Product teams rarely have the luxury of solving everything. There are competing customer needs, business goals, technical constraints, stakeholder opinions and limited time.
I use evidence to reduce uncertainty, then make the trade-offs explicit, so teams can decide what to pursue, what to challenge and what to leave out.
This gives me a clear separation from Investigation.
Investigation reduces uncertainty.
Decision-making converts that understanding into choices.
Start with the decision, not the solution
A lot of teams ask: “Which design should we choose?” when they haven't articulated: “What are we actually deciding?”. Define what needs to be decided. Before comparing solutions, I try to make the decision itself explicit.
Are we deciding which problem to solve?
Which customer to prioritise?
What behaviour to change?
Where to invest?
What not to build?
Make the trade-offs visible
Every decision has a trade-off. Name it.
For CashKaro:
Marketing wanted more promotional visibility.
Product wanted activation.
Business wanted transactions.
Support wanted education.
Customers wanted clarity.
There was no design solution that could maximise all five simultaneously. So I made the trade-off explicit: For the first fold, confidence had to win over information density.
Evidence informs judgement; it doesn't replace it
Data doesn't make the decision for you.
Numbers can tell me what is happening. Research can help explain why. Neither tells me automatically what the organisation should do. That still requires judgement. I use evidence to make the reasoning visible, challenge assumptions and understand the consequences of each option.
Choose what not to do
Prioritisation is subtraction.
Every product has more opportunities than the team has capacity to pursue. Good prioritisation isn't creating a longer backlog. It's being able to say: Not now. Not this. Not for these users. And being able to explain why.
When evidence disagrees with the room
Sometimes research contradicts stakeholder intuition. Sometimes the data contradicts the research.
Sometimes two teams are looking at different parts of the same problem. I don't treat disagreement as something to win. I use it to understand what assumption is different.
Make decisions reversible where possible
Match the decision to the level of certainty it deserves.
I don't believe every decision needs months of research.
Some decisions need stronger evidence because they're expensive or difficult to reverse.
Others can be tested, learned from and changed.
The goal isn't certainty.
It's enough confidence to make the next responsible decision.
Create commitment, not consensus
Teams don't need everyone to agree. They need to know why a decision was made.
Alignment doesn't mean everyone gets what they want. It means people understand:
what we decided,
why we decided it,
what evidence informed it,
what we're giving up,
and what we'll learn next.
Once the decision is made, the team should be able to move together even when the decision wasn't everyone's first choice.