One persona for the Experts launch. Built against the persona-gtm.md 3-stage playbook with the persona-gtm-deliverable-template.md shippable format. Pressure-tested by a 7-voice growth + AEO + guerrilla council. Solo-safe Day-0 execution plan.
Defines WHERE this persona fits in Experts' offering before drawing the persona. Confidence: Verified · Researched · Assumption.
| # | Dimension | Value | Conf. | Reasoning |
|---|---|---|---|---|
| 01 | Description | Buyer hires Experts as a callable JSON layer of stage-specific operator archetypes that they invoke inside Cursor / Claude Code / their own coding agent, to make pricing / ICP / channel / agent-role decisions without context-switching out of the build workflow. | Researched | Maps to T6 product surface: /catalog.json live, three integration snippets in hero, MCP on roadmap. Source: SESSION-HANDOFF.md. |
| 02 | Representative segments | Technical founder OR founding engineer #1–#3 at a 5–30 person seed-to-Series-A B2B SaaS or AI-tooling startup. Daily user of (Cursor, Claude Code, Windsurf) AND (LangChain, LlamaIndex, CrewAI, AutoGen, OpenAI SDK, Anthropic SDK). | Researched | Council 12899d09 unanimous. Matches catalog archetypes agentic-engineering-builder, build-in-public-agent-engineer. |
| 03 | Buyer sophistication | High technical, moderate business. Reads code before reading copy. Buys sub-$50/mo dev-leverage tools without procurement. Hostile to marketing-speak. | Researched | Inferred from monthly tool stack ($157–$546/mo). Matches Cursor's $2B ARR (Vantage 2026) at $20–$60/mo seat pricing. |
| 04 | Commercial depth | Operator $19/mo · Studio $49/mo · Charter $499 lifetime cap 1,000 · Custom Archetype $149 one-off (24h SLA). No enterprise tier in v1. No live chat. | Verified | research/structure-pricing.md. Pricing decision locked at A4 (keep until 100 paid Operators, then re-test $29). |
| 05 | Typical sales motion | Self-serve, founder-led, distribution-via-product. 2-channel core Days 0–30 (X DMs + GitHub issue/reply), Discord = named-thread only, Show HN at Day 21+, PH after ≥20 testimonials, newsletters as retargeting. | Researched | Council 40dbd358 fix #1. Solo capacity protection. |
| 06 | Strategic upside | (a) /catalog.json + MCP makes Experts callable inside Claude Code daily session. (b) Charter members create asset that compounds with each archetype. (c) Build-in-public output of users becomes free distribution. | Researched | Lighthouse-calibration doc (research/target-persona-*.md). MCP repositioned to retention hook per council fix #6. |
| 07 | Entry product | Outbound entry: Founder Stack Gap Audit (free, 7-min diagnostic, 24h SLA) → Operator $19 trial. Inbound entry: Free Browse → 7-day reverse trial of Operator. One-off entry: Custom Archetype $149. | Researched | Council fix #3: diagnostic-to-activation flow replaces generic trial CTA. |
| 08 | Target deal size | Entry $19/mo recurring · Studio $49/mo (20% attach) · Charter $499 one-time (12% attach over 24 mo) · Custom $149 each (25% attach). Blended Y1 ARPU ~$24–$30/mo equivalent. | Researched | Council 12899d09 + Cursor/Claude/Copilot 12-mo retention 70–85% (Lenny 2026 PLG benchmarks). |
| 09 | Upsell path | Operator → Studio (when 3 saved workflows OR team invites >2) → Custom $149 add-ons → Charter $499 lifetime. Wave-2 (not Stage-1 scope): fractional CMO/CTO scaling offer, accelerator/community white-label, VC portfolio license. | Researched | Council fix #6: explicit in-product upgrade triggers, not passive pricing-page upsell. |
| Category | Field | Value | Conf. | Reasoning |
|---|---|---|---|---|
| Demographics | Age | 31 | Researched | Median founder age 29–33 (Wing VC 2024); AI-startup subset modal 31. |
