People left before they saw what they came for
Context and problem. State the business problem, the user problem and the constraints.
The onboarding asked fourteen questions before showing any result. Half of new users dropped between question six and the paywall, and support tickets said the same thing in plainer words: “I just wanted to see my chart.”
The business needed the questions for personalisation and pricing tests. The constraint was to keep that data without making people wait for it.
I owned the flow; research and pricing were shared
My role. Say what you did yourself and what the team did.
I led the design of the flow from audit to rollout: mapped the existing funnel, ran five usability sessions, built the prototypes and wrote the specs. The analyst owned the funnel data, the PM owned pricing tests, and engineering proposed the partial-profile state that made the chosen option possible.
Three ways to shorten the flow, and why I picked the second
Process and decisions. Show the options you weighed, the trade-offs and the reason for the choice.
Option B meant negotiating with engineering for a profile state the app did not have. We agreed to ship it behind a flag for 10% of traffic first, so a bad result would cost a week rather than a release.
Trial starts grew, and the paywall got a warmer audience
Result. Metrics, or qualitative findings where an NDA limits numbers.
After two weeks at 10% the new flow showed more trial starts with no drop in question completion, so it went to full traffic. Where exact figures are confidential, give relative change or describe what the team decided because of the result.
I would test the paywall copy at the same time
Reflection. What you would do differently.
We changed the flow and left the paywall untouched, which made the result easy to read but left an obvious gain for later. Next time I would plan the second experiment before the first one ships.