Match Group announced during its earnings call that Tinder is testing an AI-driven feature named Chemistry in New Zealand and Australia to enhance user matching by analyzing interests and personalities through questions and permitted access to phone Camera Roll photos.
The feature, Chemistry, operates by engaging users with interactive questions designed to reveal personal details. With user permission, it accesses photos stored in the Camera Roll on their phones. This access allows the AI to examine images and infer hobbies, preferences, and traits from visual content. For instance, if a user’s photos include scenes of hiking or rock climbing, the system could identify these as outdoor activities. The technology then uses this data to suggest matches with individuals sharing similar interests, such as other outdoor enthusiasts. Match Group positions this as a core component of Tinder’s evolution, with CEO Spencer Rascoff stating it will serve as a “major pillar of Tinder’s upcoming 2026 product experience.” This statement underscores the company’s long-term strategy to integrate advanced AI for deeper user understanding.
Tinder’s implementation of Chemistry reflects broader industry trends in leveraging AI for personalization, though it is not unique in seeking access to private photo libraries. Last month, Meta introduced a similar capability in its platforms, where the AI requests permission to analyze photos on users’ phones that have not yet been shared publicly. This feature aims to propose AI-generated edits for those images. In both Tinder and Meta’s cases, the permissions expand data collection beyond what users have voluntarily uploaded to apps.
The testing of Chemistry has financial implications for Match Group. The company disclosed that the pilot program contributes to a $14 million negative impact on Tinder’s direct revenue in the fourth quarter. This deduction arises from resources allocated to development and deployment during the trial phase in the specified markets. Consequently, Match Group’s overall fourth-quarter revenue guidance ranges from $865 million to $875 million. This projection falls short of analysts’ consensus estimate of $884.2 million, influenced by the testing costs alongside prevailing trends in the dating sector.
Beyond Chemistry, Tinder employs AI in additional functionalities to improve user interactions. One such tool is an LLM-powered system that intervenes before message transmission. When the AI detects potentially offensive content in a draft message, it displays a prompt asking, “Are you sure?” This mechanism encourages users to reconsider their wording and fosters a more respectful environment. Another AI application assists users in curating their profiles by recommending the most appealing photos from their selections, optimizing visual presentations to attract potential matches.
Tinder has also introduced non-AI enhancements to drive subscriber growth and engagement. These include dating modes, which allow users to specify preferences for casual encounters or serious relationships. Double dates facilitate group interactions for safer initial meetings. Facial verification confirms user identities through photo analysis to reduce catfishing. Profile redesigns place bio information directly on the first photo card for immediate visibility. Prompts are now embedded within the photo carousel, prompting users to share more about themselves alongside images.
Despite these innovations, Tinder operates in a challenging environment. The app has experienced nine consecutive quarters of declining paying subscribers, culminating in the third quarter of this year. Young demographics increasingly prefer real-world social experiences over digital platforms for forming connections. In the United States, online daters are reducing expenditures as disposable income decreases amid economic pressures and recession indicators.
Match Group’s third-quarter results reflect these pressures on Tinder. The unit’s revenue decreased by 3 percent compared to the same period last year. Paying users dropped by 7 percent during that quarter. For the company as a whole, revenue reached $914.2 million, representing a 2 percent increase over the prior year but slightly below the expected $915 million. Earnings per share stood at $0.62, corresponding to a net profit of $160.8 million, which missed analyst forecasts of $0.63 per share.