| Demographics | Gender | Male | Researched | 78% GitHub-active AI/ML devs male (GitHub Octoverse 2024 + SO Survey 2024). |
| Demographics | Income | $85,000 base + $200k unrealized equity | Researched | Seed-stage US founder salary median $80–100k (Kruze 2025 + Pilot 2025). |
| Demographics | Occupation | Co-founder & CTO | Verified | Council A2 lock: technical founder, not employee engineer. |
| Demographics | Family status | Partnered, no kids | Researched | 60% US seed founders 28–35 partnered no children (First Round 2024). |
| Offline & Online | Key offline events | AI Engineer Summit SF (June + Dec, ai.engineer) · LangChain Interrupt (annual since 2025) · Anthropic Code with Claude (SF, launched April 2025) · ODSC West (Oct, Burlingame) · SaaStr Annual (Sept, San Mateo). 3–4/year. | Verified | Events verified against ai.engineer, langchain.com/interrupt, anthropic.com, opendatascience.com, saastr.com. |
| Offline & Online | Key online platforms | X 500 followers (posts 1–2x/week) · GitHub daily commits · LangChain Discord active in #general+#langgraph · Anthropic Discord · Cursor Forum · Latent Space Discord · HN passive read daily | Researched | Top platforms per Stack Overflow Developer Survey 2024 + LangChain/Latent-Space community patterns. |
| Role/Personal | Exact title | "Co-founder & CTO" on LinkedIn · "Founder" + lowercase handle on X bio | Researched | YC W24-W25 batch convention per public profiles. |
| Role/Personal | Highest education | BS Computer Science, University of Waterloo (or CMU, Stanford, UIUC, Cornell, Cambridge, IIT Bombay) | Researched | YC 2023-24 educational data: ~75% CS/Eng degree, top schools per public batch listings. |
| Role/Personal | Notable career highlight | Eng #7 at Series-C AI infra startup (Pinecone/Modal/W&B pattern) before quitting Q1 2024. One ~400-star personal GitHub repo in agent space. | Researched | Ex-unicorn-eng → founder dominant in YC W24-25 (Tan public commentary); 400 stars = modal credibility threshold (Latent Space, Show HN). |
| Responsibilities | Daily tasks | 5–7 hrs Cursor + Claude Code · 1 hr customer-dev · 30–60 min Latent Space + Ben's Bites + GitHub trending · 30 min cold DMs (5–10/wk) · 30 min X drafts · 15 min LangSmith/Helicone evals | Researched | Composite Brockman + Nakajima + Levels public routines + First Round State of Founder Time 2024. |
| Responsibilities | Activities relevant to Experts | Before designing a new agent role: 30–90 min scrolling X / podcasts / Notion / bookmarks for "what would a Series A B2B SaaS GTM person do" → pastes tweet fragments into Cursor → ships sub-optimal agent role. 1–2x/sprint. | Verified | Pain loop confirmed by council 12899d09 (7 of 7 verdicts identified this). |
| Needs & Objectives | Company goals | $1M ARR within 18 months. Close $5M Series A within 24 months. 3–5 enterprise reference logos. Avoid commoditization by next Anthropic/OpenAI release. | Researched | YC + a16z 2025-26 public commentary on seed → Series A bar. |
| Needs & Objectives | Personal goals | Equity outcome >$5M after Series B. Reputation in AI builder community (speak at AI Engineer Summit). Maintain agency. Avoid VC pressure that strips identity. | Assumption | Personality-fit hypothesis; not surveyed. |
| Pain-points | Functional pains | (a) Generic LLM advice with no stage/sector specificity (b) random tweets pasted into Cursor as agent spec (c) decision-fatigue on pricing/ICP/channel/role-design (d) advisor calls scarce/slow/expensive (e) no reusable decision artifact at the end | Verified | 5-pain pattern from council 12899d09 + research/target-persona-yohei.md. |
