Plan a 2-week test
Don’t plan a company. Plan the smallest experiment that proves something.
Examples for:
Don’t plan the whole company. Plan an experiment.
Your ₹10,000 isn’t for building something huge. It’s for finding out, in 2 weeks, whether your fix actually works. Small, fast, measurable.
We’ll try [what] with [how many people] for 2 weeks. It worked if [a number].
✕ TOO BIG“Build covered stalls for the whole market.”
✓ A 2-WEEK TEST“Put up one tarpaulin shade for 3 sellers. It worked if their daily waste drops from ₹200 to under ₹80.”
Budget rules
- Your total, including a ₹1,500 buffer for surprises, must be ₹10,000 or less.
- Every line needs a reason. “Snacks for the team” is not a reason.
- Before you buy, ask: can we borrow it? Reuse it? Get it second-hand?
- Keep every bill and receipt. You’ll show where the money went.
Try it
Play with a budget
Buffer for surprises (always included)₹1,500
Total₹6,600 / ₹10,000
Fits! ₹3,400 left.
Think about the finish line nowIn your application you’ll say what proof you’ll show at the end: before-and-after numbers, photos, a short video, or what customers said. Decide it now, so you collect it during the test.
For your application, you’ll need:
- Your idea in one sentence
- What you will do in 2 weeks
- A number that tells you it worked
- A budget of ₹10,000 or less (₹1,500 buffer included)
Go deeper
The Field Guide to Human-Centered Design
Read · IDEO.org · Free PDF · Dip in
How to Plan an MVP
Watch · Michael Seibel · Y Combinator · 20 min
How I harnessed the wind
Watch · William Kamkwamba · TED · 6 min
The Boy Who Harnessed the Wind
Read · William Kamkwamba · Book & film · Book
How I started a sanitary napkin revolution
Watch · Arunachalam Muruganantham · TED · 8 min
The single biggest reason why start-ups succeed
Watch · Bill Gross · TED · 7 min
Startup School library
Watch · Y Combinator · Browse