Ofcom Studied Choice Overload. Dating Apps Are Its Cleanest Case Study

Choice overload in dating apps is a documented behavioral effect, not just a complaint. When the number of options gets large enough, people choose worse, choose slower, and feel less satisfied with whatever they eventually pick — even when the extra options were genuinely good ones. This is the same mechanism researchers have measured in supermarkets, retirement plans, and speed-dating events, and it maps onto swipe decks almost exactly.

The clearest evidence doesn't come from a dating app at all. It comes from a jam display in a grocery store, a retirement-plan enrollment dataset, and a regulator's review of digital choice architecture — three independent lines of research that converge on the same finding: more options past a certain point makes decisions worse, not better. Here's what each one actually measured, what it found, and what it does and doesn't tell you about swiping.

The Study Everyone Cites, and What It Actually Measured

Psychologists Sheena Iyengar and Mark Lepper set up a tasting booth in an upscale grocery store and rotated the display between 6 jams and 24 jams. More shoppers stopped to sample the larger display — 60% versus 40% — but the purchase rate told the opposite story. Only 3% of shoppers who saw 24 jams bought one. Among those who saw 6, the purchase rate was 30%, ten times higher.

The study, published in the Journal of Personality and Social Psychology in 2000, wasn't about dating. It was about the mechanics of choosing under abundance: more options draw more attention but produce less commitment, because evaluating them consumes the confidence needed to act on any single one.

Why This Isn't Just a Metaphor for Swiping

The jam study gets invoked casually in dating-app criticism, often loosely. The stronger case for why it applies literally comes from a separate body of research on speed dating and online choice sets, which found the same pattern with people instead of preserves: as the number of potential partners presented increased, participants became more likely to choose based on easily observable traits (like height or job title) and less likely to weigh the qualities that actually predict compatibility — while reporting lower confidence in whatever choice they made.

A swipe deck is not a metaphorical jam display. It is a live version of the same experiment, run continuously, with an incentive on the other side to keep the display as large as possible for as long as possible.

What Ofcom's Work on Choice Architecture Adds

Ofcom, the UK's communications regulator, has published research on how digital platforms structure choice — originally aimed at broadband and streaming services, but built on the same underlying decision-science literature: the more a platform's design maximizes options shown per session rather than decisions completed, the worse users report their outcomes, even when they rate the individual options as fine.

The regulatory framing matters here because it treats choice overload as a design outcome, not a personal failing. Nobody is bad at dating because they feel overwhelmed by 40 profiles in an evening. The interface was built to produce that feeling, because time-on-app is the metric a swipe platform is actually optimized for.

What the Matching Algorithm Contributes to the Problem

Choice overload alone would be bad enough with a perfect matching algorithm behind it. It's compounded by the fact that the algorithm isn't doing much matching in the predictive sense. Finkel, Eastwick, Karney, Reis, and Sprecher's review in Psychological Science in the Public Interest concluded that mathematical matching algorithms have little to no ability to predict real-world relationship success from pre-contact data alone — the compatibility signals that actually matter only become visible once two people interact.

So the deck isn't just too large. It's too large and poorly sorted, which means the extra volume isn't buying accuracy — it's just buying more decisions to make under the same overload conditions the jam study describes.

What Pew's Numbers Show About Where This Lands

Pew Research Center's 2023 survey on dating apps found that about half of U.S. adults under 30 have used a dating app, and a comparable share of users overall report the experience as at least somewhat negative — including feeling frustrated by lack of response and by the sheer volume of profiles to sort through. Pew's data doesn't isolate choice overload as a mechanism, but the pattern it reports — high usage, moderate-to-high frustration, persistent use anyway — is exactly what overload research predicts: people keep engaging with the large set because stopping feels like giving up options, even as the set itself is what's driving the dissatisfaction.

