you want it in scope — I won’t force a BeatRoute pitch. Vital Wires brand and presence is the clear start. If later you want a product-led growth lane for BeatRoute, we map that separately. One starting stage at a time.”
07“What does the first ninety days look like?”
Vivek“What’s the first ninety days?”
You“Depends on what we pick today. Brand foundation: positioning and identity locked, website and proof that sell the delivery, sales materials sales will actually use. Presence-plus-pipeline: channels running on-brand, outreach your team can operate, measurement on what’s converting. Either way: Map → proposal within forty-eight hours → you approve → Build. Expand only when the first stage is working.”
08“Send me something — I’ll think about it.”
Vivek“Send me something. I’ll think about it.”
You“Absolutely — and so the proposal is useful, not generic: should we scope brand foundation or presence-plus-pipeline as the start? I’ll write to that and send it within forty-eight hours.” If he won’t pick: “Then let’s book fifteen minutes to Map priorities — so I’m not guessing.”

Do not say

  • Any price number
  • “MVP”
  • “Biryani Blues”
  • “vCMO / fractional CMO” — if HE names it, accept the frame
  • Rank Radar / Cortex — removed from deck; only if he asks what else you’ve built
  • Long METAVRASH backstory
  • “I’m just one person”
  • “Vinay” — his name is VIVEK

Vital Wires phone pack · LoopSuit AI · open anytime: #tab-vitalwires · cheat: #tab-vwcheat · objections: #tab-vwobj

Memorize this · nothing else · Vivek / Vital Wires

VW cheat sheet

The spine, five verbatim lines, and the full objection bank. Phone-ready.

8 spine beats 5 verbatim lines 8 objections

Full script with slide lines lives in VW Meeting.

Open VW Meeting ↗

The spine · 7 beats

Memorize this sequence. Story → gap → plan → proof → close.

  1. Blenz before — clinical / unscalable / no guidelines — full rebrand ask
  2. Blenz rebuild — twelve weeks, end-to-end system, quality bar
  3. Brand book — living guidelines, one-time establish / ongoing enforce
  4. Creative Hub — I designed it AND shipped it; leadership uses it daily
  5. “One method — here’s why for Vital Wires” — enterprise delivery, weak market signal
  6. 4 stages — Create / Launch = one-time · Control / Sell = ongoing
  7. Lead engine — 600/day capacity, AI-researched, human-approved
  8. Map → Propose → 48hr — then the ask, then silence

Five verbatim lines

Say the line. If he tilts his head or follows up — use the plain English under it.

Price

“Once we map the exact starting scope — custom proposal, forty-eight hours.”

Means: We pick the starting stage in this conversation (or a short Map call). I write a proposal with deliverables, timeline, and one fixed price — within two days. I won’t invent a number live.
If he pushes for a number now: “I can give you a guess, but it’ll be wrong. Same reason Vital Wires doesn’t quote an SAP implementation before discovery. Agree the stage with me — the proposal makes it concrete.”

ROI

“I won’t invent a two-million-dollar chart for you. In consulting you can’t usually prove one campaign caused one deal — you know that. What I commit to is a scoreboard we both watch.”

“Scoreboard / instrumentation” means in plain words: we track real activity numbers — people who reached out, meetings booked, proposals sent, pipeline that came from the marketing work — and we review them monthly so we double down on what works and kill what doesn’t. Not “trust me, branding is good.”
If he asks “so what are we actually measuring?”: “Before we start Sell, we lock three or four indicators with you — for example inbound inquiries, meetings from campaign or LinkedIn, proposal rate, pipeline tagged to presence work. Those go on a simple dashboard. You’re judging me on those numbers, not on how pretty the deck looks.”

Control

“You know how consulting decks get made — cut-paste from three old decks at eleven pm before the client meeting. Control is the system that catches that.”

Means: Your team still produces proposals, decks, posts. Before anything goes to a client or LinkedIn, it passes a quick on-brand check — so late-night Frankenstein decks don’t ship.
If he asks “so you police every slide?”: “We set a review path that fits how your team already works — not bureaucracy. Produce → check → approve → ship. Fast enough for real deadlines, strict enough that Vital Wires always looks like Vital Wires.”

Difference

“One senior brain, no handoffs — and the tools become yours.”

Means: You talk to me, I do the work (with a lean bench for production). No account manager translating you to a junior. Software and brand files we build — Vital Wires owns them.
If he asks “tools become ours how?”: “Files, brand system, and any app we stand up transfer to you. You’re not renting your operating system from me. If we part ways, you keep running.”

The ask → silence

“Brand foundation, or presence-plus-pipeline — which feels like the right first ninety days?”

“Brand foundation” = Create (positioning, identity, site, proof) — fix the surfaces first.
“Presence-plus-pipeline” = Launch + Sell lean — get visible and start generating conversations if the foundation is already strong enough.
Then: stop talking. Count to five in your head. If he stalls: “If it’s easier, we can book fifteen minutes just to Map — then I’ll write the proposal to that.”

Plain English · phrases that sound smart but confuse

If you wouldn’t say it to your dad, don’t say it to Vivek. Use the right column.

“Instrumentation”

Don’t say that word. Say: “a scoreboard we both look at” or “we track the numbers every month.”

“Attribution is murky”

Say: “In consulting you usually can’t prove one LinkedIn post closed one SAP deal — too many people touch the deal. So I won’t fake a ROI chart. I’ll track the leading numbers we can see.”

“Presence”

Say: “How Vital Wires shows up before anyone meets your team — website, LinkedIn, decks, emails, events.”

“Market signal”

Say: “What buyers think you are from your surfaces — right now that understates how good the delivery is.”

“One-time vs ongoing”

Create + Launch = projects with an end date. Control + Sell = monthly work that keeps the brand clean and pipeline moving. He chooses where to start.

“Map → Propose → 48hr”

Map = short conversation on priorities. Propose = I send scoped deliverables + timeline + fixed price. 48hr = you have it within two days. Nothing builds until he approves.

Objections

Full bank is not duplicated here — go read it once.

Open VW Objections ↗

Steal from NEPQ · Nick · Alex

Moves that sharpen the room — not full scripts. Full Alt answers live in VW Objections.

NEPQ · always open objections with

“Yeah, that’s not a problem.” — then handle. Never defend first.

NEPQ · smokescreen piercer

Book next step → “Before I go — what were you wanting to go over in your mind?” → real objection.

NEPQ · self-persuasion

After he agrees something would work: “Why do you feel like it would?” — he sells himself.

Alex · damaging admission first

Agree with the true part of the objection before the reframe. “You’re right — most ROI charts in B2B are theater.”

Nick · system is its own proof

On lead engine / Creative Hub: “This is production work, not a pitch mock — same class of system I’d stand up for you.”

Nick · blame scope, not evasion

Price early: don’t invent a number — Map the stage, proposal in 48 hours. Scope is why the number isn’t live.

Alex · sell the vacation, not the flight

Don’t narrate your process. Sell the end state: buyers feel Vital Wires’ caliber before the call; pipeline you can scoreboard.

Brain · don’t take the bio — take the surface

Trust objection → point at Blenz / Hub / TD work on screen. If it doesn’t match the call, he shouldn’t sign.

If he interrupts / hijacks

Interruption = engagement. Stop the deck. Answer. Re-enter at the nearest chapter.

