Match Group (NASDAQ:MTCH) outlined changes to Tinder’s product-development process, recommendation systems and artificial-intelligence strategy during a CEO connection event focused on the dating app’s recent pace of product releases.
Tinder Chief Product Officer Mark Kantor said the company has updated “nearly every part” of the app over the past 18 months, including trust and safety, recommendations and new social connection features. He said Tinder reduced the prevalence of bots and bad actors by more than 60% and introduced products including Double Date and Events.
Engineering Output and AI Tools
Tinder Chief Technology Officer Vinay Kuruvila said the engineering team is shipping product at twice the rate it was a year ago. He said the company reduced linear handoffs among product, design and engineering teams while increasing experimentation and iteration.
Kuruvila said Tinder has also invested in its technology stack, including rearchitecting systems affected by technical debt, upgrading infrastructure for recommendations and machine-learning teams, and improving its experimentation platform. The company rewrote its chat system while keeping other parts of the ecosystem moving forward, he said, and plans to focus next on onboarding.
Artificial intelligence has become central to both product development and customer-facing features, executives said. Kantor said Tinder uses AI to reduce onboarding friction, help users build profiles and choose photos, support trust and safety tools, and improve recommendations. He said AI has shortened certain work that previously took months into weeks, or weeks into days.
As an example, Kantor said the Events product moved from an initial meeting in January to rapid prototypes within days and a public minimum viable product launch in Los Angeles in March.
Kuruvila said more than 90% of new code at Tinder is AI-generated, while emphasizing that engineers review the output. According to Kuruvila, every AI-generated code submission is reviewed by two engineers, while AI agents are also used to write tests, verify code and fix simpler bugs with human oversight.
The company said it is placing greater emphasis on hiring early-career talent with AI fluency. Kuruvila said engineering candidates are asked to complete multiple tasks using AI and explain their approach. Kantor said he is seeking curiosity, initiative and evidence of personal projects from product and design candidates.
Recommendation System Focuses on “Sparks”
Kuruvila described Tinder’s recommendation work as still being in the “early innings,” saying major releases continue to produce substantial changes in core metrics. A July launch, called Queue Unification V2, combined previously separate recommendation queues into a single system optimized for Sparks and Spark Coverage.
Previously, different queues could have distinct objectives, such as maximizing revenue, supporting new-user retention or retaining existing paying users. Under the unified approach, Kuruvila said Tinder’s machine-learning algorithms are optimized around Sparks. He said the change has driven Sparks “significantly higher” for straight women, while rollout to other segments remains ongoing.
Tinder is also developing real-time adaptive recommendations, which Kuruvila said are expected to launch in late fourth quarter. Currently, a shift in a user’s swipe behavior can take up to four hours to affect recommendations, he said. The planned system is intended to respond to behavioral changes in seconds.
Kuruvila said the company’s decision to optimize for user outcomes rather than likes or revenue represents a major shift. He added that Tinder has a “user give back” budget allowing teams to pursue changes that could improve engagement even if they reduce revenue, although the company has generally found that engagement improvements also support revenue.
Social Features and Shared Technology
Kantor said user research has repeatedly shown that singles want to bring friends into the dating experience. Tinder believes social features can reduce pressure, improve safety and make interactions more enjoyable, he said.
He said that in the U.S., more than one in five Tinder users between ages 18 and 22 has a Double Date pair. Tinder is also working on group hangouts that would support more participants, Kantor said. The company is continuing to add social elements to Events, noting that users commonly bring friends rather than attend alone.
Looking ahead, Kantor said Tinder is focused on improving the post-match experience, including using its rebuilt chat infrastructure to support conversations and meetup planning. Kuruvila said Match Group is increasingly sharing AI infrastructure, trust and safety technology and development tools across its portfolio of brands, including age assurance, verification and AI moderation capabilities.
About Match Group (NASDAQ:MTCH)
Match Group, Inc (NASDAQ: MTCH) is a leading provider of online dating products and services. The company owns and operates a diverse portfolio of consumer brands that connect singles through digital platforms. Its flagship offerings include Match.com, Tinder, Hinge, OkCupid and PlentyOfFish, which together serve users looking for long-term relationships, casual encounters and social networking opportunities.
Originating with the launch of Match.com in 1995, Match Group has grown through a combination of organic development and strategic acquisitions.
