Gagan Biyani (co-founder of Udemy, Sprig, and Maven) has a different take on how to test startup ideas and it might save you years.
He calls it the Minimum Viable Test (MVT).
Most founders follow the classic advice: “Build something small. Ship it fast. Get feedback. Iterate.”
But here’s what really happens:
→ You build a mini version of your dream product.
→ Add onboarding. A dashboard. Maybe a sleek UI.
→ You launch… and realize you still don’t know if anyone truly needs it.
That’s the MVP trap.
You feel busy. You feel progress. But you’re stuck with half-working code and no clarity.
The MVT flips that script.
→ An MVP simulates a whole car.
→ An MVT? It just tests if an electric engine works better than a gas one.
It’s not about launching. It’s about learning - fast, cheap, and without writing code.
Here’s how Gagan breaks it down:
1️⃣ Start with a clear value proposition
Don’t begin by building. Start by deeply understanding your user. What are they already trying to do? What’s frustrating them?
Then ask: What’s one simple, no-brainer promise I can make to solve that?
The more obvious and specific, the better.
Example:
Good → “Healthy meals delivered in under 15 minutes.”
Bad → “We’re redefining last-mile logistics for optimized culinary efficiency.”
2️⃣ Identify your riskiest assumption
Every idea has a hidden weak point - something that, if false, kills the business.
Before you build anything, list them out.
Ask:
→ Why might this not work?
→ Is it that people won’t want it (demand)?
→ That you can’t deliver it reliably (execution)?
→ That no one will pay for it (monetization)?
→ That the market’s too small?
→ That you can’t reach users (distribution)?
Pick the one assumption that matters most and test that first.
3️⃣ Run a focused, real-world test
Now design a small experiment that tests that key risk, nothing more. Don’t ask people what they would do. Ask them to actually do it.
Skip surveys. Skip feedback forms. Instead:
→ Put up a landing page
→ Ask for payment
→ Get them to show intent with time or money
If they don’t act, that’s your signal to revise or rethink.
Gagan used the same strategy while building his own startups..
In Maven
→ He ran 5 MVTs before writing code.
→ His first test? Teaching one live course with no software.
→ Result: $150K in revenue and key insight, the course format was more important than the tech.
→ They skipped the MVP entirely and went straight to company-building.
In Sprig
→ Before building kitchens, they ran MVTs with a private chef and paper maps.
→ Validated delivery ops in two weeks.
→ Built the MVP only after proving logistics were viable.
→ Hit $1M run-rate in 6 months.
Run one test this week. Just one. You’ll learn more from that than a month of brainstorming.
MVTs won’t eliminate failure but they’ll make your failures faster, cheaper, and 10x more useful.
.
startups founders MVP startupideas buildinpublic
This post was originally shared by on Linkedin.