He jumps to price early

“Fair — I’ll get you a number. Fastest path: pick brand foundation or presence-plus-pipeline as the first ninety days, and I’ll have a fixed proposal within forty-eight hours. Want to finish the two-minute gap so the proposal isn’t generic?”

He goes deep on software / AI

Show Creative Hub for 60–90 seconds. Then: “For Vital Wires I’d still start on brand and presence — the tools support that, they don’t replace it.”

He says “we’re fine / not a priority”

“Got it — then I won’t force it. One useful next step even if we never work together: I can do a short presence sweep — what’s working vs what’s underselling — and send you the notes. No build. Worth thirty minutes?”

He compares you to a big agency

“Big agencies sell coverage. You need coherence — one senior across brand, marketing, and the system underneath, and assets that transfer. For Vital Wires’ stage, coherence beats headcount.”

Lost the thread after a tangent

“Good — let me put that on the board. Nearest chapter for Vital Wires is still: opposite problem of Blenz, then the four stages, then where you’d want the first ninety days.”

Vital Wires cheat sheet · LoopSuit AI · #tab-vwcheat

Read once · Vivek / Vital Wires · primary + NEPQ / Nick / Alex alts

VW objections

Eight answers, once. Primary = Shree-coached / your room script. Alt = second variation from Jeremy Miner NEPQ, Nick Saraev, or Alex Hormozi — try which feels like your mouth.

8 objections Alt NEPQ · Nick · Alex +1 CMO bonus

Jeremy’s NEPQ is already in Playbook → Sales. Full ladder + word-tracks live there.

Open NEPQ in Playbook ↗

How to use Alt

Default to Primary in the room. Use Alt when Primary feels stiff, or when he pushes a second time. NEPQ = agree + questions. Nick = scope / proof / system. Alex = damaging admission first, then reframe.

Objection bank

Short → expansion → Alt. No price numbers. No “MVP.”

1. What does this cost?

Short: “Once we map the exact starting scope, I’ll put together a custom proposal and send it within forty-eight hours. Fixed scope, fixed price.”

Expansion: “Same way Vital Wires would never quote an implementation before discovery — I don’t price before scoping. What I can do right now is agree the starting stage with you, and the proposal makes it concrete: deliverables, timeline, investment.”

Shree-endorsed dodge — keep verbatim as Primary.

Alt · NEPQ (Miner) + Nick

“Yeah, that’s not a problem. I won’t invent a number in the room — same reason you don’t quote SAP before discovery.”

“If the funds weren’t the issue — which feels more useful as a first ninety days: brand foundation, or presence-plus-pipeline?”

Then: “I’ll write that as a fixed-scope proposal within forty-eight hours. You approve before anything builds.”

Mechanism: NEPQ agrees first (deflates tension), isolates value from money, then Nick’s “blame scope / pull to Map.” Still no dollar figure.

2. I already have a good website — why do I need you?

Short: “You do — and a website redesign isn’t what I’m proposing. I’m proposing your brand presence: every surface a buyer crosses before they ever meet your team.”

Expansion: “Here’s what I’d actually do first: a full sweep of your current presence — website, LinkedIn, decks, collateral — what’s working, what’s underselling the delivery. Positive and negative, shown side by side with what it should communicate. If the website turns out to be fine, it stays; the gap is usually everywhere around it — the cadence, the proof, the consistency.”

Shree coaching: analysis, positive and negative, show the impact.

Alt · NEPQ dissatisfaction-finder + Alex damaging admission

“You’re right to protect a site that works — most of what I’m talking about isn’t the homepage.”

“Quick question — is there anything you’d change about how a serious buyer experiences Vital Wires before the first call, if you could? Website, LinkedIn, capability deck, the PDF sales sends?”

Let him talk. Then: “That’s the sweep I’d do — show what’s working and what’s quietly underselling the delivery. If the site’s fine, it stays.”

3. How are you different from any other agency?

Short: “One senior brain across brand, marketing, and the software underneath — nothing gets lost in handoffs, because there are none.”

Expansion: “An agency gives you an account manager between you and the work. I design the brand, I ship the marketing, and I build the tools your team runs on — you saw Creative Hub. Seven years of product design including TD Bank’s mobile app means the digital work is production-grade, not agency-decorative. And whatever we build transfers to you — I’m the only option in the room where the tools become yours.”

“Tools become yours” means: Creative Hub–class software, brand files, guidelines — Vital Wires owns them. You’re not locked into my platform forever.
Alt · NEPQ master disarm + Nick “don’t take the bio”

“Honestly — I’m not sure we’d even be useful yet. It depends whether you need another agency layer, or one senior who designs, ships, and builds the operating system underneath.”

“Don’t take the bio. Take the surface — Creative Hub is live with a Canadian agency’s leadership every day. That’s the difference: proof you can poke, and assets that transfer.”

4. What’s the ROI? Show me results.

Short: “I’ll be straight — I won’t invent a revenue chart that says I made someone two million dollars. In B2B consulting, you usually can’t prove one campaign caused one deal — and you already know that.”

Expansion: “What I will commit to is tracking the numbers that actually move: inbound conversations, meetings booked, proposals out, pipeline sourced from this work. We look at those together every month, keep what works, drop what doesn’t. Sell exists so you can judge me on pipeline — not vibes. Create and Launch have hard outcomes too: brand shipped, site live, proof assets sales can send into every deal.”

If he follows up (“what do you mean / how do we measure?”): “Same way you’d track an SAP rollout — we pick a few leading indicators before we start, put them on a simple scoreboard, and review them on a cadence. I’m not asking you to trust a story. I’m asking you to watch the numbers with me.”
If he says “so you’re not guaranteeing ROI?”: “Correct — I won’t guarantee a dollar number in a first meeting. Anyone who does is selling theater. I guarantee the system: work that ships, presence you can see in market, and a scoreboard we both look at. If the numbers don’t move after a fair run, we don’t expand — we stop or change course.”
Alt · Alex damaging admission + NEPQ self-persuasion + Nick “system is its own proof”

“Here’s what sucks about most marketing pitches — they invent attribution. I won’t.”

“If we put a scoreboard on presence work — conversations, meetings, proposals, pipeline tagged — do you feel that’s a fair way to judge whether this is working?”

If yes: “Why do you feel like that would work for Vital Wires?” (He argues for the measurement.)

“And on pipeline ops — the lead engine you’re looking at is the same class of system I run for my own studio. Proof isn’t a case-study PDF. It’s the system.”

5. Who actually does the work? How big is your team?

Short: “It’s senior-led — I personally lead design and build on every engagement. That’s the point of hiring me.”

Expansion: “Behind me is a lean bench of specialists I bring in per engagement — content production, motion, development — depending on what the work needs. What you never get is the agency pattern: pitched by a senior, delivered by a junior. The person in this room is the person on the work.”

Alt · Alex damaging admission + brain objection-map (trust)

“Fair question — and you’re right to ask. Most shops pitch with a senior and deliver with a junior. That’s the pattern I’m not selling.”

“Don’t take my word. Take the surface — Blenz, Creative Hub, TD-grade product work. If the work on screen doesn’t match the conversation, don’t sign.”

