Frits · Mortgage calculator
Structuring a mortgage calculator around where users actually dropped off
Context
Frits is the only fully independent mortgage advisor in the Netherlands, comparing every provider rather than a select panel. I worked there as UX and Product Designer, Amsterdam, 2022 to 2023.
Problem
The mortgage calculator, the entry point to a free advice conversation, asked questions that felt unexpected, ran on a dated design, and carried friction that served no purpose: a mandatory email field, and a loading wait that did nothing for the user.
Role
I owned UX and design of the calculator within a Growth team of marketing and a data analyst. I iterated weekly on tracked behaviour, tested changes live before keeping them, and designed a further redesign that a company reorganisation paused before it could be built.
Result
Completion of the calculator, Frits' primary source of leads, rose 10 to 20 percent, tracked and confirmed in live data.
01
Context and the challenge
The calculator lets visitors work out how much they can borrow: income and financial obligations in, a comparison of Dutch mortgage lenders out, with the monthly cost at the best rate and the highest amount available. For most visitors it is the first real contact with Frits, and reaching the result page led into a free advice conversation, the company's main source of revenue.
Weekly tracking showed people leaving before that point, often at the same spots: unclear questions about debts, a mandatory email address before any number appeared, and a design that had aged.
02
Research and insight
Every change started from tracked behaviour, not intuition. Click and field data showed where people stopped, and Hotjar recordings flagged anything unusual: a field reopened repeatedly, a question answered and then abandoned. Each drop-off point became something to test directly with users and live against the existing version, before a change was kept or rolled back.
03
Design decisions
The loading screen is the clearest example. It used to be empty seconds, and many visitors did not realise Frits compared every lender rather than a limited panel. I turned that wait into proof: bank logos scanning past before the result appeared, timed to make the point without adding a word of copy.
The debts question caused the most drop-off of any single step, and the least accurate answers: people skipped it for speed, or did not see why Frits needed to know. I replaced it with a plain-language modal and a short "did you know" note explaining what a debt does to a mortgage, so filling it in became something people understood rather than something they were asked to trust.
Not every change came from data alone. Required fields and button copy were a recurring disagreement with marketing, a team that included one of the company's three co-founders: search visibility pushed toward longer, keyword-heavy text that worked against clear UX writing. We tested changes against drop-off, conversion and search ranking together rather than arguing them separately. The mandatory email went the same way: at my insistence, we tested the calculator without it. Conversion rose, genuine leads barely changed, and the field stayed gone.
04
Validation
Every change went live as a test against the version it replaced, and stayed only when drop-off and completion data supported it. Nothing in the flow changed on opinion alone, including my own.
05
Result
Completion of the calculator rose 10 to 20 percent over this period, measured with marketing and a data analyst and confirmed in live data rather than a single test. A further redesign, built around the same loading-screen logic and a clearer flow throughout, was finished but never shipped: a company reorganisation paused it before development, and Frits still runs an earlier version of the tool today.
Before · live today
After · paused redesign
Next case · EIB