Dating Apps' 'Just Talking' Trap: Engagement Metrics Are Failing

Dating apps have engineered themselves into a peculiar trap: their products are now so good at facilitating conversation that users have stopped actually dating. A study published in the Journal of Couple and Relationship Therapy identifies what it calls the "just talking" phase—an extended period of digital messaging where users deliberately avoid meeting in person or defining the relationship, instead maintaining a state of prolonged ambiguity that can stretch for weeks or months. The research, based on interviews with young adults, points to two primary drivers: fear of rejection and the desire to keep other options open whilst maintaining plausible deniability.

A Product Design Failure With P&L Consequences

This isn't a user behaviour problem. It's a product design failure with direct P&L consequences. If subscribers are "just talking" indefinitely, they're not forming relationships, which means they're not churning as success stories—but they're also not getting frustrated enough to upgrade for premium features that promise better matches.

The platforms have accidentally built a holding pattern when the entire business model depends on either converting users to relationships or keeping them frustrated enough to pay for perceived advantages.

The middle state—endless low-stakes chat—monetises poorly and may be undermining the core value proposition that justifies the category's existence. The implications for product strategy are stark. Dating platforms have optimised for engagement metrics that don't distinguish between productive conversation and relationship stagnation.

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.

Engagement Theatre Versus Relationship Formation

The study's findings expose a fundamental misalignment between what dating apps measure and what they ostensibly exist to deliver. Match Group disclosed in its Q3 2024 earnings that Tinder averaged 10.8 million subscribers globally, but subscriber metrics reveal nothing about how many of those users are stuck in messaging limbo rather than progressing to dates or relationships. According to the research, participants described the "just talking" phase as a way to maintain emotional distance whilst appearing engaged—precisely the kind of surface-level activity that looks healthy in a weekly active user report but represents a failure mode for the actual service.

Users are swiping, matching, and messaging. The in-app behaviour suggests success. The relationship outcomes may tell a different story. Dating platforms have historically treated time-to-meet as a secondary metric, if they track it at all.

Bumble has made moves in this direction with features like video chat and its "opening moves" prompt system, attempting to accelerate meaningful conversation. But these features operate within the same fundamental architecture: an infinite scroll of options that, according to the study's findings, creates the very conditions for analysis paralysis and non-commitment.

The economic incentives point in opposing directions. Higher engagement typically correlates with stronger retention, which supports both advertising revenue and the likelihood of conversion to paid tiers. But if that engagement consists of users who are messaging without meeting—effectively treating the platform as a pen pal service rather than a dating facilitator—the long-term brand promise erodes.

The Abundance Problem Platforms Created

The "desire to keep options open" that the study identifies as a driver of the "just talking" phase is a direct consequence of product design choices dating apps have made over the past decade. Swipe interfaces and match queues create a visible pipeline of alternative options. Every conversation happens in the context of dozens of other potential conversations.

When users can always swipe for another match, the opportunity cost of committing to a meeting—let alone a relationship—increases. The "just talking" phase becomes a rational response to structural incentives the platforms themselves created.

This isn't accidental. It's core to the gamification strategy that made apps like Tinder scalable. But the researchers' findings suggest that abundance has crossed a threshold from feature to bug.

Hinge's positioning as "designed to be deleted" represents an attempt to solve this misalignment, though the company's integration into the Match Group portfolio in 2019 and subsequent growth focus raises questions about how seriously that promise can be maintained when it directly conflicts with engagement maximisation. The platform does include prompts encouraging users to move off-app, but these nudges operate against the grain of the core experience.

Grindr offers an interesting counter-case. The app has historically facilitated faster progression from match to meeting, partly driven by its focus on hookups rather than relationships, but also by design choices that emphasise proximity and availability. The platform's Q3 2024 results showed average revenue per paying user of $64.60, up 8% year-over-year, suggesting that optimising for actual real-world meetings rather than extended chat can support premium monetisation.

The Authenticity Paradox

The study references separate research showing that 74% of users on platforms like Tinder claim to value authenticity and meaningful connections. But the "just talking" behaviour pattern suggests a disconnect between stated preferences and revealed preferences. Users say they want authenticity whilst simultaneously adopting communication strategies designed to maintain emotional distance and optionality.

This creates a product design challenge with no obvious solution. If platforms introduce features that force faster progression—say, matches that expire unless users schedule a meeting, or chat functions that become unavailable after a certain number of messages—they risk backlash from users who feel pressured. But if they maintain the status quo, they're facilitating a user behaviour that undermines the service's purpose.

Bumble's time-limited match system attempted a version of this urgency design, requiring women to message within 24 hours or lose the match. The company has since relaxed these constraints following user feedback, extending the window and adding "extend" features. The market appears to have spoken: artificial urgency feels coercive rather than helpful.

Video features represent another attempted intervention. Bumble, Tinder, and Hinge all introduced video calling during the pandemic, partly as a COVID-19 adaptation but also as a potential bridge to in-person meetings. Adoption has been mixed. Video requires more commitment than text and raises the stakes closer to an actual date, which is precisely what users in the "just talking" phase appear to be avoiding.

Product Strategy at a Crossroads

The regulatory environment isn't offering any pressure to solve this. The UK Online Safety Act and EU Digital Services Act focus on harms prevention and content moderation, not relationship efficacy. There's no compliance requirement for dating apps to actually facilitate dates, only to keep users safe when they do.

That leaves the competitive dynamics as the main forcing function. If one platform cracks the code on moving users from "just talking" to actual relationships without feeling coercive, it could capture the segment of users who are genuinely frustrated with messaging stagnation. But it's equally possible that the "just talking" phase represents a stable equilibrium—uncomfortable for both users and platforms, but not uncomfortable enough to force change.

The study's findings suggest the current product design consensus has run its course. The question is whether dating apps will redesign around relationship formation as the primary metric, or whether they'll continue optimising for engagement and accept that "just talking" is the new normal. The latter path is easier. The former might actually require admitting that swipe-based abundance was a feature that became a liability, and that success metrics need to be rebuilt from first principles around outcomes rather than activity.