6. Our business comes from referrals — why do we need marketing?

Short: “Referrals are proof your delivery is excellent — that’s the asset. The problem is you can’t turn a dial: every new deal still needs a warm intro, so growth stays capped by who happens to know you.”

Expansion: “And even referred buyers check you out first — they open the website and the capability deck before the call. Those surfaces don’t close deals, but they can quietly kill them, and you never hear which ones died. Marketing isn’t replacing referrals — it’s making every referral land harder, and adding a second path so you’re not only waiting for intros.”

“Nothing compounds” / “starts from zero” means: last month’s intro doesn’t automatically create next month’s pipeline. Referrals don’t stack into a machine you can turn up.
Alt · NEPQ consequence + Alex “objection is the argument for”

“Yeah, that’s not a problem — referrals are the best signal you’ve got. Delivery is clearly strong.”

“Help me understand — when you want a new vertical or a new geography, how do you turn that dial up without waiting for another intro?”

“And even warm intros due-diligence you. So the surfaces that look weak don’t replace referrals — they tax them. Fixing that is how referrals land harder.”

7. After launch, do I keep paying? What are the ongoing costs?

Short: “The tools we build are meant for your team to run without me on every click — that’s how I design them.”

Expansion: “Ongoing costs are just the tool subscriptions, listed in the scope — no surprises. If you want us on maintenance, or on Control and Sell month to month, that’s your call and priced separately. You’re never locked in — files, brand system, and any software we stand up transfer to Vital Wires.”

“Everything transfers” means: you keep the assets even if we stop working together. No hostage retainer — you’re not renting your own brand system from me.
Alt · Nick deliverable pricing / offer stack

“Think of it like a scoped build first — Create and Launch are projects with an end date and a fixed price. That’s the default.”

“Control and Sell are optional rungs after — only if the foundation is earning its keep. Separate scope, separate decision. No open-ended fog, no hostage retainer.”

8. Our team can handle this internally.

Short: “They can keep it running — that’s exactly what the Control stage sets them up to do. What they can’t do alongside delivery is the rebuild itself.”

Expansion: “A brand foundation is a project with an end date, not a task on someone’s side desk. I do the concentrated build; your team inherits a system — guidelines, templates, the review flow — that makes them faster afterward, not dependent on me.”

Alt · NEPQ + brain “hire vs ship this quarter”

“I’m not sure we’d even be needed for the ongoing — that’s the point of Control. Your team should own the run.”

“The rebuild is different. Hire or assign internally when you need a permanent seat. Right now you need a concentrated project shipped this quarter — without pulling delivery people off client work. I do the build; they inherit the system.”

Bonus · if HE names the CMO shape

“Functionally yes — the accountability of a marketing head without the org overhead. Some engagements run two or three days a week, some start with Create as a defined project — the proposal lays out the options and you pick the cadence.”

Only if he names it. Never lead with vCMO / fractional CMO.

Alt · NEPQ “feel” close

“If that’s the shape you want — does it feel more useful as a defined Create project first, or a couple of days a week on presence and pipeline? I’ll write the proposal to whichever you pick.”

Smokescreen · “send me something / I’ll think about it”

Not one of the eight — but it will happen. NEPQ loop.

Primary: “Absolutely — and so it’s useful, not generic: brand foundation or presence-plus-pipeline as the start? I’ll write to that within forty-eight hours.”

Alt · NEPQ objection loop (Miner’s best move)

“Yeah, that’s not a problem. Let’s put fifteen minutes on the calendar so you’re not chasing me and I’m not chasing you.”

Book it. Then: “Before I go — what specifically were you wanting to think through? Just so I know what questions you’ll have.”

Mechanism: Pressure drops → real objection surfaces (price, trust, partner, timing) → handle THAT one now, not in email hell.

If the room goes technical

Show Creative Hub or Franchisee Shift for 60–90 seconds. Then: “For Vital Wires, I’d still start on brand and presence — the systems support that, they don’t replace it.”

Vital Wires objections · once only · #tab-vwobj · sources: Shree room notes · Miner NEPQ (Playbook) · LoopsuitBrain Nick/Alex wiki

Infrastructure · revised 2026-09-23 · the thing that is actually blocking you

The Inbox Engine.

The block is not knowledge. It is an unfinished purchase. Every hour the warmup clock is not running is an hour added to the end of the timeline. This tab removes every remaining decision so the only thing left is clicking. Markdown source: outreach/inbox-engine.md.

5 domains × 2 = 10 mailboxes
20/day per mailbox · ceiling 25
bounce < 2%, not < 5%
~$100 month one

01The one-page answer

QuestionAnswer
Which mailbox provider?Zapmail. ~$3.90/mailbox on Starter, Google or Microsoft, automated DNS. Saraev names it by name, Sep 2026: "My current choice is Zapmail."
Why not Zoho again?Zoho's AUP forbids automated/bulk sending — it killed 29 mailboxes in August. That was a policy failure, not a technical one. Check the AUP explicitly permits cold outbound before buying anything, ever again.
Google or Microsoft?Google. Same price on Zapmail, and funded startups are overwhelmingly on Workspace — provider matching then gives you Gmail→Gmail, the best-placing combination available.
How many domains, how many mailboxes?5 domains × 2 mailboxes = 10. Fills the $39 Starter plan exactly. Not 10 × 3 — see §02 for why, and for the trigger that unlocks the rest.
How many per mailbox per day?20. Hard ceiling 25. Down from 30–40 in 2025. The single most important number on this page.
Warm up, or buy pre-warmed?Both — they are different jobs. Self-warm the 10 brand mailboxes. Buy 3–5 pre-warmed today so you send this week instead of in three weeks.
Where do leads come from?Buying signals, not a database. Companies hiring a designer, or freshly funded. 15–25% reply vs 3.43% baseline. See §05.
Which verifier?MillionVerifier, valid only. Start on the 500 free credits. Bouncer is an optional catch-all recovery pass — skip it until you are throwing away 30%+ of a good list.

02The domain plan — 5 × 2, and why not 10 × 3

You own ten domains, all registered ~May 2026 (so ~4.5 months aged, which registrars and ESPs treat better than brand-new) and all valid to 2027-05-05. Use five now, hold five back.

LayoutMailboxesSends/dayZapmailBlast radius
3 domains × 39180$393 domains
5 domains × 2 ← this one10200$395 domains
10 domains × 330600$99 + $50 Instantlyall 10 domains

5 × 2 is strictly better than 3 × 3 — same price, one more mailbox, risk spread across five domains instead of three. If one gets flagged you lose two mailboxes, not three, and five untouched domains remain in reserve.

Use nowThe five sending domains

loopsuitstudio.com · loopsuitlabs.com · loopsuitbuilds.com · buildwithloopsuit.com · shipwithloopsuit.com

Chosen because they read like a real company when a founder squints at the sender address — which is question one of the four.

Hold backThe reserve five

launchwithloopsuit.com · growwithloopsuit.com · createwithloopsuit.com · labsloopsuit.com · aiwithloopsuit.com

Deliberately last: aiwithloopsuit — an "ai" prefix is an increasingly strong spam signal in 2026, so do not lead with it. labsloopsuit reads scrambled.

Neverloopsuitai.com

Your primary domain never carries outbound reputation risk. It is where buyers land after they Google you — if it gets burned, the credibility half of the funnel dies with it. Buying ten lookalikes was exactly right.

