Laguna's AI Matchmaking: Innovation or Just Artificial Scarcity?

Laguna, a dating app that launched this week, claims it can replicate the quality of boutique matchmaking services by asking users three psychological questions and delivering a single AI-generated match per day. The pitch: all the insight of a professional matchmaker who might spend hours interviewing clients, condensed into a 90-second questionnaire and an algorithm. The company, founded by Chas McCoy and Tyler Braden, says its approach addresses what multiple industry surveys have identified as the primary driver of platform abandonment—decision paralysis from endless choice.

The Anti-Swiping Playbook Gets Another Entry

Laguna joins a crowded field of apps positioning themselves against the traditional swipe-feed model. Thursday built its brand around temporal scarcity (one day per week). Once has offered a single daily match since 2015. Hinge introduced its "Most Compatible" feature in 2019, though that sits alongside unlimited browsing rather than replacing it.

Create a free account

Unlock unlimited access and get the weekly briefing delivered to your inbox.

Register free

No spam. No password. We'll send a one-time link to confirm your email.

The distinction Laguna draws is its emphasis on AI-driven psychological profiling from minimal input. According to the company's press materials, the three questions assess personality traits, values, and what it describes as "psychological compatibility factors". No detail is provided on which psychological framework underpins this assessment, whether the questions are validated psychometric instruments, or how the algorithm weights different compatibility dimensions.

That opacity matters. Academic research on relationship formation—most notably the work published by psychologists Paul Eastwick and Eli Finkel in Psychological Science in the Public Interest—has consistently shown that pre-interaction compatibility assessments explain remarkably little variance in actual relationship success. The most robust predictor of romantic interest remains post-interaction chemistry, something no algorithm can measure before two people meet.

Human matchmakers typically conduct 60- to 90-minute intake interviews, verify employment and relationship history, and often run background checks. The suggestion that three app-based questions can replicate this is, to put it generously, optimistic.

The Business Model Challenge

Laguna's pricing structure hasn't been disclosed, though McCoy indicated in an interview with TechCrunch that the company plans to operate on a freemium model with premium features unlocked through subscription. That immediately raises the question of unit economics in a market where user acquisition costs have climbed steadily—Bumble disclosed CAC of $3.47 in its most recent quarter, up 14% year-over-year.

Single-match-per-day models face an inherent monetisation tension. By design, they limit engagement frequency, which typically correlates with willingness to pay. Hinge's "Most Compatible" exists within an app where users can—and do—engage with multiple profiles daily. Thursday compressed activity into 24-hour windows but still allowed unlimited matching during those periods.

Laguna's approach means a user who isn't interested in today's match has no reason to open the app again until tomorrow. That's a challenging proposition for retention metrics. Apps live and die by DAU/MAU ratios, and Laguna's model structurally caps daily engagement. The company will need exceptionally high match quality—defined as both mutual interest and progression to offline dates—to justify asking users to return daily despite limited in-app activity.

The AI Transparency Gap

Like most dating apps now positioning themselves as "AI-powered", Laguna has not disclosed what its matching algorithm actually does. Is it a large language model fine-tuned on relationship psychology literature? A decision tree trained on user outcome data the company doesn't yet have? A rules-based system that scores responses to the three questions against predetermined compatibility matrices?

The absence of detail is standard in the industry but increasingly problematic. The EU's Digital Services Act, which came into full effect in February 2024, requires platforms to provide meaningful transparency about automated decision-making systems. Dating apps with significant EU user bases—Laguna has not specified which markets it's targeting—will eventually face pressure to explain how their algorithms work, particularly if those systems make binary include/exclude decisions about who users see.

There's also the question of training data and bias perpetuation. If Laguna's AI is trained on existing relationship outcome data, it will inherit whatever biases exist in that dataset—typically drawn from WEIRD (Western, Educated, Industrialised, Rich, Democratic) populations that don't represent global relationship norms. The company's promotional materials make no mention of algorithmic auditing, bias testing, or the demographic composition of whatever training data underpins its matching logic.

What Comes Next

Laguna's launch timing is notable. It arrives just as enthusiasm for generative AI features in consumer apps is beginning to meet scepticism about actual utility. Dating companies have rushed to bolt on AI chatbot coaches, profile optimisation tools, and "smart" matching features, but evidence that these improve core metrics—matches, conversations, dates—remains scarce.

The app will either prove that severe choice restriction paired with minimal-input profiling generates meaningfully better outcomes than existing approaches, or it will become another data point in the growing catalogue of dating apps that mistook a feature for a business. The answer will show up in retention curves within 90 days. If users aren't returning daily, and if those who do return aren't converting matches to conversations at rates well above industry benchmarks, the model doesn't work regardless of how sophisticated the algorithm claims to be.

For operators watching this space, the relevant lesson isn't about whether to adopt single-match models. It's about the gulf between marketing positioning and product substance, and whether the industry's current AI moment is producing genuinely better matching technology or simply rebranding the same collaborative filtering algorithms that have powered dating apps since OkCupid pioneered them two decades ago.