The YC Partner Interview: A Complete Guide to the 10 Minutes That Decide Your Batch

You spent months building. You rewrote your application five times. You recorded the founder video until you stopped sounding like a hostage. Then the email arrives: YC would like to interview you.

And the whole thing comes down to a 10-minute video call.

Ten minutes is not a lot of time. There is no room for a deck, no time for small talk, and no second chance to explain what you meant. A partner asks a question, you answer, and they decide almost immediately whether to dig deeper or move on. By the end of the day, you usually know the outcome.

This guide is a deep dive into that call: what it looks like, how it differs from every other interview you have done, what partners are actually listening for, the questions that come up again and again, and the mistakes that quietly sink strong companies. Everything here is drawn from what YC and its partners have said publicly about the interview, and from what founders who went through it have written about it.

YC in brief, and how you get to the interview

Most founders reading this already know the basics, so here is the short version.

Y Combinator is a startup accelerator. It funds early-stage companies in batches, gives them a few months of intense support from partners and alumni, and ends each batch with Demo Day, where companies present to investors. YC invests on a standard deal in every company it accepts; the current terms are published on YC's site, so check there rather than trusting a number you saw in a tweet.

Getting to the interview has three steps:

1. The written application. What you are building, who the founders are, how far along you are, and why you. Partners read it closely, and they will have it open during your interview.
2. The founder video. A short, unpolished video of the founders talking. It is a check on how you communicate, not a production.
3. The interview invite. If the application stands out, you are invited to a 10-minute video interview with partners.

Traction is not required to get this far. YC has said that a large share of each batch is still at the idea stage. What gets you invited is a clear idea, a credible team, and evidence that you move fast.

The rest of this guide is about step 3.

What the YC partner interview actually looks like

Here is what YC itself says about the format, and what founders consistently report.

• It is 10 minutes, over Zoom. Not 30, not "about 15". Ten.
• All founders should be on the call. YC prefers that each founder answers at least one question.
• Two to four partners usually interview you. They have read your application, it is open in front of them, and they take notes while you talk. Often, they are the partners you would work with if you are accepted.
• There are no slides and no formal presentation. YC is explicit about this: there is only time for two things, partners asking questions and looking at what you have built.
• A demo is sometimes requested. If you have software, have it loaded and ready to screen share. If you have hardware, have it with you or have a video.
• The decision usually comes the same day. If you are not selected, YC sends feedback by email.

The first 60 seconds

The interview almost always opens with some version of one question:

"So, what are you working on?"

It sounds like a warm-up. It is not. It is the most important answer of the interview. Partners need to understand your company in a few plain sentences, and many founders, even good ones, cannot do it. Dalton Caldwell has described how badly it goes when founders open with a rehearsed, Shark Tank style pitch instead of simply saying what they do.

A good first answer sounds like this: one sentence on what the product does, one sentence on who uses it, one sentence on how far along you are. Then stop. The partner will tell you where they want to go next.

How partners steer the conversation

After the opener, the interview becomes a fast back-and-forth. Founders who have been through it estimate roughly ten seconds per answer. Partners will interrupt a long answer and redirect you, not to be rude, but because they have ten minutes to understand your entire company.

YC partners have described the questions as basic and context dependent, not tricks. What they ask depends on where you are:

• Pre-launch or idea stage: When are you launching? How do you know people want this? What do people use today instead?
• Launched with users: Where do new users come from? How fast are you growing? Do users stick around?
• Revenue-generating: What is your growth rate? What is your churn? What do your unit economics look like?
• Regulated spaces (fintech, health, insurance, legal): What is your regulatory strategy?
• Crowded spaces: What exactly is different about you, and why do customers switch?
• Marketplaces: Which side is harder, and how do you bootstrap it?

Partners tend to go deepest wherever your answers are weakest. One founder described an interview in which the partner went straight to the core of their business model, challenged their pricing assumptions, and spent most of the remaining time there. If one part of your story is thin, expect most of the 10 minutes to be spent on it.

How it differs from every other interview you have done

Founders often prepare for the YC interview the way they prepared for a job interview or an investor pitch. Both are the wrong model.

The biggest difference is that the YC interview is a conversation, not a performance. You are not delivering a pitch. You are answering questions from smart people who already know your application and want to find out whether it holds up when they press on it. YC has said it actively discourages rehearsed speeches, because they produce robotic answers that do not respond to what was actually asked.

