When Attention Becomes a Liability: Auditing UK Product Metrics for Trust, Not Compulsion
Photo: Oxyman, CC BY-SA 2.0, via Wikimedia Commons
Product teams in the UK have become fluent in the language of engagement. Daily active users, session length, return frequency and notification click-through rates appear in dashboards as signs of health. They are useful signals, but they are not neutral. A metric is a design instruction, and some instructions quietly teach teams to build products that are endured rather than used.
The concern is not that engagement is inherently manipulative. A well-designed service should be used when it is useful. The concern is that engagement metrics often reward intensity without asking whether that intensity reflects value. A user who checks an app repeatedly because they are anxious about a delivery, a bill or a message is engaged. That does not mean the experience is good.
The Problem With a Flattering Number
Session length is a useful example. In a research tool, a long session may indicate deep work. In a council tax portal or an NHS booking flow, it may indicate confusion. In a social product, it may indicate compulsive checking. The same number can tell three different stories, and teams tend to choose the most flattering one.
The same applies to notifications. A re-engagement campaign that brings users back after abandonment may look efficient. But if the abandonment was caused by a confusing form, the campaign is scaling a failure. Streaks, red badges and urgency prompts can increase return frequency while reducing the user's sense that they are in control. In 2026, with UK expectations around privacy, online safety and wellbeing continuing to mature, that trade-off deserves explicit review.
A Review, Not a Redesign
The practical response is not to abandon measurement. It is to add a 'time well spent' review alongside existing accessibility and privacy checks. The review should be lightweight enough to run during product development and specific enough to surface uncomfortable questions.
It can be organised around four lenses: intent, control, outcome and exit.
Four Lenses for the Audit
Intent asks why the user opened the product and whether the product supports that job. A useful prompt is: can a user complete the core task in one visit without being nudged to return? If the answer is no, the team should know whether that is because the task genuinely takes time or because the design has inserted unnecessary return loops.
Control asks whether the user can set the pace. Prompts include: are notification defaults respectful, and is the easiest path the one that respects a pause? Can a user reduce frequency without hunting through settings? Does the product treat a break as a failure of engagement or as a legitimate use pattern?
Outcome asks whether the product is measuring value rather than activity. Teams should ask: do we measure successful resolution, confidence and reduced need for support? A support product that reduces repeat contacts may be more valuable than one that increases them, even if the engagement dashboard looks quieter.
Exit asks whether the user can leave with what they need. Is there an obvious done state? Can a user export, save or record the outcome? A product that makes completion hard in order to prolong the session is not engaging; it is obstructive.
Metrics That Deserve a Second Look
Several metrics merit a second look in any UK product review. Daily active users without a task-completion measure can reward habitual opening over usefulness. Notification-driven sessions can reward interruption. Session length in support or transactional flows can reward friction. Streak retention can reward loss aversion. Re-engagement after abandonment can reward confusion. Click-through on urgency prompts can reward anxiety.
None of these metrics is inherently unethical. Context matters. A news product and a benefits calculator have different jobs. But the review should force the context into the open. If a metric only looks good when users are uncertain, distracted or unable to finish, it is measuring the wrong thing.
Putting the Review Alongside Accessibility and Privacy
The strongest place for this review is inside existing quality gates. Accessibility checks ask whether people can use the product. Privacy checks ask whether they can trust it with their data. A time-well-spent review asks whether they can use it on their own terms. Together, these checks describe a product that is not merely compliant but habitable.
Teams should sample real journeys, including failure states and notification flows. They should involve users with accessibility needs, because interruptions and urgency prompts can fall unevenly. They should document decisions about defaults, prompts and re-engagement. And they should be willing to retire a metric that encourages the wrong behaviour.
Engagement is not the enemy. Unexamined engagement is. The aim is not a quieter dashboard for its own sake, but better evidence that the product is helping people do what they came to do. In 2026, UK product teams should be judged not only by how often people return, but by whether they leave feeling that the time was well spent.
