“Data is king.” That’s what we’re told every day. But what happens when you don’t have any data? Or worse, when the data you do have doesn’t make sense?
Most people vouch for data-driven decisions. It gives us a sense of certainty, of ROI, of safety. But data can be misinterpreted, incomplete, or simply not exist when you need it most.
In startups, this gap is where founders can freeze. They wait for data to arrive, for surveys to fill, for research to validate. Meanwhile, the decision still needs to be made.
Sometimes you need to decide without having enough of it.
Why people ask for data
When people in your team push back and ask for data, they are often asking for assurance. They want to know that the effort, time, and capital they’re about to invest won’t be wasted.
That’s reasonable. You would want the same if someone asked you to spend the next three months building something.
But what if you could try it in a week?
That changes the conversation. You still don’t know whether the idea will work, but you have a way to find out without committing to the whole thing.
Gut feeling as a legit strategy
Gut gets a bad reputation. It sounds like luck, guesswork, or even recklessness.
But sometimes you’ve spent enough time with a problem to notice something you can’t yet back up with numbers. You understand the customer’s frustration. You’ve seen where the existing experience gets in the way. You have a reason to believe another approach could work.
I think that’s worth acting on.
You should still be able to explain your thinking. What have you noticed? Why do you think it matters? What could you be wrong about? Your team needs more than “I have a feeling” to work with.
And then you need to test it.
Our leap of faith
In our case, gut led us to make a radical product move. At the time, the platforms we were looking at served both borrowers and lenders. Following the same approach would have been easy to justify.
Our gut said, “Split them.”
The idea felt unorthodox. The team asked for data. We had none that directly supported the split. So instead of trying to convince them with endless slides, we cut the commitment: built a minimal version in seven days, threw a few thousand rupees into ads, and tested it live.
Buggy. Poor UI. Half-baked. But functional.
Seven years and 1 crore+ installs later, I still consider that gut move one of our best bets. Of course, that first version wasn’t responsible for everything that followed. There was a lot of work after it.
But it got us moving. And it gave us something real to learn from.
What we changed
Two calls shaped the direction.
Separate experiences
Borrowers weren’t coming to us to understand how a lending platform worked. They wanted to know whether they could get money, whether they could trust us, and how easy the process would be.
So we launched a borrower-first app, removing friction in discovery.
The lenders had different concerns. Serving both sides of the business didn’t mean we had to put both through the same experience.
Go short, not long
Instead of ₹5 lakh loans, we moved towards ₹30,000.
The thinking was that smaller amounts could make for a quicker borrowing process and allow lenders to spread their money across more loans. We still had to assess risk and do the necessary checks. A smaller loan didn’t remove those responsibilities.
But it gave us a different offering to put in front of people.
Both decisions started with our understanding of the customer. We didn’t have proof that they would work before we tried them.
Unstructured works sometimes
An unstructured approach can be useful in the chaos of an early-stage startup. You don’t always have the time or resources to follow every step in a product playbook.
We asked: if I were the customer, would this solve my problem?
That helped us get to an idea. But putting ourselves in the customer’s place could only take us so far. We still needed to see what actual customers did.
This is where I think people get stuck. They either wait for enough data to feel certain, or become so attached to their instinct that they stop questioning it.
You have to be willing to find out that your idea wasn’t as good as you thought.
What should you do when the data isn’t there?
First, be precise about the problem. Forget “everyone with a phone.” Who needs this? What are they struggling with today? Why would they try your product?
Then shrink the commitment. Find the smallest version that lets you test the part you’re unsure about. Don’t spend months building everything around an assumption you could examine much sooner.
It still needs to work well enough to give you a useful answer. If people leave because the application keeps failing, you haven’t learned whether they wanted the offering.
Decide what you’ll look for after launch. People signing up? Completing the process? Coming back? Paying? Pick the behaviour that matters for the decision you need to make.
And if it doesn’t happen, investigate before adding more features.
Data will always matter. But there will be times when it is scarce, incomplete, or simply not there.
In those moments, I’m willing to trust my gut enough to try something small.
Then I want to know if I was right.
