A seed-stage fintech team noticed that just over half of the people who started their signup flow never finished it. The obvious suspect was the form itself: eleven fields, an identity check, and a bank connection step. The team scheduled a sprint to shorten it.
The assumption
Before writing any code, one PM insisted on five quick interviews with people who had started but not finished signup in the last two weeks. The hypothesis going in was simple: the form is too long, people get tired, they leave.
What the interviews actually said
None of the five mentioned form length unprompted. Instead, the same hesitation showed up across four of the five conversations: people got to the bank-connection step, paused, and went to search "[company name] is it legit" in another tab before continuing — or not continuing.
The "job" these users were hiring the signup flow to do wasn't just "create an account." It was closer to "let me verify this is safe enough to link my bank account to," and the flow didn't address that job at all — it just kept asking for more information.
What changed
Instead of cutting fields, the team added three small trust signals exactly at the bank-connection step: a one-line explanation of how the connection works, a recognizable security badge, and a link to a plain-language page about data handling. The number of fields stayed the same.
The result
Drop-off at that step fell by 38% over the following three weeks, measured against the same step in the prior month. Total signup time didn't meaningfully change — people weren't in a hurry, they were unconvinced.
The takeaway
- Drop-off and friction aren't the same problem; treating every drop-off as a friction problem leads to fixing the wrong thing.
- Watching where in the flow a hesitation happens is often more useful than asking people why they left.
- Five short interviews were enough to overturn an assumption the whole team had agreed on.