They had new registration designs ready. I found the problem in the funnel.
The team thought registration was losing 70% of users. I rebuilt the funnel and investigated the different paths. The step they suspected was working; the first page and the route to login needed attention.
The new registration designs were already prepared.
The product had a multi-step registration. The company believed registration had roughly 70% drop-off, rising to around 80% once the following steps were included. The team thought one of the steps was putting users off and wanted to implement a new flow.
New designs were ready. Before that work went ahead, I investigated the funnel behind the decision.
Rebuild the funnel. Follow the different paths.
I found a problem in the existing funnel, built a new one, and investigated the different registration flows.
I also looked at what people did after leaving registration. Exiting that funnel did not necessarily mean they had left the product.
Many people leaving registration were trying to log in.
The first page was misleading, and the login button was hard to find. Most users who closed the registration funnel went to login and successfully logged in.
That changed the interpretation of the drop-off: many of those people needed a clearer route into an existing account.
95% of users who started registration got through it.
The company had suspected that one particular step was putting users off. The investigation showed that 95% of users who started registration passed it.
The rebuilt funnel and the flow investigation gave the team a different picture of where the friction actually was.
The same event meant different things on iOS and Android.
During the registration investigation, I also found that iOS recorded the button tap before validation, while Android recorded success after validation. Both platforms used the same event name for different user actions.
This was a separate tracking issue, not the cause of the misleading funnel. I suggested aligning what the event represented on both platforms so their data could be compared consistently.
The team redirected the work.
The company stopped spending resources on the planned registration rework. It shifted its attention to making the first experience more straightforward and improving registration tracking to avoid the same confusion in future.
Rebuilding the funnel changed the product decision before the team implemented the new designs.
Have a related problem?
Product and funnel investigations →Bring me
the problem.
An app that needs building. A migration that needs someone to own it. Numbers that don’t make sense. A product decision your team can’t resolve. Tell me what’s happening. I’ll tell you whether I’m the right person, how I’d approach it, and what the work would cost.
Email Alex ↗alex@x91tech.com