Why not all ten now — the binding constraint is Instantly, not Zapmail

1. You already lost 29 mailboxes once. Burning all ten domains on an unproven setup leaves zero brand domains to fall back on. You want survivors if something is misconfigured.

2. You cannot feed 600/day yet. Capacity is worthless without leads and proven copy, and unproven copy at 600/day burns ten domains' reputation on v1 of a message.

3. The real cost is the plan upgrade. Instantly Growth caps at 5,000 emails/month. Ten mailboxes × 20/day × 22 days ≈ 4,400 — you are already at that ceiling. Going to 30 mailboxes forces Zapmail Growth (+$60) and Instantly Hypergrowth (+$50). That is +$110/month for capacity you cannot use.

The upgrade trigger, written down so it is a decision and not a drift: after 30 days of sustained bounce under 2% and at least 1,000 delivered, add the next five domains and upgrade both plans. Not before.

03Buy list — today, ~45 minutes of clicking

#WhatCostWhy it is on the list
1Zapmail Starter — 10 mailboxes$39/moThe fleet. 5 domains × 2. Automated SPF/DKIM/DMARC/MX. Do this first — it is the only clock that cannot be compressed.
23–5 pre-warmed inboxes~$15–25/moThe bridge. Send this week rather than in three. Also the ammunition you are willing to burn.
3Instantly — sending onlyhave itGrowth is fine at 200/day. Do not buy the Leads/Credits plan — see §05.
4Sales Navigator$0 — free trialActivate today and diary the renewal date, so month two is a decision rather than a default charge.
5Apify — pay-per-result actors~$30Sales Nav export at ~$0.01/lead, plus the job-board actors. No subscription.
6Anymail Finder$14/moPay only for verified. Risky and not-found are free. This is what structurally fixes the bounce problem at source.
7MillionVerifier$0 to startFinal gate. 500 free credits with a business email; credits never expire. Buy a pack once you know your pass rate.

≈ $100–110 for month one. Cut from the earlier version of this page: Evaboot ($79–139/mo), Bouncer, and the Instantly Leads upgrade ($47/mo) — all three were over-build. For reference, Saraev's genuine minimum-viable 2026 stack is about $50/month.

04Zapmail → Instantly, exactly

Eight steps. Do not reorder them — steps 4 and 6 are where people destroy fleets.

  1. Connect the domains in Zapmail, not in HostingerZapmail → Connect Existing Domain → pick the five from §02. Zapmail returns nameservers; in Hostinger, replace the existing nameservers with theirs. Propagation 24–48 h.
  2. Then never touch DNS againZapmail now owns SPF, DKIM, DMARC and MX and manages them automatically. Hand-editing those records is the most common way people break this.
  3. Create 2 mailboxes per domain, real-person namesjastej@ and j.sehra@ on each. Never hello@ / info@ / team@ / contact@ — role addresses are filtered harder and they defeat the entire "a human wrote this" premise.
  4. Verify green twiceWait for green in Zapmail, then check independently at mxtoolbox.com. Zapmail's dashboard sometimes reports verified before propagation finishes, and connecting early is what poisons a fleet.
  5. Warm 14–21 days until the score clears 80One click to start; it then runs itself. Do not connect to Instantly before 80. This is the step that keeps getting postponed and the only one that cannot be compressed.
  6. Export into InstantlyOne-click export authorises into the workspace. If it fails: Instantly → Email Accounts → Add → Custom SMTP/IMAP, SMTP 587 STARTTLS (or 465 SSL), IMAP 993 SSL. A port mismatch fails silently — test both before suspecting the password.
  7. Cap the volume yourself — Instantly will notInstantly does not enforce warmup limits on custom SMTP. Set 20/day per inbox, ceiling 25. Keep Instantly's own warmup running alongside live sending, permanently. 10 × 20 = 200/day.
  8. Campaign hygieneInbox rotation ON · provider matching ON · two-step sequence (Day 0 / +3) · open tracking OFF · plain text · no link in Day-0 · footer OFF · and never load an address that has not come back valid.

05Leads — buy signals, not a database

The finding that should decide your whole lead strategy

Signal-triggered outreach gets 15–25% reply rates against a 3.43% baseline — roughly 5×. Signals meaning funding rounds, leadership changes, hiring surges. Everything else is noise around that fact.

And for what you sell, one signal beats all the others: companies actively hiring a Product Designer, Design Engineer or Senior Frontend.

Database leadCompany hiring a designer
Budgetunknownconfirmed — a salary is allocated
Painassumedstated publicly, in their own words
Urgencynonean open role costs them every week
Who to contacta job titlethe hiring manager, named
Your offergenerica direct substitute for what they are already buying

That last row is the whole game. You are not pitching "AI products" — you are saying hiring takes three months and $140k/year; I will ship the thing in three weeks for a fixed price, and you can still hire. That is a real argument to a real person with money already set aside.

Why the July disaster does not apply here

You used free contact scrapers that pattern-guess first.last@realdomain — of 704 UnitOS domains, every one had valid MX records and every mailbox was invented. Job scrapers return public job postings: data that actually exists. Finding the email is a separate step, handled by a tool that only charges when the address is verified.

Run two lanes, let reply data decide

Lane ASales Navigator Spotlights

The Spotlight filters are the point, not the basic ICP filter that every database also has: Job Opportunities · Recent funding events (last 12 mo) · Headcount growth (20%+ / 6 mo) · Leadership changes (last 3 mo).

Stack them: hiring a designer + funded in the last 12 months + under 50 people. Export with an Apify pay-per-result Sales Nav actor at ~$0.01/lead — 2,000 leads ≈ $20, versus $79–139/mo for a subscription scraper.

Free for month one on your trial.

Lane BJob boards, direct

Apify actors on Wellfound/AngelList · YC Work at a Startup · LinkedIn Jobs · company career pages (Greenhouse, Lever, Ashby, Workable). Pay-per-result, ~$3/1K, no subscription and no LinkedIn account exposure at all.

ThenThe comparison

~1,000 leads each, identical copy, compare reply rate at 500+ sends per lane. The winner takes the volume in month two — and that result is also what tells you whether Sales Nav is worth $99 when the trial ends.

This is the split-test rule applied at the source, not just the copy.

⚠️ Protect the LinkedIn account

Cookie-based scrapers run through your own session, and LinkedIn does restrict accounts for it. Your LinkedIn is load-bearing — it is where buyers land after a Day-0 email, and it is part of the professional record behind the FSH story. Do not risk it to save $20.

Prefer a no-login Apify actor (several exist, ~$0.30/run + $0.01/profile). If you use a cookie-based one: a few hundred rows a day, waking hours only, never overnight — and never from the account you send connection requests from.

Why not Instantly's own lead database

$47 buys 1,500 credits ($0.031 each) and credits burn per action — AI lead generation runs ~5 credits per lead, about $0.16/lead. More to the point it is a generic 450M-row database: stale, and every other sender is emailing the same people with the same "personalised" opener. Use whatever credits are already in your subscription for enrichment. Do not buy the credit plan.

