What SwipeStats Taught Me About Personal Data
by Kristian Elset Bø, Founder
SwipeStats started as a curiosity project. Dating apps hand over your full activity history if you request it under GDPR — every swipe, every match, every message, with timestamps. It seemed like a shame for that data to live in a JSON file nobody would ever read.
The first version was a visualization: upload your Tinder export, get a dashboard showing your match rate over time, time-of-day patterns, geographic spread, conversation length distributions. I shipped it, thousands of people uploaded, and I learned something I didn’t expect.
Market vs. mirror
My initial pitch — what I thought would be interesting — was cross-user aggregates. How does your match rate compare to the average? What does the typical user’s swipe-to-reply ratio look like? How does that shift by age or city? Market statistics.
That’s not what people came for.
What people came for was the mirror. What do my own patterns look like? They scrolled past the aggregate charts and spent most of their session on the pages that showed their own behavior over time. The insight they were after was self-reflective, not comparative.
That reframing is the single most useful thing SwipeStats taught me, and it shapes everything I build now.
The lesson, generalized
A lot of data products aim for comprehensiveness — we have the most listings, the biggest dataset, the most accurate market index — and assume users will find that compelling. Sometimes they do. More often, users engage with the slice of data that’s about them.
Market statistics are a one-time read. You skim the headline number, feel briefly validated or demoralized, and move on. Personal data is a recurring ritual. You come back to it because the story is still being written.
This is why Homi’s home-comparison views lean hard on "how does this listing fit your story" instead of "how does this listing stack up against the market." The market view has a place, but it’s the opening act, not the main event.
What I’d do differently
If I were starting SwipeStats today, I’d have led with the personal dashboard and buried the market-aggregate charts three clicks deep — not because they aren’t interesting, but because they’re not what brings people back. Data products earn retention by becoming part of how users understand themselves, not by being a well-indexed encyclopedia.
The best data product is one your user checks without being prompted. That only happens when the data is about them.