| Pain-points | How Experts helps | Callable JSON archetype layer (/catalog.json + MCP at Day-14 trial milestone) · decision card per archetype (when/when-not/stage/failure-mode) · one-click export to Cursor/Claude prompt block · 5 prebuilt builder packs · 3 real before/after build-log artifacts shipped pre-launch · weekly Founder-Engineer Build Notes | Researched | Solution design + council fixes #4, #5, #6. |
| Personality | Key traits | (1) intellectually curious (2) low patience for fluff and marketing-speak (3) build-first / write-second (4) opinionated and confident in code judgment (5) default-to-public build log | Researched | Composite from yohei-nakajima.md + pieter-levels.md + marc-lou.md voice files in ops/voices/. |
| Company info | Size | 12 employees | Researched | Median 12-month-post-seed US AI startups (Pitchbook 2025 + First Round headcount benchmarks). |
| Company info | Revenue / ARR | $250,000 ARR | Researched | Post-PMF-test seed B2B SaaS $100–500k ARR before Series A (Lenny 2025 SaaS benchmarks); $250k modal. |
| Company info | Stage & footprint | Seed-funded · $2.2M raised on $12M cap (SAFE) · Delaware C-corp · remote-first · co-founder in SF Bay Area · contractors EU + India | Researched | Carta Q4 2024 seed-round median $2.5M; $2.2M leaner-cohort. |
| Power level | Signing authority | Sole <$1,000/mo · $1k–$5k requires Slack-check with co-founder/CFO · >$5k requires board-paper-trail | Researched | Carta + Pilot 2024 seed-stage spending-authority data. |
| Budget owner | Annual budget | ~$60,000/yr tools+cloud+AI API · ~$24,000 of that is "thinking/decision tools" line item | Researched | Pilot.com 2024 seed-stage SaaS spend benchmarks. |
| Budget owner | Spending autonomy | $0–$200/mo no questions · $200–$1,000 signs alone · >$1,000 discusses with co-founder | Researched | Carta + Pilot 2024. |
| Identity (council fix #2) | Hard ICP triggers | Shipped code in last 30 days · posts code or build logs publicly · uses Cursor or Claude Code · uses or discusses LangChain/LlamaIndex/CrewAI/AutoGen · seed or pre-seed funded · team 2–15 · owns product AND GTM consequences | Verified | Council fix #2 (4-of-7 voices identified this as the breaking weakness). Required filter on every targeted account. |
24 fields · 3 Verified · 19 Researched · 2 Assumption · 87% V+R against 60% playbook floor
Reply rate of 8% collapses to under 5% if list quality dilutes. Onboarding throughput suffers. Required filters on every account on the 2,500-target list.
Public GitHub commits OR public build logs in last 30 days.
Reads as a "first-class user" of one of: Cursor, Claude Code, Windsurf.
Uses or discusses publicly: LangChain, LlamaIndex, CrewAI, AutoGen, OpenAI SDK, Anthropic SDK.
Owns both code AND GTM decisions. Not employee engineer, not academic.
5–30 person team. Pre-Series-A. Capital-efficient or just-funded.
Employee engineers, fractional execs, indie hackers without agent stack, researchers without commercial path, >Series-A. Re-evaluated in Wave-2.
| Source | Count | Notes |
|---|---|---|
| US seed AI / B2B-SaaS startups funded 2024–25 | 12,500 | Crunchbase 2025 |
| EU equivalent | 3,500 | Dealroom 2025 |
| APAC + India | 2,200 | Tracxn 2024 |
| LatAm + AMEA | 600 | Crunchbase 2025 |
| Total companies globally addressable | 18,800 | Sum of regional |
| At ~2 buyer-grade individuals/company | 37,600 | Founder + first eng |
| English-language addressable | 28,000 | Final TAM |
Cross-check: Cursor reported $2B ARR Feb 2026 (Vantage) at $20–60/mo seat = 700k–1.4M paid devs; our specific founder-engineer subset = ~5% overlap = ~50k. We use lower TAM of 28k. Above 50-company niche threshold. Funnel marketing appropriate.