06The chain, and the three kill-switches

the pipeline, end to endSIGNAL SOURCE Lane A: Sales Nav Spotlights (hiring + funded + <50 people) → Apify export, $0.01/lead Lane B: Apify job boards — Wellfound / YC / Greenhouse / Lever / Ashby ↓ QA GATE — 100 rows into Claude: "how many of these are companies where a solo design engineer could realistically ship what they're hiring for?" ≥50 → proceed. <50 → fix the filter, re-pull. (60 seconds) ↓ Anymail Finder → founder / hiring manager. Pay ONLY for verified; risky and not-found cost nothing. ↓ lead_triage.py → SEND / RISKY / DEAD (guessed_pattern >10% = source is guessing) ↓ MillionVerifier → send `valid` ONLY (500 free credits to start) ↓ personalization engine → human template + fuzzy variables ↓ Instantly → loaded, rotated, capped at 20/inbox/day

KillPull verifies under 70% valid

Throw the whole pull away and change the source. Do not salvage it — the 30% that survived came from the same guessing process as the 70% that did not.

Killguessed_pattern over 10%

The source is pattern-matching, not observing. UnitOS was 13%, QSR 10% — and they bounced at 68% and 60%. Change actor or provider.

KillLive bounce over 2%

Pause immediately. Above 2% sustained, Google penalises the entire domain, not just the campaign. Instantly hard-kills at 5%.

07Pre-warmed — the bridge, and how to allocate it

Pre-warmed mailboxes are rented, on the provider's generic domains (internetgrowth.ai and the like), warmed 20–30 days in advance. You do not choose or own the domain. The entire value is that you send on day one.

Job 1The bridge

Warmup means your ten Zapmail boxes send nothing until roughly 7–10 October. Three pre-warmed boxes × 20 = 60/day starting this week — about 1,000 emails of real data you would otherwise simply lose. You need 1,000 before any conclusion about copy is legitimate.

Job 2The ammunition

These are the inboxes you are allowed to burn while establishing a feedback loop. Rented, generic, replaceable. Your ten brand mailboxes never get burned.

TradeoffThey weaken question two

A founder who gets mail from internetgrowth.ai and Googles it finds nothing — which undercuts the "who the hell is this?" answer your whole message architecture depends on.

So allocate deliberately: pre-warmed for copy testing and volume on a segment you do not mind spending; your five domains for the leads you actually want to win, where jastej@loopsuitstudio.com does credibility work. Cancel the pre-warmed after month one if the brand fleet is healthy.

08Health monitoring

WeeklyPull the numbers

python loopsuit-automations/outreach-tools/scripts/instantly_pull.py → bounce %, reply %, per-inbox volume.

DoThe honest podcast test

Keep a small, real interview-invite campaign running from the same inboxes. Almost nobody declines a guest slot, so its reply rate is close to a pure read on inbox placement. When it drops, placement dropped — not the copy.

Saraev runs a version implying a channel he is not really booking for and calls it "a bit mean". Run the true one: you are publishing anyway, so every acceptance is also a warm ICP conversation and a piece of content.

QuarterlyRe-tender the stack

Mailbox providers go stale fast. The one that was best in July is often not best in October.

Reading a decaying campaign

7–10% reply that decays to 0.5–1% within two weeks is reputation erosion, not copy. Check in this order: (1) is any mailbox above 20/day? (2) bounce above 2%? (3) spam-trigger vocabulary in the body? (4) manual spam complaints — one per 200 sends does real, lasting damage. (5) is anything sending from a free/personal mailbox?

09Today, in order

Two clocks should be running by tonight: the Sales Nav trial and the Zapmail warmup. Everything else can slip a day.

  1. Activate the Sales Navigator trial · 2 minFree, 30 days, starts the moment you click. Put the renewal date in your calendar now so month two is a decision, not a default charge.
  2. Zapmail → buy Starter · connect the five domains · Start Warmup$39. Paste the nameservers into Hostinger. The warmup clock starts the moment DNS resolves — the only irreversible-if-delayed step on this list.
  3. Create 10 mailboxesjastej@ + j.sehra@ on each of the five. Google.
  4. Buy 3 pre-warmed inboxesThese are what you send from this week. Connect these to Instantly only — leave the Zapmail ten alone until the score clears 80.
  5. Build one Spotlight searchHiring a designer + funded in the last 12 months + under 50 people, US/CA. Export 200 rows with a no-login Apify actor.
  6. Run the 100-row QA gate through Claude"How many of these are companies where a solo design engineer could realistically ship what they're hiring for?" 50+ or fix the filter — before spending another cent on enrichment.
  7. Anymail Finder → MillionVerifierFounder or hiring manager per row. Load valid only.
  8. Load the campaign on the 3 pre-warmed boxes60/day, two-step, open tracking off, no link in Day-0.
  9. Diarise 14 days outCheck the Zapmail warmup score ≥ 80, then export the ten into Instantly and step up to 200/day.
the Day-0 this list writes for you · copy as MDSubject: your product designer role Saw you're hiring a product designer — went through the posting and then your onboarding flow on my phone, and hit a dead end on step 2. I'm a design engineer; three years shipping TD Bank's mobile app, now I design and ship AI products end to end, solo. That role is realistically three months to fill and onboard. I'd ship the same work in three weeks, fixed price, and you can keep hiring. I already redrew those three screens — happy to send the mocks either way. Worth 15 minutes?

Answers all four questions, is entirely true, and arrives with work already attached.

When the last box is ticked

LoopSuit has a live outbound funnel for the first time, mail is going out this week rather than in three, and the expensive clock is running in the background instead of waiting on a decision.

One honest caveat: 15–25% is the reported ceiling for signal-based outreach and assumes tight targeting and a relevant offer. Plan on the 0.25% booking rate in the Playbook and treat anything above it as upside. The reason to pick this source is not the headline number — it is that everything downstream gets easier when the prospect has already told you what they need and budgeted for it.

Herk's pricing system + Hormozi's close · September 2026

Price on value. Never on hours.

The whole method in one line: make them say out loud what the problem costs them, then charge a fraction of the number they just said. Markdown source: 06-pricing-and-packaging.md.

10–20% of annual value
show 10× first-year ROI
3 tiers — middle wins
50% up front, always

01The calculator

→ Open the client-facing quote sheet — that is the one you screen-share on a call. This page is your prep tool.

Put in what the problem costs them. Get out the number you quote and the line you say when they ask where it came from — and they will ask. Prep with this; never share this page on a call. Build a one-page client-facing quote sheet for the screen-share.

Use Hours for ops/automation work. Use Hire for the founder lane — the value is the designer they’re trying to recruit, and the three months it takes.

Hours lost per week

20 hrs/week

Fully-loaded cost per hour

$40/hour

Other annual leakage (errors, missed deals)

$0/year

Your capture rate

12% of annual value

say this when they ask where the number came from

Sanity checks. If the quote is under $3,000, either the problem is too small to be worth your delivery hours or you have not counted the leakage. If your fee is more than ~20–25% of what you are promising to generate, the offer is too expensive — fix the fee or fix the promise. And if you cannot produce a number at all, that is not a pricing problem: you do not yet understand the value, and you are not ready to write the proposal.

02The rule

Cost does not justify price. Value justifies price.

The gardener mows your lawn for $100 a week. He buys a new truck, his costs rise, and he asks $200 for the same lawn. Nobody accepts that — his costs are not your problem. Keep three words separate: cost is the floor you walk away below · value is the ceiling, what it is worth to them · price is anything in between you can agree on. In AI and product work that gap is enormous, which is exactly why hourly billing is so destructive here.

