Phase 1
Frame the problem
Map the user journey, business lever, constraints, and available evidence before choosing a feature.
In practice: Led FTC pricing-transparency work and built a dynamic fee module across OYO US and G6.
How I make product decisions.
I align the customer behavior, business lever, system constraint, and measure of success before committing to a solution.
Phase 2
Choose the lever
Turn the goal into a decision: the metric to move, the mechanism, the tradeoffs, and the smallest useful test.
In practice: Used engagement and funnel analysis to ship dashboard, migration, and retention improvements at SplashLearn.
Phase 3
Ship the whole change
Stay with the work through UX, instrumentation, QA, launch readiness, and the operating changes around it.
In practice: Carried D2C improvements through UX, analytics, QA, launch readiness, and operating setup at Honasa.
Phase 4
Review the evidence
Agree on success and guardrails before launch, then use outcomes and failure modes to decide whether to scale, revise, or stop.
In practice: Built evaluation into the NL-to-SQL product so answer quality could be measured against known-good queries.
Working principles.
Write down the decision
Make the customer behavior, business lever, constraint, tradeoff, and measure of success explicit.
Work at implementation depth
Use prototypes, platform-native surfaces, logs, and data to keep product judgment honest.
Separate signal from attribution
Use the evidence to distinguish what changed, what the product likely influenced, and what should not be claimed.