How to Demo Your Product in a YC Interview

A well-executed 90-second product demo in a YC interview does more than any amount of verbal description: it makes the product real, it proves the product works, and it shows the user experience in a way that no explanation can replicate. A poorly executed demo — one that breaks, loads slowly, uses fake data, or takes more than 2 minutes — does the opposite.

This page covers exactly when to demo, how to prepare for it, how to execute it in under 90 seconds, and what to do when something goes wrong.

Not every product benefits from a live demo in a 10-minute interview. The decision depends on two questions: is the core value of your product visible in under 90 seconds, and does seeing it provide more information than hearing a description of it?

Demo when:

• Your product has a distinctive UX that is genuinely surprising or clarifying when seen
• The core value is visual or interactive — a dashboard, a workflow, a generated output
• You can show a real user doing a real task with real data in under 90 seconds
• Seeing the product immediately answers "what does this actually do?" better than words could

Do not demo when:

• Your product is infrastructure or an API with no visible UI
• The most important part of your product takes more than 2 minutes to demonstrate
• Your best demo requires a lengthy setup or explanation before the payoff
• You are building hardware and do not have the physical product on camera
• The demo environment is unreliable or requires special conditions to work correctly

The default should be: offer the demo, let partners decide. "Want me to show you the product quickly?" asked after your one-sentence description is a natural, low-pressure way to introduce the option without assuming partners want to spend time on it.

A YC interview product demo should follow this exact sequence:

Step 1 — One sentence of context (5 seconds)

Name the user and the task before showing anything. "This is what a pharmacy owner sees when she checks her expiry alerts each morning."

Step 2 — Show the starting state (10 seconds)

Open the product to the relevant screen. Do not navigate through menus — go directly to the most important view. "She opens WhatsApp and sees this."

Step 3 — Trigger the core action (20 seconds)

Show the single most important thing your product does. Not a feature tour — the core. "She sends the stock update command. Our system processes it against the supplier database and flags these three items expiring in the next 14 days."

Step 4 — Show the output (20 seconds)

Show the result of the action. If there is a generated output — a report, an alert, a recommendation — show it. Make sure real data is visible, not placeholder text. "She gets this alert — item name, quantity, expiry date, and a one-tap return request to the distributor."

Step 5 — One sentence of evidence (15 seconds)

Name a real metric the demo illustrates. Do not narrate what just happened — give the number that makes it meaningful. "That workflow saves her 45 minutes of manual cross-checking per week. Our 23 customers report an average of ₹12,000 per month in prevented expiry losses."

Total: approximately 70 seconds. That leaves 20 seconds of buffer and gives partners a natural opening for a follow-up question about what they just saw.

Demo preparation is where most founders underinvest. A demo that hesitates, loads slowly, requires login, or shows placeholder data destroys the impression a live product should create. Prepare the following before the interview:

Use a real account with real data. Never demo a new account with placeholder or test data. Partners can tell the difference between a product with real users and a product with fake data. "This is a real account from one of our pharmacies in Nashik — I've anonymized the owner's name" is more compelling than any polished demo account.

Pre-navigate to the right screen. Before the interview starts, have your product open at exactly the point where your demo begins. You should not need to log in, navigate through menus, or load a dashboard during the demo itself. The first thing partners see should be the starting state of your core workflow, already on screen.

Test your screenshare setup. Test screen sharing on the specific device you will use for the interview, in the specific video platform YC uses (typically Google Meet), the day before. Know which application to share, whether audio shares along with video (usually turn audio sharing off unless your product has audio), and that the screen resolution is legible at normal video call size.

Have a fallback. If your live product breaks mid-demo, have a short screen recording (30-60 seconds) of the same workflow ready to share. It is not as compelling as a live demo, but it is significantly better than describing a product whose demo just failed. Tell partners once: "Let me switch to a recording — I want to make sure you see the workflow clearly." Move on without dwelling on it.

Proof of existence

The most fundamental thing a live demo proves is that the product exists and works. This sounds basic, but in an era when AI products can be built rapidly and may be more demo-ready than production-ready, seeing a real product with real users and real data carries significant evidential weight. Partners discount descriptions of products they have not seen; they cannot discount a product that is working in front of them.

User experience clarity

Some products are immediately intuitive when seen and confusing when described. A WhatsApp-native pharmacy inventory product is easier to understand from a 30-second WhatsApp screen recording than from a 3-minute explanation. If your product's UX is its differentiation — if the reason users choose it over competitors is that it fits into a workflow they already have — showing it is the clearest possible proof of that claim.

Real user data as social proof

A demo with real user data — even anonymized — is a form of social proof. It proves the product has real users, that those users have entered real data, and that the product is working on that data in production. A demo account with placeholder names and round-number inventory quantities looks like a product no one is using. A demo with a real pharmacy's data, redacted appropriately, looks like a product people rely on.

Mistake 1: Demoing too early without context

Opening your screen without first naming the user and the task leaves partners watching a product they do not understand yet. Always establish who the user is and what they are trying to do before anything appears on screen.

Mistake 2: Doing a feature tour instead of a workflow demo

"Here is our dashboard, here is the analytics tab, here is the settings menu, here is our reporting feature" is a feature tour. It tells partners what exists. It does not show them a user accomplishing a goal. Demo the core workflow — one user, one task, one outcome — not every feature the product has.

Mistake 3: Demoing a product that is not production-ready

If your product requires a careful setup to avoid showing bugs, slow loading times, or incomplete features — do not demo it live. Use a screen recording of a reliable run. A live demo that breaks in front of YC partners is significantly worse than no demo at all.

Mistake 4: Using the demo as an escape from hard questions

Some founders offer to demo specifically to avoid answering a retention or traction question. Partners notice this. The demo should add information, not deflect from questions. After the demo, return immediately to the conversation.

Mistake 5: Demoing for too long

A product demo in a YC interview should not exceed 90 seconds. Every second over 90 is a second not spent on traction, competition, or team questions that partners care about more. Time your demo. Cut it if it runs long. The discipline of keeping the demo under 90 seconds is itself a signal of good judgment.