StopHourly punishes you for improving

Two developers, same $150/hr, same output. The fast one bills 1 hour; the slow one bills 3. You just made 3× from the worse worker. And when Claude Code makes you do in half a day what took a week, hourly means getting better at your job cuts your income.

Hourly is fine — good, even — for the first two or three projects, when you need trust more than margin. ~$100/hr. Then stop.

DoCount more than wages

Missed opportunity — a launch they have not got to in two quarters. Error cost — Herk's most profitable build converted a crew's phoned-in orders to text. It saved 45 min/day but prevented ~$12k/month of scheduling errors; the time saved was almost irrelevant. Revenue per unit — if the meetings you book close at $5k, price against bookings, not the coordinator's hours.

DoSell revenue, not savings

Money made = ROI × social proof ÷ resistance. "Eliminate admin → frees time → you make money" is a three-link chain and every link leaks belief. Skip to the end. Marketing and sales are standardised across industries so proof transfers; a bespoke back-office automation proves nothing to the next prospect.

03Find the constraint before you price anything

Revenue is water in a pipe. A blockage in the middle — plenty going in, little coming out — is supply. No water in the pipe at all is demand.

SupplyThe question

"If you 10×'d your business tomorrow, what would break first?"

Then say nothing. "First" is the load-bearing word — whatever breaks first is the constraint.

DemandThe question

"What would you have to do to 10× your business tomorrow?"

Then say nothing. Build the systems that support whatever they name.

WhyThis is the retainer

Solve the first constraint and a new one appears. That is not a problem — that is the recurring contract, and it is an honest justification for one, unlike a subscription for vague availability.

What they ask for is almost never what they need

Herk has taken 300+ intro calls; the stated request is almost never the real one. Over half of all AI automation requests are lead-gen/sales/marketing because "we need more leads" is everyone's first instinct — and it is frequently wrong.

The med-spa case, which matters because med spa was a LoopSuit lane: the owner asked for a lead-generation agent. Leads were arriving fine — they were not showing up to appointments, and nobody followed up with the ones who did. The pipe was full and leaking. The right build was reactivation and appointment reminders, not lead gen.

Two corollaries. Being an AI consultant means knowing when not to use AI — sometimes the honest answer is a deterministic script and a Slack notification, or "hire another person," and saying so is why they keep you. And clients often want relief, not value — the ability to say they have an AI strategy. Do not sell relief. "Boring is beautiful": deterministic automations are easier to build, easier to evaluate, less risky.

04Pick ONE KPI, with them, before you take money

Herk's first serious failure: he built a client a personal assistant and was told "we're not getting the value we paid for." Three mistakes — it was not the bottleneck, there was no KPI, he guessed the price. (The real bottleneck was that the client hand-wrote every proposal.)

ObjectiveProvable

"We book 5 meetings a week from manual coordinators today. Ideally the coordinators get 10 hours a week back and we book 10 meetings a week."

Baseline → target. Unarguable.

SubjectiveRefuse it

"I want to feel less overwhelmed." There is no way to prove you delivered it, which means there is no way to defend your invoice or price the next client. Refuse the engagement or renegotiate the metric.

ThenMeasure at 1, 2, 3 months — and show them

Herk's Goldman lesson: he built the automations, presented them on a 25-person call with his name on screen, and the manager replied "Good job, Arthur." If you do not make the value visible, you do not get the credit — even when the money is real. Capture the baseline before you build; his stated biggest regret is never capturing the "after".

05The proposal and the close

DoThree tiers, never one number

Open with three short paragraphs that are not about you: where they are now (their words, their numbers), where they want to go, why your expertise closes the gap. Then ~10% / 25% / 50% of annual value, with genuinely more outcome at each level. The middle is designed to win.

One number makes the decision "should we work with this person?" Three numbers make it "how should we work with this person?"

DoThe menu close Hormozi

1. Unsell what they don't need. "This is fine, you've got a couple of years on that." This is what buys the trust — you just proved you are not maximising the invoice. 2. Prescribe what they do need. 3. Fake choice — "Monday or Tuesday?" 4. Card on file.

"Getting the first sale is the hard part. Getting the approximate sale is significantly easier." Sell while the hood is already popped.

NeverDiscount. Reduce scope.

"Sounds like $15k isn't in the budget this quarter. Let's cut it back and call it v1 — when it's running and paying for itself we take the next step, and we've already scoped it."

Discounting on a frown teaches them that frowning works. Reducing scope keeps the rate intact and pre-sells phase two.

DoObjective milestones, every 30 days

✅ "The owner can ask the system a question, it pulls from the database and answers inside a minute." Provable, demonstrable. ❌ "The inbox agent works as expected." You will argue about "expected" for three weeks, unpaid.

Scope creep is a good sign — they are imagining more work with you. Never say no: "Great idea, let's put it in the v2 backlog so we hit the defined milestones fast and you get the benefit sooner."

MoneyPayment and running costs

50% up front, 50% on delivery. Never net-60. API keys, tokens and cloud on the client's card, in the client's name — every time. Your fee is design, consulting, development, testing; tokens are utilities. Include an estimated monthly run-cost with stated assumptions. Build testing into the price (+$1–3k); pay for pre-launch testing yourself, swap to their keys at launch.

WalkThey demand a number before discovery

They are price-shopping to compare you against Upwork, not hiring a partner. "Some people are looking for the cheapest labour. Don't take it personally."

Why this / why now / why me

Why this — is the thing they asked for actually the thing that produces the result? Why now — urgency is real money; a closing window or a competitor is worth a premium. Why me — "why not do it internally? Couldn't you vibe-code it? Couldn't an intern?" Awkward, and the most valuable question on the list.

When they say "honestly, my nephew could probably do that" — do not argue. Say: "He could probably build a version of it. Who's watching it at 2am when the model updates? Who scales it when volume is ten times what you expected?" Then stop talking. Their answer is the real reason they will hire someone.

When they won't give numbers (and they are not obliged to), ask the scale questions instead — they measure the ceiling, not the cost: How much time does this take you today? How many people touch it? What happens when it goes wrong? How many locations — how many of these come in on your busiest day?

06The LoopSuit offer ladder

RungOfferPriceNotes
0Working session (from cold)Free · 15–20 min⛔ Never a paid audit from cold. Arrive having already done work — redrawn screens, a flow teardown.
0bPaid diagnostic (inbound / referral only)$300–500Fine once they already know you. Wrong from cold.
1Design Sprint — one flow redesigned and shipped$3,000The floor, and the wedge. Fast, low-risk, provable — and it chunks up.
2Ship Sprint — designed and built, 2–4 weeks$6,000–12,000Priced at 10–20% of annual value, never off a day rate.
3Support retainer$400–800/moKeeps it working to spec: breakages, API changes, model updates. New features are a separate sale. Flat and standard across clients.
4Fractional design engineer$6,000–9,000/mo1–2 days/week embedded. Market rate $110–155/hr for AI-interface design engineers. The anchor.
5Priority / speed premium+20%Front of the queue. Same work, pure margin.

HormoziSpecialise into what everyone else won't do

He took a roofer off $18–25k full replacements and onto repairs — smaller tickets nobody wanted, but high volume, short cycle, reliable cash, and every repair puts you on the roof where the replacement gets sold. $32k → $247k/month in 90 days.

