Real estate investment · Research
User research for an investment property app
Two rounds of interviews with real investors turned the next release from an instinct into a stated, tested need.
The problem
The app was being built the way the category builds these products — sensible patterns, borrowed from the rest of the market, with no event tracking and no usability testing behind them yet. For an MVP that's a reasonable starting point. What nobody could answer was whether those patterns held for this particular audience: financially literate private investors buying income property abroad. Decisions about what to build next rested on instinct rather than on anything a client had actually said.
What we built
A research loop timed to the product, not to a calendar. We mapped the events worth tracking and verified the implementation, then ran two rounds of task-based usability tests at the two moments the product had enough surface to test — once the catalog shipped, and again once the client dashboard shipped. Respondents were real prospects and real clients, recruited through their own managers rather than a paid panel, and every session ended with questions about coming back, not just about completing the task.
What it changed
The next release answered a question clients had asked
Asked what would have made them bookmark the catalog, the most common answer was some version of "a personal account." The client dashboard became the next release because of that — a specific, testable need instead of a roadmap item chosen by instinct — and round two tested the dashboard built in response.
A label that misled people got fixed
A bell icon saved a property, under the label "Favorites." Respondents read the bell the way it behaves everywhere else — as a notification toggle — so the action people were meant to take was the one they avoided. Renaming it "Watchlist" put the label and the icon back in agreement.
Failures no analytics dashboard would have surfaced
Watching people use the app caught a tab animation that read as a new page, so tapping back threw users out of the property entirely; a follow button that silently does nothing in Safari until the app is added to the home screen; and a scroll that could quietly reload a results page and clear the filters someone had just set.
How it worksthe engineering detail
Written for the person your CTO forwards this to.
Event tracking first, as groundwork
Mapped the events worth tracking — booking, watchlist activity, and the other core actions — and checked the implementation once the development team had it in place. It ran alongside the interviews rather than standing in for them.
Round 1 — the catalog, with a contrast group
Five people who actively invest in real estate and five who don't, on the same tasks: find a property under construction using at least two filters, request a consultation, subscribe to a property's updates. The non-investors were there to show which confusions were about the interface and which were about the category.
Round 2 — the dashboard, with existing clients
Existing clients, screen-recorded in the live app: log in, save a search by budget and location, save a property to track. The real question underneath was whether people could find the properties they had already bought.
Retention questions, not just task success
Every session went past "did they finish" to "would they come back" — what would have made you bookmark this, how much is following a property worth, would you message your manager or open the app next time you had a question.
Recruiting through managers, not a panel
Managers asked their own clients who would be willing to take part, and we reached out to each person individually. Slower than a usability panel, and the only way to watch the actual audience rather than a proxy for it.
Findings handed over as a working backlog
Each finding came with the change it implied, so the output was a queue the team could act on. Some shipped immediately — the price filter now defaults to price per unit, because nobody wanted to do the per-square-meter math in their head. Others, like leading the property screen with the monthly payment, are scheduled and traceable to what respondents said.
Patterns — Task-based usability testing · Contrast group · Research-to-roadmap loop
Stack