Online Dating's Real Success: It's Not About Finding Soulmates

Approximately 360 million people used dating apps globally in 2024, according to Business of Apps. The dating services market generates over $8 billion annually. Yet the question that matters most to users - does this actually work? - has received surprisingly little definitive scientific attention. The answer, based on more than a decade of accumulated research, is: it depends on what 'works' means, and the evidence is more encouraging than the industry's critics suggest but less transformative than its marketing claims.

The most rigorous longitudinal data available comes from a 2013 study by John Cacioppo and colleagues, published in the Proceedings of the National Academy of Sciences. Analysing a nationally representative sample of 19,131 Americans who married between 2005 and 2012, the researchers found that couples who met online reported slightly higher marital satisfaction and slightly lower rates of marital breakup than couples who met offline. The effect sizes were small but statistically significant. Critically, couples who met through dating websites (as distinct from social media or other online contexts) showed the most favourable outcomes.

More recent research has complicated this picture without overturning it. A 2020 study by Josue Ortega and Philipp Hergovich, published in the Journal of Political Economy, modelled the effects of online dating on society and found that it increased interracial marriage rates and produced stronger marriages (measured by breakup rates) compared with offline meeting contexts. The mechanism proposed was that online dating connects people who would otherwise never have met, expanding the pool of potential partners beyond the limitations of social networks, workplaces, and geographic proximity.

The DII Take

Online dating works in the sense that it successfully connects people who form lasting relationships. What online dating does not do, as the compatibility algorithm literature makes clear, is identify uniquely compatible partners through algorithmic matching.

It works primarily as an introduction mechanism - expanding the pool of potential partners beyond what offline social networks can provide. This is a genuinely valuable function, and the dating industry would be better served by marketing it honestly than by overpromising algorithmic compatibility. The evidence supports the claim that dating apps help people meet. It does not support the claim that dating apps find people their soulmate.

What the Research Shows About Relationship Outcomes

The academic evidence on online dating outcomes clusters around several key findings.

Meeting online has become the dominant pathway to romantic partnership. A widely cited 2019 study by Michael Rosenfeld, Reuben Thomas, and Sonia Hausen, published in the Proceedings of the National Academy of Sciences, found that by 2017, approximately 39% of heterosexual couples met online - surpassing all other meeting contexts including through friends, at bars, at work, and through family. For same-sex couples, the proportion meeting online exceeded 65%. This structural shift means that online dating is no longer an alternative to 'normal' dating. It is normal dating.

Relationship quality comparisons between online and offline meeting contexts show mixed but generally positive results. Beyond the Cacioppo et al. findings, a 2020 meta-analysis by Stephanie Tong and colleagues examined 30 studies comparing relationship outcomes for couples who met online versus offline. The overall effect sizes were small and inconsistent across studies, but the meta-analysis found no evidence that online-initiated relationships were inferior in quality to offline-initiated ones. Some studies found modest advantages for online-initiated relationships on specific measures (communication satisfaction, relationship stability), while others found no difference.

The speed-dating literature provides a useful comparison. Research by Eli Finkel and Paul Eastwick using speed-dating paradigms has demonstrated that initial attraction is highly unpredictable from profile information alone but becomes measurable within seconds of face-to-face interaction. This suggests that the value of online dating lies not in the profile-matching phase but in the introduction phase - getting two people into the same room (or onto the same video call) who would not otherwise have met.

User satisfaction with the process, as distinct from outcomes, remains problematic. Pew Research Centre surveys have consistently found that while many Americans have used dating apps, satisfaction is mixed. A 2023 Pew survey found that while 53% of Americans who had used dating sites or apps said their overall experience was positive, 46% said it was negative. Among women, negative experiences were more commonly reported, driven by unwanted contact, harassment, and the emotional toll of the process.

The Meeting Context Effect

One of the most important but least discussed findings in the online dating outcomes literature is that how couples meet shapes the early dynamics of their relationship, even if it does not determine long-term success.

Couples who meet online have, by definition, chosen each other based on curated self-presentations before experiencing each other in person. This creates what researchers call an 'expectation gap': the difference between the person anticipated from the profile and the person experienced in reality. Research by Sharabi and Caughlin (2017) found that expectation violations during first dates negatively predicted romantic interest, even when the violation was objectively neutral or positive. A person who turns out to be more attractive than their photos suggested still violates expectations, and the violation itself introduces cognitive dissonance into the interaction.

By contrast, couples who meet through friends, at events, or through shared activities encounter each other without pre-formed expectations. The 'getting to know you' process unfolds organically, with expectations calibrated by direct experience rather than profile curation. This difference may partly explain why some studies find a slight advantage for offline-initiated relationships in early-stage satisfaction, even if long-term outcomes converge.

The implication for platform design is significant. Features that reduce the expectation gap - video profiles, voice introductions, real-time interaction before the first date - should improve first-date satisfaction by aligning expectations with reality. Hinge's voice note feature and the growing adoption of video-first dating formats both move in this direction.

Gender Differences in Online Dating Outcomes

The research reveals consistent gender differences in online dating experience that platforms must account for.