| Stage | Need | Buyer's words | Experts solution | Channels | Notes |
|---|---|---|---|---|---|
| Awareness — 4 needs | |||||
| Awareness "Who are you?" | Discovery | "Who are you and why should I care?" | Show HN one-shot (Day 21+) · Product Hunt (after ≥20 testimonials) · selective Latent Space placement (Wave 3 only) | Show HN · PH · Latent Space | DELAYED per council fix #1 |
| Credibility | "Is this real or slop?" | 3 real build-log artifacts using /catalog.json in Cursor + Claude Code · open-source MCP recipe · public changelog · weekly Founder-Engineer Build Notes (teardown + benchmark + failure) | GitHub README · X build-log threads · Antonis LinkedIn | Council fix #4: proof asset on every outbound touch | |
| Specificity-proof | "Show me how it's different" | Comparison page (Experts vs prompt-ChatGPT vs X-threads vs advisor-calls vs CrewAI) · live curl /catalog.json block in hero · "Founder-Engineer Build Notes" archive | Programmatic SEO/AEO pages · X "vs" threads | Council fix #7: SoftwareApplication + FAQPage + Organization + BreadcrumbList JSON-LD · llms.txt · "Last verified" timestamps | |
| Reachability | "Where do I keep seeing this?" | GitHub-issue reply-guy with archetype slug + JSON snippet on real issues in agent-framework repos | GitHub issues (LangChain, LlamaIndex, CrewAI, AutoGen, llama_index) · Discord = named-thread only | Quota ~5/day; Wave-1 channel | |
| Activation — 3 needs | |||||
| Activation "Does this work for me?" | First-value | "Can I use this in 30 seconds?" | Free Browse · live /catalog.json snippet above the fold · one free pack (5 archetypes) downloadable as JSON without signup | Landing page hero · free pack · free /catalog.json | First action = paste-in-Cursor; eliminate friction |
| Diagnostic | "Tell me what's wrong with my stack first" | Founder Stack Gap Audit · 7-min personalized teardown from public stack signals · 24h SLA · free for ICP-fit accounts | X DM "reply stack" CTA · landing CTA "Get your Founder-Engineer GTM teardown in 24h" | Council fix #3: diagnostic-to-activation flow replaces generic trial CTA | |
| Trial | "Let me try the full thing" | 7-day reverse trial of Operator $19 · downgrades to Browse on expiry (no lockout) · Day-3 mandatory artifact export: ICP pressure test OR 30-day channel sprint OR 10-message outbound pack OR AEO-ready page block | Signup flow · in-app onboarding | Council fix #5: if no artifact in 48h, manual founder email. Activation = Weekly Activated Operators (North Star) | |
| Revenue — 3 needs | |||||
| Revenue "Show me it works" | Pricing fit | "$19 feels right; what do I get for $49?" | Three tiers visible: Operator $19 (solo validation, single lane, weekly sprint) · Studio $49 (team handoff, 3+ saved workflows trigger) · Charter $499 once (founding 1,000 · public counter · roadmap input · quarterly teardown) | Pricing page · in-app upgrade triggers | Council fix #6: usage fences + explicit upgrade copy · plan names unchanged |
| Custom on-demand | "I need an archetype you don't have" | Custom Archetype $149 · 24h SLA · manual fulfillment Y1 · pay-after-delivery · only after paid milestone event | /request · X DM · post-trial email | Founder capacity caps at ~50/year | |
| Scarcity | "Will I miss the founding window?" | Charter $499 lifetime cap-1,000 · public counter ("x/1000 filled") visible on pricing page · founding-member opt-in listing | Pricing page Charter card · public Charter members page | Y1 headline number, not MRR | |
| Retention — 2 needs | |||||
| Retention "How do we expand?" | Freshness | "Is the catalog alive?" | Weekly catalog drops (5+ archetypes/week) · public changelog · weekly "Operator Performance Digest" email with shipped tasks, time saved, unresolved backlog | Email · changelog · X/LinkedIn post · in-product weekly Build Review | Council fix #5: weekly ritual sustains 70% Y1 retention |
| Embedded workflow | "Still callable from Claude Code?" | MCP server unlocked at Day-14 trial milestone for qualified paid users · "Founding Operator Access" · weekly office-hours cohort · shared implementation playbook · new archetype auto-publishes as MCP tool | MCP server · automated build pipeline | Council fix #6: MCP as retention hook, not awareness hook · waitlist-only until 100 paid Operators | |
| Referral — 2 needs | |||||
| Referral "What's in it for me?" | Public proof | "Why would I tell my followers?" | Charter member opt-in public listing with handle + project · build-log artifact format ("I used Experts to ship X, here's the JSON I called") · anchor-account RT cycle | Charter members page · X build-log examples | Vanity + capped supply |
| Affiliate | "Financial upside?" | 30% recurring on Operator/Studio for 12 months · 20% one-time on Charter | /affiliate · in-app share link · Charter dashboard | Benchmarked to ConvertKit, Substack, Linear referral programs | |
| Stage | Primary KPI (movement) | Secondary KPI | Y1 Target | Benchmark |
|---|---|---|---|---|
| Awareness | Targeted-account → reply rate | Cumulative unique landing visitors | 8% reply · 25,000 visitors | Personalised DM 8–15% (Cleverly 2026) |
| Activation | Visitor → Stack Gap Audit signup | Audit → trial conversion · paste-in-Cursor rate | 40% audit→trial · 25% paste | Diagnostic-first vs trial-first +15–25% (council fix #3) |
| Revenue | Trial → Paid Operator | Operator → Studio attach % · Day-3 artifact export rate | 25% trial→paid · 20% Studio · 60% artifact | Dev-tool trial→paid 15–35% (Vantage 2026) |
| Retention | 12-month Operator retention | Custom-request attach / quarter · Weekly Activated Operators (NSM) | 70% net retention · 1.5 customs/qtr/Studio | Dev tool 12-mo retention 65–80% (Lenny 2026) |
| Referral | Referrals per paying logo Y1 | % Charter opting into public listing | 0.4 referrals/logo · 60% public opt-in | Affiliate 0.2–0.6 per dev-tool comps |
Year-1 capacity constraint: founder time on Custom orders caps at ~50/year before pulling time from product OR content cadence. Demand is not the bottleneck; Antonis is.
