For builders Batch 01 · Fall 2026

The playbook.

Ship weekly. Show the log. Answer like a founder, not an applicant.

We can help you get in. Here is how. Partners read thousands of applications in a few weeks. The ones that stand out are short, specific, and backed by things that already exist. This page is the playbook we run on our own apps: build, talk to users, ship, write it down, post it. Bring your draft to the wall and we will mark it up.

Forma is an independent studio and is not affiliated with Y Combinator. No one can promise admission; we can promise you will apply with a real log.

How this works

Three things we actually do

01

Ship in public

Post what went out every Friday on your own ship log. We feature the best ones on the Forma home page.

02

Office hours in the Lounge

Bring your one-liner, your numbers, your rejection. Get a straight answer the same week.

03

Application review

Paste your draft answers on the wall below. We mark them up: what is vague, what is a real insight, what to cut.

What we shipped

The log we keep on ourselves

The playbook is only worth reading if we run it. Five lines from this week; the full log is on the home page.

  • Evensong submitted to App Review
  • Seventeen paths rewrittenShipped
  • Tomorrow You submittedIn review
  • Landed 1.1 submitted, subscriptions includedIn review
  • Nibble (formerly Macra) 1.4 liveLive

The full ship log

The application

The eight questions that get you in

The form changes a little every batch. The questions underneath it do not. For each: what a strong answer looks like, and the mistake we see most.

  1. What are you building?

    Strong. One sentence a stranger could repeat. Who it is for, what it does, what changes for them. "Landed keeps the weight off after Ozempic."

    Mistake. A category and three adjectives. "An AI-powered platform for holistic wellness" says nothing anyone can check.

  2. Who needs it, and why now?

    Strong. A named kind of person with a named week. Something changed recently that makes the old way stop working. Say what changed.

    Mistake. "Everyone" and a market-size figure copied from a report. Big markets are the result of a specific person, not a substitute for one.

  3. What have you built and shipped?

    Strong. Dates. "Submitted Sep 3, live Sep 9, 1.1 with billing Sep 14." A log that shows the pace, not a roadmap that promises it.

    Mistake. A demo that has never met a user, described as "an MVP". If it is not in someone's hands, say so plainly and say when it will be.

  4. How do you know people want it?

    Strong. Numbers with denominators. "Eleven of the forty people we showed it to opened it the next morning." Retention beats signups; money beats retention.

    Mistake. Waitlist size, likes, or "everyone we talked to loved it". Enthusiasm from friends is not demand.

  5. Why you?

    Strong. A reason you were already close to the problem before you decided to solve it. You lived it, you sold to these people, or you built the last version of this at someone else's company.

    Mistake. A list of credentials with no line to the problem. Partners are not hiring; they are betting.

  6. What is your unfair insight?

    Strong. One thing you believe that most competent people in the space would disagree with, and the evidence that made you believe it.

    Mistake. Restating the problem in a louder voice. An insight is something you learned, not something you noticed.

  7. How far along are you?

    Strong. The current state in plain numbers: users, revenue, weeks since launch, what shipped last week. Then the one thing you are stuck on.

    Mistake. Rounding up. "Pre-launch with strong traction" is two words that cancel each other out.

  8. How will you make money?

    Strong. Who pays, how much, how often, and whether anyone has already. A price you have charged beats a model you have modeled.

    Mistake. "Freemium, then enterprise, then ads." Three plans is zero plans. Pick the one you can test this month.

The weekly cadence

Five days. Same order every week.

This is the loop that produces a log worth showing. Miss a day and pick it up the next; the order matters more than the streak.

  1. MonBuild

    Pick the one thing that ships Friday. Write it in one sentence. Start.

  2. TueTalk to users

    Three conversations, thirty minutes each. Ask what they did last week, not what they want.

  3. WedShip

    Whatever exists goes out. A TestFlight, a page, a build in review. Small counts.

  4. ThuWrite the log

    Date, what went out, one number. Twelve words. Add what broke.

  5. FriPost it

    Put the log where people can see it. Then write Monday's sentence.

Bring us your draft

Post your one-liner or your draft answers. We reply on the wall. Rough is fine.

Forma is an independent studio and is not affiliated with Y Combinator. No one can promise admission; we can promise you will apply with a real log.

YC publishes its own application advice, and it is the first thing to read: ycombinator.com/apply. Questions about this page go to tryformaapp@gmail.com.