Y Combinator Application Writer and Editor

Read every question first, and fill in every field. Longtime YC partner Dalton Caldwell says in his 2023 talk How to apply and succeed at YC² that many weak applications skip the founder biographies. Partners also weigh whether you followed the directions, video included.

Caldwell reads each application by telling himself a short story: who the founders are, what they've built and whether anyone wants it. "If I can tell this kind of story about your company by reading your application," he says, "it means it was a good application."

Graham starts with "What is your company going to make?" because, he writes, he needs "something to hang the application on." One test is how close your answer brings a reader to reproducing the idea. "We are going to transform the relationship between individuals and information" fails it. "A database with a wiki-like interface, combined with a graphical UI for controlling who can see and edit what" passes.

Skip the hype: Graham says marketing-speak is "just noise" to partners. Don't fear sounding narrow, and compare yourself to something familiar when it's accurate, as in his example "It's eBay for jobs." YC still publishes Dropbox's original application³ as a model.

For each founder, the form asks for something impressive they've built or achieved, and Graham calls it "the most important question on the application." Size matters more than type, and one specific example beats any adjective: "Jordan is an exceptionally dedicated person" carries no weight. He'd leave out the startup itself; YC already knows about it. If nothing stands out, name the hardest thing you've done.

Graham treats the question about a time you hacked a non-computer system to your advantage as a wildcard. A good answer, he says, can send him back for a second look at an application that otherwise seemed unpromising. Tell one true story with a clear result.

Make technical skill easy to see. Caldwell says a team gets five times better odds of an interview when at least one founder could be hired into a technical role at a top YC company. Name what each technical founder built, and where.

Describe what exists today with counts and dates ("14 bakeries pay $80 a month, up from 3 in March"), not "strong early traction." Say how long each founder has worked on it, and how much of that full time. Caldwell counts leaving a job for it as "evidence of seriousness."

Being early is normal. YC's FAQ⁴ says on average 40% of the companies it funds in each batch are just an idea, and most have no revenue. If that's you, say so in one line, then show what you've learned; a customer discovery synthesis turns interview notes into counts a partner can weigh.

Saying your product will be well designed and easy to use "is not an insight," Graham writes; whoever built the current one presumably tried that too. Say exactly what you'll do differently and why it will be enough. YC's interview guide⁵ expects you to know the existing products in your market and "what, specifically, is wrong with them."

Then raise the hard problems yourself. "If we think of a problem you don't mention, we'll assume it's because you haven't thought of it," Graham warns. Serious obstacles don't put YC off, he says, as long as you show you're aware of them and have at least a theory for overcoming them.

Don't skip the question about other ideas you considered. Graham says it's "quite common" for YC to fund a team to work on one it listed as an alternate.

When you explain how you'll make money, say who pays, how much and how often. Then size the opportunity from customers you can count; a TAM, SAM and SOM calculator does the bottom-up math.

Give exact percentages and amounts in the equity answers. You don't need to incorporate first, the FAQ says, but cofounders should agree on the split. YC's 2007 form put it well: "This question is as much for you as us." A founder equity split template helps you settle it.

Before you submit, take Graham's advice: print the application, "take a red pen and cross out every word you don't need," and make what's left as specific as you can. Caldwell puts it more bluntly: "Don't write a novel."

Length is now a warning sign. In August 2026, Entrepreneur reported⁶ partner Tyler Bosmeny's observation that YC applications had grown 60% longer in three years. Partner Pete Koomen charted the word "wedge" rising from under 1% of applications in spring 2025 to over 20% in summer 2026. The article reads both as signs of AI writing.

Finally, check every claim. Partners have your application open during the 10-minute interview, and the interview guide says YC may ask you to verify numbers you state there. Keep a source for each figure, and have a cofounder read it all aloud.

Working from scattered notes and an old draft? The tool at the top of this page drafts every answer from them and marks what only you can write.