There is a phrase that appears in nearly every corner of modern business: drive engagement. It sounds healthy. Energetic, even. People are engaging with the product, engaging with the content, engaging with the brand. The graph goes up, the dashboard turns green, somebody adds a celebratory arrow to a slide deck, and the meeting ends with the comforting belief that people wanted more of whatever we made.
Sometimes they did. Sometimes engagement means exactly what it is supposed to mean: a person found something useful, interesting, entertaining or valuable enough to spend time with it and perhaps come back. But the word has become so elastic that it now covers a much less flattering category of behavior.
We call it engagement when people cannot easily look away. We call it engagement when they are irritated enough to respond. We call it engagement when the interface creates friction around leaving, when a notification manufactures urgency, when the next piece of content begins before the person has decided whether they want it, or when a system quietly turns a momentary curiosity into a forty-minute detour.
At some point, “driving engagement” stopped describing a relationship and started describing a technique. The distinction matters because one is evidence that people value what you made. The other may simply be evidence that you got very good at pushing the right buttons.
The corporate phrase is useful precisely because it removes the mechanism from the sentence. “We increased engagement by 18 percent” sounds considerably nicer than “we made the red notification badge harder to ignore.” “The new experience improved retention” sounds more sophisticated than “we made cancellation take seven screens.” “Users are spending more time in the app” avoids the uncomfortable question of whether they are staying because they are delighted, confused, angry, compelled, lost or waiting for the thing they actually wanted to load.
This is what euphemisms do well. They take a messy human behavior and sand off the parts that would make the room uncomfortable. Engagement can mean attention, loyalty, affection, outrage, compulsion, confusion, argument, anxiety or simple failure to find the exit. Once all of those are compressed into the same metric, the company no longer has to explain which one it created.
The problem is not measurement. Measuring behavior is useful. The problem begins when the measurement becomes the objective and the human reason behind it stops mattering. A person who rereads an article because it changed how they think and a person who rereads a paragraph because the page keeps jumping under an advertisement can both produce more time on page.
A customer who opens an app every morning because it helps organize their life and one who opens it because a glowing red badge has been engineered to feel unfinished can both improve daily active-user numbers. The spreadsheet sees activity. It does not automatically see consent.
That gap creates an enormous temptation. Genuine engagement is difficult. You have to make something worth choosing. You have to earn attention repeatedly. You have to accept that some people will not care, that some sessions should be short, that a customer may successfully complete a task and leave. Manipulated engagement is much easier to optimize because it gives the designer more levers.
Add another reminder. Hide the dismiss button. Delay the unsubscribe link. Insert a streak. Make the default public. Introduce artificial scarcity. Place the emotionally loaded option in the bright button and the sensible one in gray text. Turn a finite feed into an infinite one. Let the next video start before the viewer can make a fresh decision.
None of these tactics needs to be dramatic to work. That is part of their power. The modern engagement machine is rarely a villain twirling a mustache over a control panel. It is a thousand tiny decisions, each defensible in isolation. The notification is only a notification. The countdown timer is only a timer. The streak is only a little flame icon. The extra cancellation screen is only one more confirmation. The autoplay is merely convenient. The recommender is just helping people find more of what they like. Put enough of those choices together, however, and convenience can quietly become choreography.
Then the company observes the choreography it designed and reports the resulting movement as evidence of enthusiasm.
Social platforms provide the clearest version of the problem, but they are hardly alone. Retailers create urgency around carts that were never truly scarce. Subscription businesses make joining frictionless and leaving archaeological. News sites place the emotional temperature of a headline ahead of its informational value because anger travels. Games stretch reward schedules just enough to keep the next action feeling close.
Dating apps can profit from the continuation of searching even while presenting themselves as tools for ending the search. Productivity products interrupt people to announce features intended to improve their productivity. Everywhere, the behavior being measured can drift away from the benefit supposedly being delivered.
The language keeps everything respectable. We do not say that we are attempting to make someone anxious enough to reopen the app. We say we are improving re-engagement. We do not say that an argument is useful because angry people produce comments.
We say the content sparked conversation. We do not say that we intentionally made the process more annoying to abandon. We say we reduced churn. We do not say that the customer clicked because the alternative was visually buried. We say the call to action performed well.
There is a reason people increasingly feel suspicious of experiences that technically “work.” They have become fluent in being managed. They recognize the fake countdown clock. They know why the notification says somebody “liked something you shared” instead of simply telling them what happened. They notice when a free trial asks for a credit card before proving its value.
They know that “only two rooms left” may be less a piece of information than an instruction about how quickly they are supposed to feel. Consumers may not know the vocabulary of behavioral design, but they know what it feels like when a product keeps putting its thumb on the scale.
And that may be the most dangerous long-term consequence of confusing manipulation with engagement: you can improve the metric while degrading the relationship.
A trick can be tremendously effective once. It can even work hundreds of times if the user never quite notices the pattern. But every small coercion teaches the person something about the institution behind it. The cancellation maze says, “We value your subscription more than your decision.”
The manipulative notification says, “Your attention is available for us to summon.” The outrage headline says, “Your emotional state is a distribution channel.” The fake urgency says, “We are comfortable borrowing your anxiety if it improves conversion.” The message accumulates.
That is why the better question is not whether something drives engagement. The better question is what kind of engagement it creates and whether the person would still choose the behavior if the pressure were removed.
Imagine removing the badge, the streak, the countdown clock, the default opt-in, the inflammatory headline and the extra friction around leaving. Does the person still come back? Do they still read? Do they still buy? Do they still recommend the thing to somebody else?
If the answer is yes, you may have created genuine engagement. If the answer collapses the moment the behavioral scaffolding disappears, the business may not have an engagement strategy at all. It may have a collection of increasingly sophisticated tricks.
This is not an argument for making every digital experience passive or refusing to persuade anyone. Products should be designed well. Journalism should have compelling headlines. Retailers should make buying easy. Entertainment should be difficult to put down when it is genuinely entertaining. There is nothing inherently sinister about understanding human behavior. In fact, good design depends on it.
The dividing line is whether the design helps people do what they reasonably came to do or quietly recruits their reflexes against them. Good persuasion clarifies value. Manipulation obscures choice. Good engagement leaves the person feeling that their time was well spent. Manufactured engagement often leaves behind a stranger residue: Why am I still here? Why did I click that? Why was that so hard to stop?
Businesses should probably be more frightened of that feeling than they are. The dashboard cannot easily quantify it, but it is a form of debt. Each tiny betrayal borrows against trust. Eventually the customer begins treating every prompt as suspicious, every recommendation as self-serving and every “personalized” experience as another attempt to extract something. The company has trained the audience, just not in the direction it intended.
Real engagement is almost embarrassingly simple by comparison. Make something worth people's time. Help them accomplish something. Surprise them. Entertain them. Teach them. Respect the moment when they are finished. Give them a reason to return that survives the removal of the tricks.
That kind of engagement is harder to manufacture because it is not really manufactured at all. It is earned. And earned attention is less convenient on a quarterly dashboard because it cannot always be forced into the shape a growth target demands.
Which is exactly why the phrase deserves scrutiny whenever it appears in a meeting. When someone says, “This will drive engagement,” the next question should be automatic: How? What are people actually doing, and why are they doing it?
Because if the answer is that we irritated them, cornered them, confused them, scared them, baited them or made leaving unnecessarily difficult, then we should stop congratulating ourselves for engagement.
We are just tricking people.
Cultural framing is RMN's.