PMF gate (council fix from growth-hacking voice): at 50 active trials, run the Sean Ellis 40% test ("how would you feel if this disappeared?"). If <40% "very disappointed" → freeze acquisition expansion, iterate onboarding. If ≥40% → expand outreach volume and add Wave-2 channels.
| Product | Price | Attach | GM |
|---|---|---|---|
| Browse (free) | $0 | n/a | n/a |
| Stack Gap Audit (free, ICP-gated) | $0 | lead-gen | n/a |
| Operator (entry) | $19/mo | 100% | 90% |
| Studio (upgrade) | $49/mo | 20% | 88% |
| Custom Archetype | $149 each | 25% | 60% |
| Charter (lifetime · cap 1,000) | $499 once | 12% over 24mo | 95% |
| Entry deal | $19 first month |
| Y1 weighted total | $260 |
| Y2 weighted total | $174 |
| Y3 weighted total | $128 |
| 3-yr weighted LTV | $562 |
| CAC (blended · community-led) | $135 |
| LTV : CAC | 4.2 : 1 |
| Payback | 9.0 months |
| Sales cycle (first close) | 14 days median |
| Sales cycle (expansion) | 90 days median |
Sensitivity callout: the single assumption that flips this model is 12-month Operator retention <60%. At 50%, LTV drops to $410 and LTV:CAC = 3.0×, payback >12 months. Defending retention requires both the weekly catalog drop AND the MCP server (Day-14 milestone). Without both, retention slips below 60%.
| Year | Team | New Operators | New Studios | New Charters | New Customs | Total revenue | Coverage |
|---|---|---|---|---|---|---|---|
| Y1 | 1 (Antonis) | 250 | 50 | 80 | 40 | $109,000 | Beachhead only |
| Y2 | 2 (+ part-time community) | 600 | 180 | 250 | 120 | $321,000 | Beachhead + Wave-2 fractional |
| Y3 | 3 (+ archetype curator) | 1,400 | 400 | 420 | 280 | $681,000 | Beachhead + Wave-2 + accelerator/white-label |
| 3-yr | 2,250 | 630 | 750 | 440 | $1,111,000 |
[X DM template — to @handle who shipped a public agent build-log this week] hey @handle, saw your [specific build] thread. reply stack and i'll send your 7-min Founder Stack Gap Audit. no slides. one teardown + 3 operator moves you can ship this sprint. free for seed/preseed founders shipping code weekly. — a. [Landing CTA] Get your Founder-Engineer GTM teardown in 24h. → paste your repo or company URL [After audit delivery] Run your first Operator on a live backlog item in 12 minutes. → 7-day trial · no card
Growth-hacking PMF + four-fits + atomic-network + AEO + zero-click + attention-arbitrage + community-led. Council 40dbd358.
Experts · Specifics · ExpertCrew. Phase-7 trademark research recommends defensive register expertcrew.com ($10–$15).
decide todayCouncil was 4-of-7 for Stack Gap Audit (lower friction, earns the trial). 3-of-7 for direct trial activation. Run 100-account A/B in Wave 1.
A/B 50/50 in Wave 1Path A (recommended by growth-hacking voice): waitlist until 100 paid Operators + month-2 retention signal. Path B (zero-click + community-led): unlock at Day-14 trial milestone for qualified paid.
Path A unless MCP is <3 days of buildPhase-7 recommends $29/$79/$799. Council leans keep $19 until 100 paid Operators + activation data clears, then re-test pricing.
keep $19 → raise at 100