What This Research Does Not Prove

• It doesn't prove smaller is always better. The overload effect has a floor. Too few options creates its own dissatisfaction; the jam study's 6-item display wasn't chosen at random — it sat inside a range prior research had already identified as manageable.
• It doesn't prove every user is equally affected. Some people report genuine enjoyment browsing large decks, and individual tolerance for evaluating options varies. The literature describes an average effect, not a universal law.
• It doesn't prove swipe apps are deliberately harmful. The overload effect is a predictable byproduct of a design optimized for engagement time, not proof of malicious intent. The incentive is documented; the motive is a separate claim this research doesn't make.
• It doesn't prove fewer choices alone fix a match's quality. Cutting deck size addresses the overload; it doesn't address whether the underlying set was well-chosen in the first place. Both matter.

This distinction is worth sitting with, because it's easy to overcorrect the takeaway. The research is precise: past a threshold, more options degrade decision quality and confidence. It is not a blanket case against choice itself.

What Actually Reduces Choice Overload, According to the Research

Across the overload literature — from Iyengar and Lepper's original studies to later work on retirement-plan enrollment — the interventions that worked shared one property: they moved the first filter off the individual and onto a smaller, pre-curated set, chosen by criteria the person specified in advance rather than criteria they had to infer while scanning.

1. Reduce the set before the person sees it. Overload research consistently shows the fix is upstream of the decision, not a willpower fix applied during it.
2. Filter on stated criteria, not surface traits. The speed-dating research found people default to easily observable traits under overload precisely because deeper criteria take more effort to apply at volume. Filtering on the deeper criteria first removes that shortcut's appeal.
3. Cap the decision frequency, not just the set size. A large set delivered slowly overwhelms less than the same set delivered all at once. Pacing is part of the fix, not a side note.

This is the structural argument for agent-mediated matching, and it's a mechanical one rather than a marketing one: an agent that has been briefed on what you're actually looking for can do the pre-filtering the research says needs to happen before you ever see an option, rather than asking you to do it in real time against forty simultaneous choices. It's also the same logic behind the broader case against infinite swiping as a strategy and the deck-size mechanics covered in the match-quality myth piece.

How This Should Change What You Actually Do

If you're currently dating through a swipe app, the research suggests a concrete adjustment rather than an abstract one: shrink your own working set deliberately. Set a hard cap on profiles reviewed per session — the overload research suggests somewhere under a dozen before quality of judgment starts declining — rather than scrolling until you run out of patience.

If you're evaluating alternatives, the research gives you a specific question to ask any service, matchmaker, or agent-mediated platform before you commit: who is doing the filtering, and on what criteria? A service that reduces your deck size but filters on the same surface traits you'd default to anyway hasn't actually fixed the mechanism — it's just given you a smaller version of the same problem. neverswipe is built around the earlier finding, not the later one — pre-filtering on a briefing, not a photo grid.

Frequently Asked Questions

Is choice overload in dating apps a real, studied phenomenon?

Yes. It draws on decades of decision-science research, most famously Iyengar and Lepper's jam study, and has been directly extended to partner-choice contexts through speed-dating research showing the same pattern of lower confidence and shallower criteria use as options increase.

How many options is too many before overload sets in?

The research doesn't specify a universal number, but the overload effect in Iyengar and Lepper's studies appeared between 6 and 24 options, with satisfaction and follow-through both dropping sharply at the higher end. Most overload research points to somewhere in the low double digits as the point where quality of decision-making starts to degrade.

Does a matching algorithm fix choice overload by picking better options?

Not on its own. Finkel et al.'s review found matching algorithms have limited predictive power for pre-contact compatibility, so a large algorithmically-sorted deck is still a large deck — the sorting doesn't remove the overload, it just relabels it.

Does this mean fewer choices always produce a better outcome?

No — the research describes a range, not a rule that smaller is always better. Too few options creates its own dissatisfaction. The finding is that reducing an oversized set to a manageable, well-filtered one improves both decision quality and confidence.

What's the practical fix if I'm not ready to change platforms?

Cap how many profiles you review per session, and decide your filtering criteria before you start browsing rather than while you're mid-scroll. Both moves mirror the interventions that worked in the underlying research.