What partners are actually evaluating

Strip away the format and partners are trying to answer a handful of questions about you. These are drawn from what YC and its partners have written and said publicly.

1. Can you explain it clearly?

This is the strongest signal, and it comes up in almost everything YC partners say about interviews. Dalton Caldwell has said that if he cannot understand the basics of what a founder is trying to tell him, the founder is not even in the ballpark. Michael Seibel's advice is to make your description so simple it sounds almost too simple, and to avoid words like "platform". Paul Graham's version: explain it so a partner could repeat it accurately to their colleagues after you hang up.

2. Do you know your own business?

Founders who do well understand their own numbers, understand their risks, and have a crisp idea of what to work on next. That means knowing users, paying customers, revenue, growth rate and retention without hesitation, and stating them precisely. "We have 2,000 users" is weaker than "We have 2,000 sign-ups, 400 weekly actives, and 60 paying customers at $20 a month." Partners notice the difference immediately, and YC has said it may verify the numbers you give.

3. How fast do you move?

YC has said that demonstrating fast progress is the most reliable way to impress. Seibel frames progress as the ratio between what you have done and how long you have been working on it. Product and customers count far more than fundraising or partnership announcements. If you made real progress between submitting the application and the interview, say so.

4. Do people actually want this?

Partners want evidence of demand, not belief in it. Real user conversations, usage, payments, waitlists with people who pre-paid, and above all things you learned from users that surprised you. YC has said it is impressed by founders who know a lot about their users.

5. Do you have an insight?

What do you understand about this problem that most people do not? Seibel's warning is useful: a good insight is specific and plainly worded, and saying a weak insight forcefully makes you look worse, not better.

6. Are you honest?

YC has said it would rather hear a smart founder sincerely reason through an unexpected question than hear a glib answer. "I don't know yet, here's how we would find out" is a strong answer. Stretching the truth gets low scores, and misrepresenting numbers, such as passing off annual revenue as monthly, is disqualifying.

7. Is this a team that will last?

How did the founders meet? Who does what? Who writes the code? Is everyone full time? Is equity split roughly evenly? Partners are also asking themselves a quieter question: could we work productively with these people for the next few months?

8. Are you formidable?

Paul Graham's essays describe the founder qualities YC looks for: determination above all, plus flexibility, imagination and a willingness to break rules that do not matter. His word for the combination partners hope to see is "formidable", which he describes as justifiably confident. Not loud, not arrogant: confident because the facts back you up.

Want to rehearse it before the real call? We built a free, YC-style partner interview on Roundz. It is 10 minutes long, asks the questions partners ask, pushes back when an answer is vague, and gives you bullet-point feedback on what landed and what to fix. Practise your YC interview for free.

The questions that come up again and again

There is no fixed script, but the same themes appear in YC's own guidance and in founders' write-ups. Here they are, grouped, with what a strong answer looks like.

What you are building

• What does it do, concretely?
• What exists today? When do you launch?
• Can you show us?

A strong answer is plain and specific: who uses it, for what, and what state it is in right now. If it is not launched, give a date and mean it.

Users and traction

• How many users do you have? How many are paying?
• Where do new users come from?
• How fast are you growing, week over week?
• Do users stick around?
• Who was your first paying customer, and how did you get them?

A strong answer gives the number immediately, with the right qualifier (active or registered, monthly or annual, paying or free), and then stops.

Demand, if you are early

• How do you know people want this?
• What do they use today instead? And what do they really use?
• How many potential users have you talked to? What did they say?

A strong answer names real conversations and real behaviour: "We interviewed 25 teachers, 19 said scheduling costs them three hours a week, and 8 pre-paid for the first month."

Market

• How big can this get?
• How many customers could buy this, and what would each pay?

A strong answer is bottoms-up: the number of customers you could realistically sell to, times what they would pay. Partners have little patience for "the market is $7 trillion, and if we get just 1%...".

Competition and insight

• Who are your competitors, and what is wrong with them?
• Why hasn't someone already done this?
• What do you know that others don't?

A strong answer names the real alternatives, including spreadsheets, agencies and "doing nothing", and explains specifically why customers switch to you. "We have no competition" is one of the fastest ways to lose credibility.

Business model

• How do you make money?
• Why that price?

A strong answer is one simple model you are already testing. A long list of possible revenue streams sounds like you have not decided.

