Test the Idea Before You Build It
Test the Idea Before You Build It
A large part of my work on digital acquisition and onboarding has been running A/B tests to check a hypothesis before the full feature is built. The habit sounds obvious. Doing it well takes more discipline than it seems.
Write the hypothesis down first
A test without a written hypothesis tends to find something, because there is always some metric that moved. Before a test starts, I want one sentence that says what we are changing, for whom, and what we expect to happen. If the team cannot agree on that sentence, the test is not ready.
Pick the metric that matters to the journey
In an acquisition journey, it is easy to measure clicks on a single screen. What matters is whether more people finish the journey and become customers. A change that lifts one step and quietly hurts the next is not an improvement. Looking at the whole funnel keeps the team honest.
Let the data overrule opinion, including yours
An idea can feel certain in a planning meeting and still fail in a test, while a small change nobody was excited about can win. That is the point of testing. The result is only useful if the team is willing to drop an idea it liked when the numbers say so.
Pair numbers with research
A/B tests tell you what happened, not why. Market and competitor research, and the qualitative insights we gather with UI/UX designers, fill in the why. The two together give a much better basis for the next idea than either alone.
Make testing part of the roadmap
Tests take time to design, run and read. If they are not planned into the roadmap with engineering, they get squeezed out by delivery deadlines. Treating validation as real work, with its own place in the plan, is what keeps it from being skipped.
Where AI fits
Building AI products in the Curious PM cohort showed me how quickly a working prototype can now be put in front of users. That makes it cheaper to test an idea early, but it does not remove the need for a clear hypothesis and a fair measurement.
Poorvi skipped presentations and built real AI products.
Poorvi Nigam was part of the August 2026 cohort at Curious PM, alongside 24 other talented participants.