Revenue growth shows the AI spend is paying off

Grindr is betting that artificial intelligence can do more than just improve the dating app experience.

It believes AI can simultaneously create a new premium subscription business while also dramatically changing how quickly its engineers build software.

“Our strategy has always been to use AI everywhere we can,” said CEO George Arison in an exclusive interview with CNBC.

The LGBTQ-focused dating platform on Thursday reported second-quarter revenue of $138 million, up 33% from a year ago. The company also raised its 2026 guidance and now expects full-year revenue of approximately $540 million, up from $535 million, and adjusted EBITDA of approximately $232 million, up from $227 million.

Arison said the company’s results are early evidence that his AI strategy is starting to pay off.

Rather than focusing solely on consumer-facing features like its new premium AI companion called “Edge,” Arison is also deploying AI across its business to reduce the cost and time required to build new software products.

Grindr estimated that total engineering output increased roughly 2.5 times between July 2025 and April 2026, the earnings presentation showed, despite maintaining roughly the same-sized engineering team.

“Before GenAI, producing that much technical output would have required roughly 200 additional engineers and approximately $60 million in annual cost,” the earnings presentation said.

The gains have allowed Grindr to release products faster without meaningfully increasing engineering headcount, a trend investors are increasingly watching across the software industry as companies look for tangible financial returns from generative AI investments.

While spending on large language models continues to rise, he said the return on investment has made the decision an easy one.

“We were on track to spend $6 million on tokens this year,” said Arison. “I have zero qualms about that because the amount of productivity increase that I’m getting from that is orders of magnitude more, like 10-times more than the money that I’m spending on tokens. So it’s a total no-brainer.”

Arison said Grindr has not eliminated any jobs because of AI adoption, but has aggressively expanded the use of coding assistants and software development tools from Cursor, Anthropic’s Claude and Devin.

“You can run these businesses in a far leaner way than people think you can, but you need much better management,” said Arison.

Grindr is also using AI to create an entirely new premium subscription tier.

The company is testing pricing for its AI-powered “Edge” companion in select markets, including New York, where access costs as much as $350 per month.

Initially, management expected “Edge” to primarily attract upgrades from its highest-paying subscribers. Instead, Grindr said, the customer mix has been broader.

“Our assumption was that only people who are ‘Unlimited’ subscribers would move up to ‘Edge,'” Arison said. “The data so far does not support that. There are people who are not ‘Unlimited’ subscribers who were also moving to ‘Edge’, including people who don’t subscribe at all.”

While Grindr did not disclose how many users have subscribed to “Edge” or where pricing will ultimately land, Arison said the company is “happy with the test results so far.”

The early adoption rate could help validate one of Wall Street’s biggest questions surrounding generative AI: whether consumers are willing to pay substantially more for AI-powered experiences. Morgan Stanley in July upgraded the stock to overweight from equal weight and raised its price target from $15 to $18, citing the “ultra-premium” subscription tier.

The company’s financial results suggest those investments are beginning to translate into business performance.

Arison said subscriber retention has remained stronger than expected despite price increases across Grindr’s premium membership tiers.

“Churn is actually lower than we had expected it to be, which has led to more organic growth in the first half of the year,” said Arison.

Paying user count grew in the second quarter to 1.4 million, up 16 over last year, with average revenue per paying user reaching $25.51, a year-over-year increase of 12%.