Zero to $3,000

The session is a case study; the guide is how you run the same play. You leave with a first channel, a budget, a go/no-go number, and a funnel that's actually instrumented.

Five-minute setup

Do this before the session starts, or in the first ten minutes while the framing happens. Nothing here is required to attend.

1. Have an agent installed

   Claude Code, the Codex extension, or whatever you already run. Every skill file below is plain markdown with frontmatter — it works in anything that reads a skills directory.

2. Get the skills

   Download this session's bundle and unzip it at the root of your project. It writes into .claude/skills/ and nothing else.

   unzip -o zero-to-launch-02-b2c-ggbucks-case-study.zip -d .

3. Not sure which skill you need?

   Run /ask-santos and describe the problem in your own words. It names one skill, says why that one, and tells you what comes after it. You don't have to learn the catalogue.

4. Point it at a real product

   An idea in a notes app counts. The skills read your repo, your site and your README before asking you anything, so run them from inside the project rather than a blank folder.

5. Make somewhere to put the output

   The template repo already has the folders. If you're working in your own project, a docs/gtm/ directory is fine — the skills write markdown, they don't care where.

The steps

1. 01

   Pick one channel, not a mix

   /channel-plan

   The first thirty days of ggbucks were one channel at a time. Two channels at once means you learn nothing from either.

   Done looks like: Channel, the specific test, the budget, and the number that decides whether you keep going.

2. 02

   Write the go/no-go number down before you spend

   Decided in advance it's a threshold; decided afterward it's a rationalisation. Cost per signup, reply rate, whatever fits — one number, one date.

   Done looks like: "If it isn't ___ by ___, I stop." In the file, not in your head.

3. 03

   Instrument the funnel now, not in October

   /analytics-wire

   Events per stage, then verify each one actually fires. Receipts are 30% of the October score and this is where they come from.

   Done looks like: Every stage in your session-01 funnel has an event, and you've seen each fire in the dashboard.

4. 04

   Take a baseline snapshot

   Today's numbers, saved with today's date. In eight weeks this is the before half of your launch story.

   Done looks like: A screenshot or a table committed to the repo.

5. 05

   Decide whether paid is worth trying at all

   /paid-test

   The gate, not the execution. It checks the funnel is instrumented, there's organic demand, and you can afford a test big enough to read. If those two budget numbers don't overlap, it tells you to stop — which is the useful answer.

   Done looks like: A go or a no-go, with the number that decided it written down.

6. 06

   Treat community as a funnel stage

   /community-plan

   Superfans → ambassadors → organic. It's a stage with a conversion rate, not a Discord you open and hope.

   Done looks like: You can name the step that turns a user into someone who brings other users.

7. 07

   Wire up the weekly report

   /bi-agent

   An agent that reads your analytics on a schedule and tells you what moved, in plain English. Set it once in August and it reports all the way to the hackathon.

   Done looks like: One report has landed. If it hasn't run once, it isn't set up.

What you leave with

02-b2c/ — channel-plan.md, analytics.md with a verified event per stage, community-plan.md, and a bi-agent/ that has reported once.

If you get stuck

No traffic to instrument yet? Wire the events anyway. Ten minutes now beats reconstructing eight weeks of history the night before judging.

Or ask in Discord — mentors are in there between Wednesdays, and everyone else's work in progress is too.