# First Dollar Playbook Status: current Updated: April 14, 2026 Use [COMMERCIAL_TRUTH.md](COMMERCIAL_TRUTH.md) for live pricing and traction claims. Use [VERIFICATION_EVIDENCE.md](VERIFICATION_EVIDENCE.md) and [RELEASE_CONFIDENCE.md](RELEASE_CONFIDENCE.md) for engineering and package-publish proof. ## Important correction ThumbGate has **no external paying customer and no external customer revenue yet** — 0 paid orders and 0 paid signups. Two historical Stripe charges totaling $20 exist in the ledger but are not external customer traction and do not count as the first dollar. The job is to book the **first real external dollar**. The shortest honest path is: 1. qualify one painful workflow 2. sell the **Workflow Hardening Sprint** 3. attach the proof pack 4. expand into Team rollout only after the first workflow proves out ## Why broad launch activity did not create paying customers The failure mode is not missing infrastructure. It is weak demand capture: - Low-audience social posts create impressions, not buyer conversations. - Generic "agent memory" copy sounds replaceable by `CLAUDE.md`, `.cursorrules`, or another prompt file. - A GPT, plugin, npm package, and docs page are distribution surfaces, not revenue by themselves. - Revenue starts when one buyer names one repeated failure they already care about and agrees that blocking it before the next tool call is valuable. The first-dollar motion therefore has one job: move a cold user from curiosity to one caught repeat. ## First-dollar activation ladder 1. **GPT proof:** send the user to the live ThumbGate GPT and ask for one bad answer, risky command, deploy, PR action, or agent plan they do not want repeated. 2. **Typed feedback:** have them type `thumbs down:` or `thumbs up:` with one concrete sentence. Do not claim ChatGPT's native rating buttons feed ThumbGate. 3. **Local enforcement:** install with `npx thumbgate init` where the agent executes, then show the saved lesson, generated rule, or caught repeat. 4. **Solo Pro trigger:** upgrade only after a real caught repeat when the buyer needs more captures, the dashboard, DPO export, or proof-ready evidence. 5. **Team trigger:** if there is a shared repo, review queue, deploy workflow, migration process, or release approval problem, convert to the Workflow Hardening Sprint. ## Operating thesis - Do not rely on low-intent self-serve traffic to produce the first repeatable revenue. - Do not lead with "agent memory" or generic developer-tool subscriptions. - Lead with one workflow that is not safe enough to ship or defend today. - Sell proof, enforcement, and rollout confidence around that workflow. ## Best buyers - Claude-first consultancies delivering workflows for clients - platform teams owning standards, approvals, or release controls - AI product teams with one workflow owner and one repeated failure pattern ## What to say > We are not pitching another agent platform. We harden one workflow that already matters, so it becomes safe enough to ship team-wide. The offer is a Workflow Hardening Sprint: one workflow, one owner, one proof review. ## 14-day execution loop ### Days 1-2 - Send the four warm outreach messages from [marketing/team-outreach-messages.md](marketing/team-outreach-messages.md). - Add ten more qualified targets from consultancies, platform teams, and AI product teams. - Use [OUTREACH_TARGETS.md](OUTREACH_TARGETS.md) to reject weak-fit calls early. ### Days 3-7 - Book qualified calls only. - Ask for: - the workflow with business value - the repeated failure or rollout blocker - the buyer or champion - the runtime already in use - Send the sprint brief and proof pack the same day. ### Days 8-14 - Turn every qualified call into one of two outcomes: - a paid sprint proposal - a named pilot agreement - If a lead wants "more info" but cannot name a workflow, push them back to the free path, guides, or Pro page instead of dragging the sales cycle. ## Required proof pack - [WORKFLOW_HARDENING_SPRINT.md](WORKFLOW_HARDENING_SPRINT.md) - [VERIFICATION_EVIDENCE.md](VERIFICATION_EVIDENCE.md) - [RELEASE_CONFIDENCE.md](RELEASE_CONFIDENCE.md) - `proof/compatibility/report.md` - `proof/automation/report.md` ## Commercial guardrails - Do not count traffic, stars, installs, directory listings, or compliments as revenue proof. - Do not promise ROI before the buyer names the workflow and the blocker. - Do not let Team pricing outrun [COMMERCIAL_TRUTH.md](COMMERCIAL_TRUTH.md). - Do not let package-publish claims outrun [RELEASE_CONFIDENCE.md](RELEASE_CONFIDENCE.md). ## Definition of done - one booked workflow hardening sprint - or one named paid pilot agreement Anything weaker than that is still pipeline, not revenue.