Eye-Tracking Study Exposes Dating Apps' Misguided Design Priorities

Match Group and Bumble have spent a decade perfecting interfaces that foreground physical attraction. Eye-tracking research from Australia now suggests they may have optimised their products for precisely the wrong hierarchy of information. A study from Queensland University of Technology reveals that both men and women assess economic signals alongside physical attractiveness within the first 10 seconds of viewing a dating profile—contradicting the core assumption underpinning swipe-based design.

A study from Queensland University of Technology, published in Evolution and Human Behavior, used infrared eye-tracking technology to measure what 289 heterosexual adults actually looked at when viewing dating profiles. The findings, reported by PsyPost, contradict a core assumption underpinning swipe-based design: that appearance dominates initial selection. According to the research, both men and women assessed economic signals—occupation and income—alongside physical attractiveness within the first 10 seconds of viewing a profile.

Men fixated on income information for an average of 3.49 seconds when viewing women's profiles. Women spent 4.02 seconds scrutinising income data on men's profiles. That doesn't sound like incidental browsing. Those are deliberate, sustained assessments happening in the narrow window when most dating apps assume users are deciding purely on looks.

Create a free account

Unlock unlimited access and get the weekly briefing delivered to your inbox.

Register free

No spam. No password. We'll send a one-time link to confirm your email.

What users look at versus what apps show them

The study used what researchers described as 'barebones' profiles—headshot, first name, age, occupation, and income. No curated photo grids, no prompts, no biographical essays. Just the core information set. Participants' eye movements were tracked across multiple profiles to determine fixation patterns and duration.

The key finding: economic information wasn't secondary. It was processed almost simultaneously with physical appearance, and in some cases commanded more sustained attention. Women allocated 38.5% of their total viewing time to income and occupation fields when viewing male profiles. Men spent 26.8% of their time on the same fields when viewing women.

Contrast that with how Tinder, Bumble, and Hinge present profiles. All three prioritise full-screen or near-full-screen imagery. Occupation appears as a single-line subtitle at best, often requiring users to tap through to a full bio. Income, where disclosed at all, is typically several interactions deep.

Hinge's prompt-based structure buries career information beneath dating preferences and personality markers. Bumble's career and education badges exist, but they're tertiary visual elements competing with photos, prompts, and verification icons. The design philosophy is explicit: attraction first, everything else later.

The gender complexity the narrative missed

The research also complicates the industry's operating assumption about gendered selectivity. The common narrative—repeated in earnings calls, investor decks, and product justifications—is that women are choosier, more complex in their preferences, and therefore require more sophisticated matching systems. Men, the assumption goes, are simpler: they see, they like, they swipe.

The Queensland data shows something different. Men's attention patterns varied significantly based on women's stated income. When viewing profiles of women with lower incomes, men's gaze shifted toward physical attractiveness. When income was higher, attention to economic signals increased. That's not simple binary evaluation.

Both genders are doing rapid, simultaneous evaluation of multiple signals. Current interfaces assume a linear progression—looks, then maybe bio, then possibly career details. Users appear to want all three at once.

Women's patterns were similarly complex but inverted. They spent more time examining profiles of lower-income men, though the research doesn't clarify whether this represents scepticism, compensatory assessment of other attributes, or something else. The researchers offered an evolutionary psychology interpretation—women assessing resource-provision potential—but sociocultural explanations could equally account for the behaviour.

Women in Australia, as in most Western economies, still face wage gaps and career interruptions tied to caregiving. Economic assessment may reflect practical risk evaluation rather than inherited mate-selection heuristics. What matters for product teams isn't the theoretical explanation. It's the observed behaviour: both genders are doing rapid, simultaneous evaluation of multiple signals.

The design gap and its consequences

Platforms have good reasons for their current design choices. Large images perform better in A/B tests. Swipe mechanics require simplicity. Cognitive load research suggests too much information reduces decision-making speed. But those metrics optimise for engagement, not match quality.

If users are actively seeking economic information in the first seconds of profile evaluation, and platforms require additional taps to surface it, one of two things happens: users make decisions with incomplete information, or they spend more time per profile extracting what they need. The first degrades match quality. The second degrades engagement metrics.

There's a third possibility: users adapt by inferring economic status from visual cues—clothing, settings, activities depicted in photos. That introduces noise and potential bias. A lawyer in casual dress gets misread as lower-earning. Someone who can afford professional photography signals higher status regardless of actual income. The platform loses control of the signal quality.

Operators in subscriptions-focused verticals—eHarmony, Match.com, The League—have always surfaced career and education details earlier in the profile hierarchy. They charge more and target users explicitly seeking long-term partners. The Queensland research suggests that information architecture may not be a premium feature. It may be table stakes for anyone trying to facilitate decisions based on compatibility rather than just attraction.

The study has limitations worth noting. The sample was entirely Australian, predominantly white, and exclusively heterosexual. Cross-cultural patterns may differ significantly. Income as a disclosed field is uncommon on most Western platforms, so ecological validity is uncertain—users may behave differently when income isn't explicitly stated. But the core finding—that economic signals command substantial, immediate attention—holds even accounting for those constraints.

What happens when Hinge's next product iteration lands, or when Bumble's AI-powered features roll out more broadly? If those tools continue to prioritise imagery and textual prompts while economic information remains buried, they'll be optimising for metrics that matter to platforms, not decisions that matter to users. The gap between what converts well in testing and what actually facilitates better matches has always existed. This research just makes it harder to ignore—particularly as dating apps face growing scrutiny over their relationship with mental health and user wellbeing.

D