YC Application Advice: What We Learned Going Through the Process

I went through the YC application process and got in. I also know a lot of founders who didn't get in and should have, and founders who did get in and probably shouldn't have. The selection is imperfect. But there are things that reliably increase your odds, and I'll share them here.

This isn't a generic "work hard and believe in yourself" post. These are specific things that worked and didn't work.

What the YC Application Actually Tests#

The YC application is testing a small number of things:

1. Can you build something? — Demonstrated by what you've shipped
2. Do users actually want it? — Demonstrated by traction, even if tiny
3. Do you understand your market clearly? — Demonstrated by the quality of your problem framing
4. Is the team credible? — Domain expertise, previous building experience, co-founder chemistry

Everything else in the application is noise. The four things above are signal.

The Most Important Part of the Application#

YC asks: "Why did you pick this idea to work on?" This is the most important question on the application.

Most founders write something about market size or trends. That's wrong. The right answer is personal and specific. Why are you uniquely positioned to build this? What did you experience that others haven't? What insight do you have that made you start this instead of something else?

For DenchClaw, the honest answer was: I spent years building enterprise agentic workflows at scale. I saw how much value AI could deliver when it had deep context on your data and could act on your behalf. I also saw how bad the existing tools were for knowledge workers who weren't enterprise customers. That combination — lived experience + a gap I personally felt — is what YC looks for.

Generically: "AI is transforming X" is not a compelling answer. "I spent five years in X and I know exactly which part of the workflow is broken and why nobody has fixed it" is.

On Traction#

You don't need much traction to get into YC, but you need some. Zero traction with a great idea is less compelling than small traction with an okay idea.

For early-stage applications, "traction" can mean:

• 10 users who love your product enough to tell you when you break something
• A waiting list that you've validated converts (by accepting people and checking if they use the product)
• Early revenue, even $100/month from someone who found you organically
• A pilot with a company willing to put their name on it

What traction proves is that the problem is real and that you can execute. Both matter.

What Gets You an Interview vs. Rejected#

I've talked to enough founders who applied multiple times to notice the difference between applications that get interviews and ones that don't.

Gets an interview:

• Clear, specific problem framing ("dentists spend 2 hours per day on appointment scheduling; we cut that to 10 minutes")
• Evidence you've talked to users and know them deeply
• Technical founders who have shipped real code
• Genuine insight about why the market is changing now

Gets rejected:

• "AI is eating X industry" without a specific wedge
• Market size math that assumes 1% of a massive market
• Non-technical founding teams building technical products
• No product yet, just an idea

The Interview#

If you get an interview, it's 10 minutes. That's not long enough to tell a complex story. Optimize for clarity, not completeness.

The interviewers will test your conviction, your understanding of users, and your ability to handle pressure. They'll ask hard questions designed to find the holes in your thesis. The right response to a hard question is to engage with it honestly, not to deflect or oversell.

Things we prepared for DenchClaw's interview:

• A 30-second product demo we could do verbally
• Our three most convinced users and why they use it
• A clear answer to "why is this better than HubSpot" that was honest about where we're behind and clear about where we're differentiated
• The thing we're most uncertain about and how we plan to resolve it

After the Decision#

If you don't get in, apply again. YC acceptance rates are low and the process is imperfect. Garry Tan has been explicit that the process is meant to cast a wide net and will make mistakes in both directions.

Most successful YC companies applied more than once. Use the rejection to fix the specific things that weren't convincing. Get a user to tell you in two sentences why they love your product. Use their words in your application.

Frequently Asked Questions#

Do you need a co-founder to get into YC?#

YC strongly prefers founding teams of 2-3. Solo founders do get in, but they're at a disadvantage. If you're a solo founder, be very clear about why you're the right person to build this alone and how you'll compensate for the lack of a co-founder.

How important is the idea vs. the team?#

YC invests in teams more than ideas, because ideas change. A great team with a mediocre idea will pivot to something better. A mediocre team with a great idea will execute poorly. But you still need a credible initial direction — the idea is evidence you can identify real problems.

Does prior YC experience help?#

Alumni companies get some benefit of the doubt in the screening process. But the bar for getting in is the same — it's based on what you're building, not past affiliation.

What's the one thing you'd tell a first-time applicant?#

Get 10 users who care about your product before you apply. Not 10 users who signed up because you asked them to — 10 users who would be genuinely disappointed if the product disappeared. That converts an idea into a business in YC's eyes.

Ready to try DenchClaw? Install in one command: npx denchclaw. Full setup guide →