Team and commitment

• How did you meet? How long have you known each other?
• Who does what? Who builds the product?
• Is everyone full time? How is equity split?
• What will you do if we don't fund you?

A strong answer is short and unrehearsed. If one founder is not full time yet, say so plainly and say when that changes.

Risks and learning

• What is the biggest thing that could kill this company?
• What is the biggest mistake you have made so far?
• What has surprised you about your users?

A strong answer is candid. YC has said it is more convinced by honest discussion of difficulties than by glib dismissal.

Red flags that sink good companies

These come up repeatedly in what YC partners say about interviews that go badly.

• Jargon and buzzwords. "An AI-native orchestration layer for the knowledge ecosystem" tells a partner nothing. Say what it does for one specific user.
• Rehearsed speeches. If your answer does not respond to the question that was asked, partners notice.
• Inflated or fuzzy numbers. Leading with total downloads when the honest figure is weekly actives, or rounding monthly revenue into a "six-figure run rate", destroys trust the moment a partner does the maths.
• "There's no competition." There always is, even if it is a spreadsheet.
• Top-down market sizing. A big report number with no path to real customers.
• A potpourri business model. Five ways you might make money instead of the one you are doing.
• No evidence of talking to users. Especially at the idea stage, this is often the whole interview.
• Glib answers about risk. Dismissing an obvious risk is worse than admitting it.
• One founder doing all the talking, or founders talking over each other.
• Arguing with the partner. Treating the interview as a debate to win, rather than a conversation, is a bad sign about working together.
• Leaning on the demo to explain the company. If you need the demo to say what you do, the explanation is not clear yet.

What to do before, during and after

Before the interview

• Make progress. YC's own advice is that the best preparation is improving your company between applying and interviewing. Launch, ship, get users, grow revenue.
• Nail your one-sentence description. Seibel's test: email two sentences to a smart friend and ask them to explain your company back to you. If they cannot, rewrite it.
• Write your numbers down. Users, actives, paying customers, revenue, growth rate, retention, burn, runway. YC does not expect you to memorise them; having them in front of you is fine.
• Think through the obvious questions. Competitors, risks, why now, why you. Not to memorise answers, but so you are not hearing the question for the first time.
• Make sure every founder can answer something. Decide roughly who takes product, who takes numbers, and who takes the technical questions.
• Have the demo loaded. Logged in, on the right screen, ready to share in one click.
• Practise the format, not a script. YC warns against over-rehearsing, and they are right: memorised answers sound robotic. What helps is getting used to the pace, short answers under rapid follow-ups, so ten minutes does not feel like ambush.

During the interview

• Answer the question that was asked, then stop. Short answers give partners room to ask the next question. Long answers use up your ten minutes on one topic.
• Lead with the number, then the context. "Sixty paying customers. Most came from one community post."
• Say "I don't know" when you don't, and say how you would find out.
• Let the partners steer. If they interrupt, they are not being rude; they want to get to something they care about.
• Stay earnest. YC's guidance is to be sincere, straightforward and natural. Treat it as a conversation with smart people who might soon be on your side.

After the interview

• Expect a decision quickly, usually the same day.
• If it is a no, read the feedback carefully. YC sends rejected founders feedback focused on the most important factors in the decision.
• Reapply. A large share of each batch has applied before. Dalton Caldwell has said that showing you internalised previous feedback and made progress since is heavily weighted in your favour, and that YC keeps track of it.

Walk in ready

The YC partner interview is short, fast and unforgiving of vague answers. But it is not mysterious. Partners want to understand what you are building, see that you know your business, believe that people want it, and trust that you will keep going. Founders who do well are not the most polished speakers. They are the ones who can say what they do in a sentence, give a real number without flinching, admit what they do not know, and keep up when the questions come quickly.

The only thing most founders never get is practice under real conditions: ten minutes, rapid questions, someone pushing on the weak spots.

That is why we built a free, YC-style partner interview for the founder community. It runs for 10 minutes, starts with "What are you working on?", asks the questions partners ask, pushes back when an answer is vague, and then gives you clear, bullet-point feedback: your strengths, your red flags, and the one thing to fix before the real call.

Practise your YC partner interview for free on Roundz.

Ten minutes decide your batch. Spend ten minutes practising them first.

Roundz is an independent practice tool and is not affiliated with or endorsed by Y Combinator. Descriptions of the YC process are based on YC's public guidance at the time of writing; always check ycombinator.com for the current process and terms.