The State of Dating Apps 2026: Broken by Design, Abandoned by Users

The swipe apps are not having a bad year. They are experiencing the structural consequences of a model that was never built to get you into a relationship. The data is now comprehensive enough to say so plainly.

Thirty per cent of American adults have used a dating app, according to Pew Research Center's landmark 2023 survey — a number that climbs to 53% among 18-to-29-year-olds. Online dating is not a niche. It is the default. And yet, with scale has come a reckoning the industry can no longer manage with a rebrand and a price hike.

How widespread is dating app burnout?

A Forbes Health / OnePoll survey of 1,000 Americans conducted in April 2024 found that 78% of dating app users feel fatigued with the experience — sometimes, often, or always. Women reported it at slightly higher rates (80%) than men (74%). The top cause, cited by 40% of respondents, was simple and damning: the inability to find a genuine connection with someone.

That is not a complaint about bad luck. It is a structural verdict. An app built to maximise engagement will produce exactly this outcome at scale: a large proportion of its users working harder to achieve something the app is not actually trying to deliver.

Pew Research's 2023 report adds the texture. Nearly half (48%) of online daters have experienced at least one form of unwanted behaviour on a platform — unsolicited explicit messages, persistent contact after rejection, offensive names, or physical threats. Women aged 18 to 50 were twice as likely to have experienced these as men. A striking 71% of online daters told Pew that lying to appear more desirable is "very common" on these platforms. Only 3% considered it uncommon.

The picture that emerges is not of a product people love but sometimes find frustrating. It is of an infrastructure that routinely extracts time, money and emotional energy while delivering results that roughly half its users describe as negative.

The revenue chart that explains everything

Tinder's financials are more telling than any survey. In 2025, Tinder recorded its first-ever annual revenue decline — down 5.2% to $1.8 billion, with paying subscribers falling to 8.9 million. That followed a year in which Tinder's direct revenue grew just 1% over the whole of 2024, as paying users dropped 7% year-on-year despite aggressive price increases. The company held paying-user revenue flat only by charging the shrinking base more.

Match Group — which owns Tinder, Hinge, OkCupid and Match.com — saw total paying users fall 5% across its portfolio to 14.9 million in 2024. At their December 2024 investor day, Tinder's own leadership warned of flat-to-declining direct revenue through 2026 before any expected recovery.

Bumble's story is grimmer. Its share price has collapsed more than 90% from its 2021 peak. The company posted its first quarterly revenue decline since going public in 2021, and analysts project its revenue could shrink by an average of 1.3% per year for the next three years. Even Bumble's own founder, returning to lead the company in mid-2025, told Fortune that today's dating apps "are rooted in rejection and judgement" and called them "not healthy dynamics."

In the UK, Ofcom data showed that Tinder lost 600,000 users in a single year, part of a broader pattern in which all four of the major platforms — Tinder, Bumble, Hinge and Grindr — saw declining user numbers.

Why structural misalignment is the only honest explanation

Industry observers tend to frame these declines as a product problem — bad UX, wrong age targeting, too many subscription tiers. That framing lets the business model off the hook. The more uncomfortable and accurate diagnosis is that the business model is the product problem.

A dating app paid primarily through subscriptions and engagement-boosting features (paid visibility, boosts, "who liked you" unlocks) earns its revenue from people who are still searching, not from people who found what they came for. A subscriber who meets a partner and cancels is a churned customer. The incentive is not to help you succeed; it is to keep you engaged at the exact level of hope that keeps you paying.

This is not speculation about intent. It is the observable output of the model at scale: engagement-optimised infinite feeds, artificially paywalled useful features, and a pool that mixes casual and serious intent to maximise volume. The paradox of choice research — Iyengar and Lepper's well-documented finding that more options reduce satisfaction and decision quality — predicts exactly the fatigue the surveys are measuring.

The one bright spot in Match Group's portfolio, Hinge, grew direct revenue 39% in 2024 by leaning into serious-relationship positioning. But Hinge is still owned by Match Group, still paywalls key features, and still operates on a swipe-adjacent model. Its growth is the market signalling what it wants — genuine outcomes — not proof that the existing model has fixed itself.

The shift that is actually happening

The dating market is not dying. It is sorting. Users who have been burned by the swipe model are not swearing off dating; they are looking for something structurally different. The online dating market globally was valued at $10.28 billion in 2024 and is projected to grow; the question is who captures that growth.

The structural shift is towards AI-driven compatibility matching. Not AI as a marketing adjective on top of a swipe feed, but AI that genuinely replaces the photo-first, volume-first judgement model with something more useful: a system that understands personality, values and goals, and introduces a curated few people who fit — rather than presenting a bottomless catalogue and charging you to be seen in it.

The relationship-science case for this model is robust. The similarity-attraction literature — from Byrne's foundational work (1971) through Luo and Klohnen's couple studies (2005) and Gaunt (2006) — consistently finds that shared values and personality are among the strongest correlates of relationship satisfaction. A photo predicts almost none of it. Swipe apps have spent a decade optimising for the thing that predicts the least and ignoring the things that predict the most. The data on burnout and user attrition is the predictable result.

What Lamp is doing instead

Lamp is built on the premise that the model itself must change, not just the interface.

No swipe feed. Instead, Lamp builds an AI model of your personality, values, lifestyle and goals — reading your plain-English Wishes and building a picture of who you are and what you genuinely need in a partner. It then introduces a curated few people who fit, with the reasons shown upfront. Genie, Lamp's AI assistant, helps you turn an introduction into a real conversation and a date.

No paid visibility. Lamp does not throttle your presence in the app and sell you back access to it. Core matching and Genie are free on iOS. You are introduced to someone because the AI concluded you fit — not because you outbid the queue.

No engagement optimisation. The measure of success is a relationship, not another session. That means the product only works if the matching actually works — which is exactly the accountability the swipe model has spent years avoiding.

The swipe apps are in decline because their model was always in tension with their users' actual goal. Lamp is designed around that goal from first principles. The data on what is broken is now thorough enough to treat as settled. What comes next is the product built to replace it.

Download Lamp free on the App Store and find out what dating looks like when the app is actually on your side.