The mainstream dating apps are losing users, revenue, and goodwill at a pace that has caught executives off guard. Tinder reported an 8% subscriber drop in Q3 2025. Bumble’s revenue fell 10% year over year, and its paying user base shrank 16% in the same quarter. The decline is platform-specific rather than industry-wide. Niche apps grew alongside matchmaker firms while the mainstream swipe-and-match category contracted.
The cause is structural, and most of it traces back to the algorithm. The systems that run Tinder, Hinge, and Bumble are optimized for engagement metrics rather than for matching outcomes, and the gap between those two objectives has widened to the point where users are quitting in measurable numbers. The exodus is now visible in earnings reports, regulator filings, and survey data, and the same pattern shows up across all three major platforms.
The Engagement-First Algorithm Problem
Modern dating algorithms are built to maximize time on app, return visits, and conversion to paid features. Showing the perfect match too quickly defeats those goals. The systems instead create artificial scarcity, throttle visibility based on payment behavior, and dispense matches in spaced intervals to keep users swiping. A 2025 ScienceDirect paper modeled engagement cycles in swiping apps and found that the apps’ core mechanics map cleanly onto variable-ratio reinforcement schedules, the same pattern used in slot machines. The simple swipe right action itself, independent of match outcomes, produces small dopamine spikes followed by dropoffs that prompt more swiping.
Research summarized by industry analysts in 2025 documented the same conclusion from a different angle. 73% of dating apps fail within their first two years because their founders treat them as connection products instead of as algorithm-driven revenue machines. The survivors accepted the trade-off, which means accepting that their best operating metric is retention rather than pairing. That choice has shaped the user-facing product in ways most users have begun to notice.
User Decline by the Numbers
The exodus is now well-measured. Tinder lost 594,000 UK users between May 2023 and May 2024, per Ofcom, with Bumble down 368,000 and Hinge down 131,000. The drops continued through 2025. By Q3 of that year, Bumble’s paying users were down 16% year over year and Tinder’s subscriber count had slipped 8%. Hinge has held up better than its peers but reached its growth largely on AI-powered features rather than on its core swipe loop. The company internally acknowledges that the swipe-based portion of its product is no longer driving acquisition, which is the strongest tell yet that the format itself is the problem.
The age and gender breakdown explains the trend. 79% of Gen Z users and 80% of women report burnout from app dating. Reports of unsolicited contact correlate with women’s faster departure from the platforms. Men remain more active but report lower match rates and fewer in-person meetings per session, which produces frustration that compounds across weeks of usage.
From ELO to Engagement Scoring
Tinder’s original ranking system was an ELO score, the same model used for chess players. Profiles received a desirability rating based on right-swipe ratios, and the algorithm matched users within similar score bands. Tinder publicly retired the ELO model in 2019 in favor of an opaque scoring system that incorporates real-time activity, message quality signals, and engagement metrics. The replacement model is harder to game but easier to weaponize against the user.
Recent algorithm updates have made matters worse for inconsistent users. A 2026 Tinder change introduced harsher penalties for low-engagement weeks, which can reduce a profile’s visibility for as long as a month after a single quiet stretch. Hinge introduced “Your Turn Limits” in 2025, which automatically end conversations that go unanswered. Both changes are framed as quality improvements while conveniently increasing pressure on users to upgrade to paid tiers.
Niche and Alternative App Categories
The niche dating market hit $12.52 billion in 2026 and is projected to nearly double by 2035. Faith-specific apps, sobriety-focused apps, ethical non-monogamy apps, and lifestyle-specific apps have grown while the mainstream platforms shrank. Reported match satisfaction on niche platforms runs 20% to 30% higher than on the mainstream apps within the same demographic groups.
Alternative dating apps, specialty platforms in fields like co-parenting, fitness culture, and creative communities, and human matchmaker services have together absorbed the demand the mainstream apps have failed to retain. Users are voting with their installs.
Swipe Fatigue and the Neurochemistry of Repetition
Repeated swiping engages the same dopaminergic pathways that respond to slot machines and short-form video. Neuroscientific work from the past five years has documented that the swipe motion sustains itself even when matches are sparse. The motor pattern becomes automatic, the cognitive load drops, and the actual decision-making degrades. Users describe the late stages of a session as feeling numb rather than excited.
This has measurable downstream effects. Self-esteem scores on standardized scales are lower in heavy app users than in light users, with effect sizes that grow with session frequency. Match quality also decreases as fatigue rises. The algorithm rewards prolonged sessions, but the sessions produce worse matches, which produce more frustration, which the algorithm interprets as engagement and rewards with more sessions. The loop sustains itself even as the user reports diminishing satisfaction.
Platform-Side Experiments
The mainstream apps have started experimenting with new AI features for matching and conversation prompts. Industry analysts following dating apps have warned that these AI tools are not yet replacing the human judgment that users actually want. Hinge’s “Convo Starters” tool drove a 15% lift in matches and message exchanges in early 2025 testing. Bumble has rolled out an AI compatibility filter. The bets are real, but they are also defensive. The companies are trying to retrofit a connection product onto an engagement infrastructure that was never built for it, and the retrofits are not yet showing results in the user-base numbers.
Hinge has had the most success because it leans hardest into prompt-based discovery, which produces signals that an algorithm can use without forcing the user back into the swipe loop. Tinder and Bumble have been slower to commit to that direction. Their core revenue still depends on the swipe-driven flow, and pivoting away from it would shock the financial model that investors expect.
The Path Forward for Dating Tech
The longer-term outlook hinges on the question of how far the apps are willing to go in de-emphasizing engagement metrics. Investors have historically opposed that move, while users have started rewarding it through their feet. The next two years of app metrics will show which incentive proves stronger. The data so far suggest that the platforms are losing the users who matter most for long-term growth, the ones who came in looking for relationships and have started leaving for channels that take that goal seriously. The mainstream era of swipe-based dating is not finished, but its growth chapter has run its course. What replaces it will be built around different incentives, and the apps that catch the next wave will be the ones willing to give up engagement metrics in exchange for outcomes their users actually want.
David Prior
David Prior is the editor of Today News, responsible for the overall editorial strategy. He is an NCTJ-qualified journalist with over 20 years’ experience, and is also editor of the award-winning hyperlocal news title Altrincham Today. His LinkedIn profile is here.
