How to Validate a Startup Idea at Purdue

You validate a startup idea by getting evidence from real potential customers before you build — not by asking friends if they like it. At Purdue you are surrounded by test users, and a useful first round of validation costs about two weeks and zero dollars.

Key takeaways

• Talk to at least 10 people who actually have the problem, and ask about their behavior, not your idea.
• Build the smallest possible test — a landing page, a form, a manual version of the service — before any real code.
• Look for evidence that costs the customer something: time, a sign-up, a preorder. Compliments are not evidence.
• Decide in advance what result would make you stop. If nothing would, you're not validating, you're collecting encouragement.

Start with the problem, not the pitch

Write down who has the problem, when it happens, and what they do about it today. If you can't name a specific kind of person at Purdue or beyond who hits this problem in a normal week, that's your first finding.

Then find 10 of those people and ask about the last time the problem came up. Don't describe your idea first — people are polite, and a pitch turns the conversation into feedback on your delivery instead of facts about their life. Questions that work:

• "When did this last happen to you? Walk me through it."
• "What did you try? What did that cost you in time or money?"
• "Have you looked for a fix before? What did you find?"

If most people can't recall a recent, specific instance, the problem may be real but not painful enough to build on.

Run the cheapest possible test

Once interviews suggest the problem is real, test whether people will act, not just agree.

Pick the one that matches your idea and set a threshold before you start — for example, "20 sign-ups from 100 visitors" or "3 of 10 interviewees use the manual version twice." Hitting the threshold means build the next smallest version. Missing it means change the idea or the audience, and that's a successful test too.

Use the campus around you

Purdue makes the expensive part of validation — finding people to talk to — cheap. BuildPurdue events and Nightshift put you in a room with students who will give you honest reactions and introductions. Classes, clubs, and dining courts are full of whatever niche your idea serves.

• Find BuildPurdue events
• Work alongside other builders at Nightshift
• How to start a startup at Purdue
• How to find a cofounder at Purdue

What to avoid

• Don't survey your friends. They know you, so they'll be kind, and kindness is bad data.
• Don't build for three months "to have something to show." The point of validation is to earn the right to build.
• Don't count "I'd totally use that" as a yes. Count sign-ups, usage, and money.
• Don't run one test, get a mixed result, and quit or commit. Mixed results usually mean the audience is wrong, not the idea.

FAQ

How many interviews are enough?

Ten good ones beat fifty rushed ones. Stop when new interviews stop surprising you — when you can predict the answers, you've learned what this audience has to teach.

What if someone steals my idea?

Ideas are cheap and execution is the moat. The risk of building the wrong thing in secret is far higher than the risk of someone hearing your idea and out-executing you on it.

Do I need to validate if I'm building it anyway?

Yes — validation changes what you build. Even a project you'd do for fun ships better when the first version targets something real people demonstrated they want.

Wrap up

Validation is two weeks of conversations and one cheap test, run honestly. If you want structure and people who will hold you to your threshold, apply to the cohort or bring your idea to the next BuildPurdue event.