Women report more negative experiences than men across virtually every survey. Pew Research Centre data shows that women are more likely to report receiving unwanted sexual messages, being contacted after saying they were not interested, and feeling harassed on dating platforms. These negative experiences reduce women's willingness to engage with platforms, contributing to the gender imbalance that plagues many dating apps (most platforms have more male than female active users).

However, women who do form relationships through dating apps report outcomes at least as positive as women who meet partners offline. The Cacioppo et al. study found no gender difference in the marital satisfaction advantage of online-initiated relationships. This suggests that the problem is not with online dating itself but with the quality of the in-app experience for women, particularly the filtering and safety mechanisms that determine which interactions reach them.

Men face a different challenge: lower match rates and higher rejection rates, which research links to diminished self-esteem and dating burnout over time. The asymmetry of dating app dynamics - where a minority of male profiles receive the majority of female attention - creates a frustration pattern for many male users that is distinct from but equally corrosive to the harassment pattern that many women experience.

What 'Working' Should Mean for the Industry

The research suggests that the dating industry should redefine its success metrics.

Currently, most platforms measure success in terms of engagement (time in app, swipes, messages) and revenue (subscriber conversion, ARPU). These metrics do not correlate with the outcomes users actually want: meaningful connections, enjoyable dates, and lasting relationships. Hinge's 'We Met' feature, which tracks whether matches lead to real-world dates, represents the most important metric innovation in the industry because it measures an outcome rather than an activity.

A more comprehensive outcome framework would track: match-to-date conversion rate, date satisfaction, relationship formation rate, and relationship durability. Platforms that optimise for these metrics rather than engagement metrics will produce better outcomes and, ultimately, better commercial results, because user satisfaction drives referrals and re-subscriptions more effectively than compulsive engagement.

The academic evidence, taken as a whole, supports a cautiously optimistic verdict. Online dating works as an introduction mechanism. It has expanded the pool of potential partners, increased the diversity of partnerships formed, and produced relationships at least as stable as those formed through traditional channels. What it has not done is crack the code of compatibility prediction or eliminate the challenges of the dating process itself. The platforms that acknowledge these realities - marketing themselves honestly as sophisticated introduction services rather than compatibility engines - will build the most sustainable businesses.

What the next decade of research needs, and what the industry should fund, is longitudinal outcome tracking at scale. The Cacioppo et al. study remains the gold standard, but it was conducted using data from 2005-2012, before Tinder existed and when online dating meant desktop-based matchmaking sites. The dating landscape has transformed fundamentally since then. A contemporary equivalent - tracking relationship outcomes for tens of thousands of app-initiated connections over multiple years - would provide the evidence base the industry needs to genuinely improve its products. Hinge's 'We Met' feature generates the seed data for this kind of research. The company, or the broader industry, that commits to systematic outcome tracking and academic partnership will produce the first genuinely evidence-based matching system in dating history.

The Nuances That Matter for Operators

Several research findings carry direct product design implications.

Time-to-meeting predicts satisfaction. Studies consistently show that prolonged messaging before meeting correlates with lower satisfaction when the meeting eventually occurs. The explanation is straightforward: users build idealised expectations during extended text exchanges that reality rarely matches. Platforms that facilitate faster transitions from match to meeting - through voice notes, video features, or event-based meeting contexts - align with the research evidence.

Platform design affects behaviour and outcomes. Research by Toma and colleagues has shown that the design of dating profiles systematically encourages mild deception (users exaggerate height, understate weight, use flattering but unrepresentative photos). Platforms that reduce information asymmetry - through video verification, real-time features, and post-match feedback mechanisms - produce more honest interactions and better outcomes.

Relationship formation through apps follows similar psychological pathways to offline relationship formation, but on a compressed timeline. Self-disclosure, responsiveness, and escalating intimacy follow the same patterns identified in decades of relationship psychology research. The medium has changed, but the underlying human dynamics have not. This means that relationship science insights about communication, attachment, and compatibility remain directly applicable to app-based dating.

The overall verdict from a decade of research is cautiously optimistic. The competitive advantage in dating is not in predicting compatibility. It is in facilitating the highest-quality introductions and supporting the transition from digital match to real-world relationship.

The Platform Comparison Question

Users frequently ask which dating platform produces the best relationships, and the honest answer is that no rigorous comparative research exists. The Cacioppo et al. study predated the current app landscape. No academic study has compared relationship outcomes across Tinder, Hinge, Bumble, and other platforms using matched samples and longitudinal tracking. This is a remarkable gap given the industry's size and the intensity of competitive marketing claims.

What indirect evidence exists suggests that platform design influences early-stage dynamics rather than long-term outcomes. Platforms that encourage deliberate engagement (Hinge, Coffee Meets Bagel) may produce fewer but higher-quality initial connections. Platforms that encourage volume (Tinder) may produce more connections of more variable quality. Whether these early-stage differences translate into differences in relationship longevity remains empirically unknown.

The absence of comparative outcome data represents both a research opportunity and a commercial one. The first platform to publish rigorous outcome data - demonstrating that its users form more lasting relationships than competitors' users - would possess the most powerful marketing asset in the industry. The reluctance to collect and publish such data suggests that platforms may fear the results as much as they desire them.

More articles by DII Editorial