Go-to-market with Claude Code
Oct 29, 2026
5 min read

Attention first, then the app: a launch checklist

The four-step checklist from “Who's watching?” for launching an app when anyone can build one in a weekend.

Dan and Claude on a dock at blue hour, with hundreds of paper lanterns rising over a still lake.
Every lantern over the lake is somebody's launch. From Who's watching?

This is the checklist from Who's watching?, the episode where Dan asks Claude why nobody is using the app he built in a weekend. (If you commented LAUNCH on the Reel, this is what you asked for.)

You can build an app in a weekend now. So can everyone else, which is why most launches go up with nobody watching. The fix is to flip the order: get people first, then build. Find a real complaint, build an audience around the problem, and only build once they've shown up. You don't need to code for any of the first three steps.

Tick each step off as you go. Your ticks, your niche and your numbers are saved in this browser, so you can come back to it later.

0 of 4 done
  1. 01

    Find a complaint.

    People are asking for apps out loud. Search Google for site:reddit.com "is there an app that" plus a niche you know.

    site:reddit.com "is there an app that" your niche

    • Google
    • Reddit
  2. 02

    Film the problem, daily.

    Before the app exists, make a 15-second clip about the problem. End with “waitlist, link in bio.” Post one a day. The comments are your research, and whoever asks “when?” is your first user.

    Example clip · 15 s

    “Photographers, how long do your clients take to pick their photos?”

    …then: “Waitlist, link in bio.”

    • TikTok
    • Instagram
    • YouTube Shorts

    15 seconds and one a day are Claude's example numbers, not rules. Find a pace you can keep.

  3. 03

    Set up a waitlist with MCP, plus a $10 reservation.

    The waitlist

    Collect emails in Kit, Resend or beehiiv. All three have MCP, so Claude can check your signups and draft the launch email.

    • Kit
    • Resend
    • beehiivbeehiiv

    The proof

    Add a Stripe Payment Link that reserves it for $10, refundable. No code needed.

    • Stripe
    An email is a maybe. Ten dollars is a yes.

    Taking money for something that isn't built yet? Make the refund clear and automatic, and check your local consumer rules before you launch the link.

  4. 04

    Hit your number, then build.

    Pick the number before you post. Write it down here so you can't move the goalposts later:

    orin

    100 signups or 10 reservations in 14 days is Claude's example bar, not a rule. Pick yours.

    Missed it

    Move on to the next complaint.

    You just saved six months.

    Hit it

    Now build it.

    Claude Code builds the app, Supabase holds the data, Stripe takes the money, Vercel puts it online. All four have MCP, so you run them from one chat.

    • Claude Code
    • Supabase
    • Stripe
    • Vercel
  5. +

    BonusOn launch day, post it on Product Hunt.

    Your waitlist is your day-one crowd: email them the link before you post.

    • Product Hunt

What it looks like

This is the example from the episode: the kind of page you'd put behind "link in bio". PickPal isn't a real product; it's only there to show the shape. One clear promise, one email box, and one button that asks for real money, refundable.

Example · not a real product

PickPal

Clients pick their favorite photos in one tap.

For wedding photographers.

you@studio.comJoin waitlist
Reserve · $10 refundable

Go deeper

Each step has its own guide on the blog:

Sources