The spouse-invite feature the research didn't support
A redesigned advisor-scheduling flow included an expensive in-app feature for inviting a spouse. My research ranked it near the bottom. It was built anyway, and when measured it had no positive effect on the main KPIs.
Nobody we asked would invite their spouse through the app. They'd sit together or forward the invite.
The request
Product owners and PMs, carrying a director-level goal of connecting clients with financial advisors, wanted a redesigned flow for booking a first meeting with an advisor. Part of it was an in-app way to invite a spouse or partner to that meeting, which would take substantial engineering effort.
I asked who had identified scheduling as the problem, and whether clients or advisors had ever said so.
What the evidence said
- The research behind the feature was flawed. Participants used the feature, then were asked whether they liked it and would use it, with nothing to compare it to.
- The recruiting profile (over 35, above an income threshold) missed the younger clients the CEO said they wanted. The average client was about 62.
- Interviews showed people form relationships with financial institutions much earlier than assumed, through referrals and existing accounts like a first 401(k) or a student loan servicer. Nobody went looking for an advisor at 35.
How I tested it
I asked an open-ended question first, inside interviews with the same recruiting profile, to hear what people might actually do before any survey existed. I built the survey options from those answers, alongside the options the team had assumed, like an in-app form, so the assumed options competed against ones users named.
In a prioritization survey of about 200 people, the spouse invite ranked second lowest. A short survey of people who’d recently met an advisor asked how they’d invite their spouse. No one picked an in-app invite or a form. They said they’d sit together or forward the invite.
What happened
The feature was built anyway. When it was measured, it had no positive effect on the main KPIs. I re-ran the prioritization survey with the same sample size and got the same ranking.
The prioritization survey was small, and I don’t know the sample size of the short spouse-invite survey.