The transfer: every AI agency is chasing the $15k+ "full transformation" and nobody wants the $3,000 job. So take the $3,000 job. The Design Sprint is LoopSuit's repair — cheap to say yes to, provable, produces a case study, and it puts you inside the product, which is where the Ship Sprint and the fractional retainer get sold.

DoSpeed is pure margin

Supply-constrained — more demand than hours, which is exactly the FSH constraint — means you can sell front-of-line. "We're usually two weeks out; there's a 20% priority option that puts you first and guarantees Friday, or we don't charge it." Identical work, reordered queue, more money. Hormozi's roofer sold it for +$300 on the first attempt and the client immediately bought the next job.

DoProject → retainer, the actual transition

1. Sell one value-priced project against a named constraint. 2. Prove the KPI at 1, 2, 3 months. 3. Then: "We can move a lot faster if we stop re-scoping and re-pricing every project. Let's just keep removing constraints."

The hours→product conversion: a "20 hours at $50 = $1,000/month" retainer becomes 15 hours + a weekly strategy call + a Slack Q&A window. Same money, but the ratio shifts from 100% hours to ~50/50 — and your effective rate now climbs when you get faster instead of falling.

07The road to $10–15k/month — LoopSuit alone

FSH salary and the Shree retainer are deliberately excluded here. The question this answers is narrower and harder: can the LoopSuit skillset, sold to strangers, reach $10–15k USD/month? Yes — and here is the arithmetic and the sequence.

MixCompositionMonthlyVerdict
A · builds only2–3 Ship Sprints at ~$5k$10–15kFragile. Stops the month you stop selling, and 2–4 builds/month is the hard solo ceiling alongside FSH.
B · anchor + builds1 fractional @ $7k + 1 Ship Sprint @ $5k$12kThe right shape. One relationship carries most of the number.
C · mature1 fractional @ $7k + 1 Sprint @ $5k + 6 support retainers @ $500$15kWhere B goes on its own after two quarters.

Anchor on ONE fractional engagement. Lowest volume, lowest stress, survives a bad sales month, and it is the offer you are most obviously qualified for. Builds and support retainers stack on top.

what the pipeline has to produce (corrected 2026 numbers)1 close ≈ 4–5 booked calls ≈ 1,600–2,000 new-lead emails ≈ 3,200–4,000 total sends (two-step sequence: ~half your sends are follow-ups) At 200 sends/day (10 inboxes × 20) ≈ 4,400/month → ~1 close/month At 360 sends/day (18 inboxes) → ~2 closes/month

Cold email alone gets to roughly one close a month — at $5k that is half the target. The other half comes from Upwork, referrals and the fractional anchor. This is precisely why the channels must run in parallel: the cold fleet is not, on its own, a $12k/month machine at this inbox count.

WindowWhat happensLoopSuit revenue
OctPre-warmed inboxes sending. Upwork streak at 3/day. Zapmail fleet warming. First give-first Looms out.$0–3k
NovFull 9-inbox fleet live at 180/day. First ~2,000 sends. First case study published with before/after numbers.$3–6k
Decn ≥ 3,000 delivered — reply data finally readable. Copy test #1 concluded. First support retainer converted.$5–8k
Jan18 inboxes. Fractional conversation live from the strongest relationship. Two case studies.$8–12k
Feb–MarFractional anchor signed. Builds stacking. Retainers compounding.$10–15k

The honest framing

This is a two-quarter build, not a two-week one. Saraev's own figure is ~3 months to real results because the market reacts on a ~60-day lag — and the warmup clock sits in front of that. The single most important number in the table above is October's, and it is not revenue — it is whether the warmup clock started. Everything downstream is gated on it.

1Daily sends, without exception

Justin Lob — $20,354/month solo while working a 9-to-5 — did 3 Upwork applications a day and a running cold campaign, and did not break the chain. First $5k project closed around day 20. His mechanism was a red X on a whiteboard in his office where he could not avoid seeing it. "You don't know your limit until you reach it."

2Case studies with before/after numbers

Without the "after" you cannot price the next client, and the last one does not feel the value. Capture the baseline before you build.

3Qualify on the call, never on the reply

Terse, lowercase, unimpressive replies become thousands of dollars. You cannot tell from four words. Book it.

4First 5–10 clients aren't optimised for money

They are optimised for the review, the case study, the referral and the skill. Price to get the yes, then deliver disproportionately.

The three-month gate

Justin's own rule before leaving his job: three consecutive months at or above his salary. He hit it in January, February and March, and only then moved.

For you this is not a quit trigger — FSH is the residency vehicle and it stays. But it is the right financial gate for everything else: do not hire, do not take a VA, do not raise fixed costs, and do not treat LoopSuit as load-bearing income until it has cleared a real number three months running.

Message architecture · September 2026 · the biggest copy upgrade in this revision

Four questions, in two seconds.

A cold recipient asks four questions in a fixed order, in about two seconds. Answer each one before they ask it, in that order. Saraev's claimed delta from fixing only this: 0.06% → 0.3–0.5% booking rate — a 5–10× funnel on the same list.

01The four questions

#Their questionWhat answers itWhere it lives
1"Is this spam?"Something genuinely personal, about themSubject + first ~100 chars (the preview window)
2"Who the hell is this?"Social proof with a real number — "I built X for a business a lot like yours"Sentence 2
3"Why are you contacting me?"The observed inefficiency + a quantified claimSentence 3–4
4"So what do you want?""I've already started on X — here's a piece of it." One 15-min ask, zero management on their sideSentence 5–6
worked example · funded-founder lane · copy as MDSubject: your onboarding screens Saw the Series A note and went through the signup flow on my phone — three screens in I hit a dead end on step 2 (screenshot below). I'm a design engineer; I spent three years shipping TD Bank's mobile app and now I design and ship AI products end to end, solo. Flows like that one usually cost 20–30% of signups before anyone ever sees the product, which at your stage is real money. I already redrew the three screens — happy to send the mocks whether or not we ever talk. Worth 15 minutes?

The visibility hierarchy — fix in this order

1. Sender name → 2. Subject → 3. Preview text (first 100–150 chars — never waste them on "Hi [Name],") → 4. Offer. "If you can get your email clicked, that's 90% of the battle."

02Personalization without AI slop

NeverLet AI write the email

"I rarely, if at all, use AI in my copywriting these years… people that use AI really intensely in the process tend to suck."

Given the freedom, the model ruins itself — it produces the "the answer isn't more words, it's making them count" register that everybody now reads as a robot. AI manages most inboxes now; hyper-optimised "every line is value" copy is pattern-matched as spam instantly.

DoHard variables vs fuzzy variables

Hard — copied verbatim from the row: firstName, company, city.

Fuzzy — written per-recipient by the model from the row's other fields: theirThingInCommon, paraphrasedChallenge, oneThingIdFixOnTheirSignupFlow.

DoThe four disciplines

1. Name them as instructions — camelCase, long, descriptive: the variable name is the prompt. 2. Cap length hard — ~10 words for reason-type, ~5 for a paraphrased challenge; long output reads as AI. 3. Force per-row variation: read every row, never sample; even when two rows share source text the output must read differently. 4. Give 3–5 real examples of each variable — few-shot beats instructions for tone.

