The science of creepy
You’ve probably heard someone say it: “I’m convinced Facebook is listening to me.” It’s rarely meant literally. It’s as a joke or an expression of something subtler. Marketers tend to dismiss these episodes as misunderstandings. I, for one mentioned the Baader Meinhoff phenomenon on this blog before.
Consumers don’t get the technical implications of what it would mean to process everything everyone is saying all the time. They underestimate the signals they leave behind in scrolling behaviour, search history and other online metrics.
The strategic friction
Digital marketing has long relied on the belief that more data equals more relevance. But relevance is only valuable when it feels explainable.
At various points in the customer journey, sometimes early in consideration, sometimes after purchase, a personal message arrives that is too well-timed. Not wrong. Just uncannily precise. And because the consumer cannot reconstruct the causal chain, the moment is read as surveillance rather than service.
What the science actually tells us
Recent work by Petrova, Malär, Hoyer & Krohmer (2025) gives an unusually precise account of what we casually call “Facebook listening to me.” Creepiness is not a technological malfunction but a consumer-level emotional episode following a predictable sequence:
Ambiguity: the consumer cannot explain why a message appears.
Intrusive surveillance appraisal: the brain generates the hypothesis of being watched.
Uneasiness: the interoceptive discomfort of unresolved prediction error.
Reactance: an action tendency to pull away.
Decreased purchase intention: the commercial echo of an emotional event.
The study reinforces earlier findings in marketing on perceived intrusiveness, persuasion-knowledge activation, and algorithmic opacity. Across disciplines, one conclusion stands firm: humans dislike uninterpretable accuracy.
And crucially: once creepiness is triggered, very little can undo it. Most managerial interventions fail.
What this means for brands and for journey design
For brands, the key insight is deceptively simple: creepiness is a model-mismatch problem. A journey breaks when the brand’s logic outpaces the consumer’s ability to explain it.
In our customer-journey workshop, we often see the same pattern: teams map touchpoints, flows and triggers, but they forget about how clients would interpret it. When you redesign journeys through the lens of cognitive plausibility rather than technical possibilities, people get the feeling they are not doing enough to reach the right people at the right time.
Perhaps the challenge of modern marketing isn’t to make personalisation smarter but more human. Creepiness and the Baader-Meinhof effect are two sides of the same cognitive coin: the first is the mind recognising patterns it expects, the second is the mind confronted with patterns it cannot explain.
Design journeys with humans in mind
People will continue to say that Facebook listens to them. Not because the microphones are active, but because their cognitive alarms are. Between relevance and intrusion lies a narrow path that’s easier crossed than you like.
Designing journeys with that insight doesn’t make the world less complex. It just makes it more human. And strangely enough, that tends to make it more effective too.
