Founder: Dr Malpani, everyone tells me to “run experiments.” But honestly, it sounds vague. What does that actually mean in the real world?
Dr Malpani: Good. You’re already ahead of most founders—because you’re skeptical.
“Run experiments” is one of those phrases investors love to throw around, like confetti at a wedding. Looks impressive. Means very little unless you define it.
Let’s make it concrete.
An experiment is simply a structured way of learning something you don’t know—without burning too much time or money.
Think of it like this:
You’re not building a company. You’re running a series of small, cheap tests to discover what customers actually want.
The Anatomy of a Good Experiment
Founder: Okay, so how do I structure one properly?
Dr Malpani: Use a simple template. No jargon. No MBA frameworks. Just clarity.
• Experiment
• Rationale
• Timeline
• Costs
• Expected Results
• Actual Results
• Lessons Learnt
Let’s walk through each. This is where most founders either learn fast—or go bankrupt slowly.
1. Experiment
Founder: This sounds obvious.
Dr Malpani: It isn’t. Most founders define experiments too vaguely.
Bad example:
“Test marketing strategy.”
Good example:
“Run ₹5,000 Facebook ads targeting working mothers to validate demand for home-cooked baby food subscriptions.”
Specificity is power. If you can’t describe your experiment clearly, you won’t learn clearly.
2. Rationale
Founder: Why does this matter so much?
Dr Malpani: Because this is where you expose your assumptions.
Every startup is a bundle of guesses:
• Customers have this problem
• They’ll pay for this solution
• This channel will reach them
Your rationale should answer:
What exactly are you trying to learn—and why does it matter?
Example:
“We believe working mothers value convenience over price when choosing baby food.”
If this assumption is wrong, your entire business collapses. Sotest it early.
3. Timeline
Founder: I tend to let experiments run until I “feel” I have enough data.
Dr Malpani: That’s founder code for procrastination.
Set a fixed timeline:
• 3 days
• 1 week
• 2 weeks
Why? Because constraints force decisions.
A long experiment is usually a lazy experiment.
Speed matters more than perfection. You’re not writing a PhD thesis—you’re trying to avoid building something nobody wants.
4. Costs
Founder: But isn’t spending necessary to get good data?
Dr Malpani: That’s what ad platforms want you to believe.
The goal is not to prove you’re right.
The goal is to learn cheaply when you’re wrong.
Set a hard budget:
• ₹1,000
• ₹5,000
• ₹10,000
Frugality is not a constraint—it’s a competitive advantage.
Bootstrapped founders learn faster because they can’t afford to be stupid for long.
5. Expected Results
Founder: This part feels like guesswork.
Dr Malpani: Exactly. And that’s the point.
Before running the experiment, write down:
• What you expect to happen
• In numbers
Example:
• 100 clicks
• 10 signups
• 2 paid customers
Why? Because without a prediction, you can’t measure learning.
Otherwise, you’ll fall into the classic trap:
“Results were interesting.”
That’s not insight. That’s storytelling.
6. Actual Results
Founder: And then we compare?
Dr Malpani: Yes. Brutally honestly.
No excuses. No “market wasn’t ready.” No blaming the algorithm.
Write what actually happened:
• 120 clicks
• 3 signups
• 0 paid customers
Data is your friend—even when it hurts your ego.
Especially then.
7. Lessons Learnt
Founder: This seems like the most important part.
Dr Malpani: It is. And also the most ignored.
Most founders jump from experiment to experiment without extracting real insight.
Ask:
• What surprised me?
• What assumption was wrong?
• What will I do differently next time?
Example:
“Users clicked but didn’t convert → Messaging unclear or trust missing.”
That’s gold.
Because now your next experiment is smarter.
The Meta-Lesson Most Founders Miss
Founder: This sounds almost too simple.
Dr Malpani: It is simple. But not easy.
Here’s the uncomfortable truth:
Most founders don’t fail because they lack ideas.
They fail because they refuse to learn.
They run experiments to confirm their beliefs—not challenge them.
That’s like going to a doctor and only accepting the diagnosis you like.
How to Maximize Learning (Not Vanity Metrics)
Let me give you a few practical rules:
• Run small experiments, not big launches
Big launches are ego-driven. Small experiments are learning-driven.
• Test one variable at a time
If you change everything, you learn nothing.
• Talk to customers after every experiment
Numbers tell you what happened. Conversations tell you why.
• Kill bad ideas quickly
Dead ideas are expensive. Kill them early.
• Double down on what works
If something works, don’t overthink—scale it carefully.
A Final Reality Check
Founder: So success is just about running enough experiments?
Dr Malpani: Not quite.
It’s about running good experiments, learning honestly, and adapting quickly.
Think of your startup as a laboratory—not a temple for your idea.
Ideas are cheap. Learning is expensive.
Your job is to make learning as cheap—and fast—as possible.
If you do this well, you won’t need luck.
You’ll have something far more reliable:
evidence.
Want to learn more about bootstrapping and creating sustainable businesses? Explore more insights and resources for entrepreneurs at www.malpaniventures.com . Let’s build businesses that put customers first!