The governing rule underneath all four: the fewer AI-generated words as a proportion of the total, the lower the chance anyone clocks it. His claimed delta: ~1–2% fully templated → ~5% genuinely personalised.

03The give-first play

The highest-converting single message there is. Saraev hired a stranger off it, and uses it himself.

the original, verbatim in spirit"Hey Nick, I love your channel, been following a while. I think your headlines could be better, so I went through and wrote a bunch of options for you — they're all in this doc. I don't need you to work with me or anything, I just wanted to give you a way to improve your brand."

The explicit disclaimer of wanting anything is what makes it work. It is disarming, it triggers reciprocity, and it is completely honest if you actually do the work. That message got its sender hired for several months.

Your version is unusually strong

Because your give is visual and takes twenty minutes: redraw three of their screens and send the mocks. Nobody else in a founder's inbox is attaching design work. Keep the Loom to 60–120 seconds; longer and it does not get watched. Manual sends from a personal Gmail: 10–15/day maximum.

04Subjects, CTAs and the banks

UseSubject lines

3–6 words, informal, "do I know this person?" ambiguity. your [thing] · [their company] · [outcome] in 2 weeks · one thing · your onboarding screens · step 2 of your signup

KillSubject lines

quick question · fast question · anything with "partnership / opportunity / collaboration" · ALL CAPS · emojis · their full name in the subject.

UseCTAs that convert to calls

✅ "X or Y work for you?" (two concrete times) · ✅ "Worth your time? Let me know"

❌ "Reply yes and I'll send a Loom" — indirect asks convert worse to actual calls.

SeasonVoluntary disclosure

Deliberately reduce value density with one weird, personal, slightly off-topic line before the pitch: "I'm from Vancouver originally — anyway…" Unprompted self-disclosure makes recipients feel closer to you and breaks the AI-spam pattern.

Your real material: TD Bank stories, FSH design work, East Van, building products at 2am. One per email. Don't wreck the email.

05The offer rules that changed

DeprecatedPaid audits from cold

"The worst way to do it is: 'Hey, tell me everything about your business — to me, a complete stranger who clearly wants to sell you something. It'll only take an hour, with no proven or guaranteed ROI. After I'm done turning your business inside out, I might sell you something. Sound good?' No one will agree to that."

Paid diagnostics work only on warm traffic. From cold: free, 15–20 minutes, and you arrive having already done work.

DoAssume the problem

"Pick a bunch of problems, assume they have them, offer them, and see what people actually respond to." You are not qualified to diagnose a stranger before the call, and asking permission to try is the ask that kills the reply. State a specific problem as though it is true. Which ones are real is answered by reply data, after the fact.

DoGuarantees — the power is in the Z

"I guarantee X result in Y time, or Z." Refund, keep working free until it's hit, a $250 gift card — something concrete. A guarantee is not "I will change the universe for you"; it is "I'll do my best, and if it doesn't land you still get something."

Convert percentages into dollars and filter the list to match: "25% of revenue" → "an extra $15,000/month", and only mail businesses doing $60k+/month. Same promise, far more attractive, and the list does the qualifying.

NeverA price in a cold email

It disqualifies people before they have met you. Plenty of people who would happily pay $4,500 will close the email the instant their eye lands on the number. Price is a call conversation, always.

06Volume before cleverness

Burn the mailboxes to the ground

Saraev's business partner, to someone agonising over copy after 200 sends: "Right now you should send as many emails as possible. You should burn your mailboxes to the ground if you have to. You should burn your domains to the ground if you have to — because you don't have anywhere near enough data to draw any meaningful conclusions. You haven't even reached the point where you're good enough to provoke a reaction."

The progression is: get any reaction at all (even negative) → burn those inboxes → get new ones → get lukewarm replies → then get good ones. Momentum and feedback loop before optimisation.

How this applies here, honestly: your 9 Zapmail inboxes are a brand asset on domains you own — those never get burned. Obey the rule without destroying the fleet by buying 3–5 cheap pre-warmed throwaway inboxes and treating them as ammunition. They exist to generate data fast and be replaced.

Test500–1,000 sends per variant

Never 50–100. Judge on reply rate — positive-reply rate is too sparse at 1k scale. One variable per week. Big forks first, micro-iterations later.

TestSplit-test every campaign, always

You are going to run 100 campaigns anyway. There is no reason not to have 100 campaigns' worth of comparative data at the end of it.

TestThree motions in parallel

Don't pick one niche and hope — pick three, set the daily volume for each in advance, and let reply data crown the winner. "Treat it like a science experiment."

RuleQualify on the call, never on the reply

"People respond 'yes' and you think 'I don't want to give them the time of day' — some of those turned into thousands of dollars. If you don't call them, you'll never know." A terse, lowercase, unimpressive reply tells you nothing.

07Channels that have no warmup clock

Cold email has a 3-week warmup clock and a ~60-day market lag. These do not — and they are how the first money arrives. Target 50–100 outreach interactions per day across all channels.

ChannelVolumeWhy
Upwork3 applications/day
non-negotiable		Pre-qualified demand — people who have already said they want to buy, so it needs far less sales skill than cold email. Justin Lob closed a $5,000 project on day ~20 of exactly this streak while working a 9-to-5. Of the 50 applicants on a good job only 5–10 are real; the rest are copy-paste. Always attach an asset — a 3-min Loom, a doc, a diagram; ~40 of 50 attach nothing. After 10–20 of them you have a library and each application takes 5 minutes, not 30.
Founder DMs10–20/dayConnect first, wait 1–2 days, then DM. Never pitch in the connection request.
Agency white-label5 conversations/moGive enormous value free to agency owners with overflow, ask for nothing, become the obvious person they call when over capacity. Saraev's own overflow went to a man who did exactly this — he built it to $30–40k/mo solo. You have an unusually strong version of this available through Shree and the FSH network.
FiverropportunisticGenuinely working in 2026 — Saraev reports members clearing $3k/month on it, to his own surprise.

Your LinkedIn profile is load-bearing before a single message is read

You are judged on ~10 signals before anyone opens the DM: name, how you use the platform, whether you're online now, last-active, connection degree, verified badge, photo, headline. Fixing these alone roughly doubles booking rate.

Photo: foreground must contrast the background. Dark suit + dark hair + dark background = invisible; change the background colour. Headline: ❌ "I help X do Y" — generic, transactional, everybody says it. ✅ a specific proven result: "Ex-TD Bank mobile · I design and ship AI products end to end". Anyone can claim to help. Very few can point to what they shipped.

SkipCold SMS

Risk/reward is far worse than email — registration ties your full legal identity to the sends and penalties are severe. Saraev does not do it.

SkipAI voice agents for inbound speed-to-lead

People feel disrespected and close rates drop ~20%. Real case: $4k/month of humans produced $20k in closes; the $100/month voice agent that replaced them produced $5k. If a form fill is worth money, put a human on it.

DeferThe content treadmill

Saraev and Herk both put the threshold at $10–20k/month solo before a daily content habit earns its hours. Survivorship bias is why everyone says otherwise — the only people telling you to build an audience are the ones for whom it worked.

Proof content is different — case studies, a clean GitHub, three real project pages are sales collateral and should exist now. The daily-posting treadmill can wait.

Copied as Markdown