Why customized ads are so creepy, even when they miss their target
Why customized ads are so creepy, even when they miss their target — “What is it about my data that suggests I might be a good fit for an anorexia study?” That's the question my friend Jean asked me after she saw …
Context & Ripple Effects
This Atlantic essay extends a critique that has been building since at least 2008, when The Washington Post described how Facebook ads target people where it hurts. Six years on, the argument has moved past accuracy: Jean's anorexia-study ad was a miss, yet the inference itself — drawn from her browsing without her knowledge — is what unsettles her.
The piece traveled widely through independent tech commentary on the day it published, picked up by Stowe Boyd, Marginal Utility, and others, suggesting the creepiness-of-targeting question had become a live debate among media analysts rather than a niche privacy complaint. The confirmed criticism circulating at the time was blunt: customized ads feel invasive because of how personal data is used to build the target, whatever the outcome.
First-order effects
- Users who see misfired targeted ads receive an unwanted glimpse of the inferences profiling systems have made about them — the reveal of what the platform thinks it knows lands harder than the ad itself.
- Advertisers buying behavioral audiences from platforms like Facebook pay for impressions delivered to wrongly inferred segments, absorbing waste whenever the model's guess misses.
Second-order effects
- Every visible misfire contradicts the industry pitch that personalization makes advertising more relevant, handing privacy critics a concrete, shareable anecdote instead of an abstraction.
- The discomfort shifts public attention from how well ads convert to how much data is collected to make the inference possible — pressuring platforms to defend collection practices they had treated as invisible plumbing.
Third-order effects
- If misfire stories keep accumulating, the policy argument moves from measuring ad effectiveness to asking what inferences about health, vulnerability, or identity companies should be permitted to draw at all — a framing that invites regulation rather than mere consumer annoyance.
The trend: Behavioral targeting is generating its own opposition: each visible misfire exposes the depth of behind-the-scenes profiling to ordinary users and turns ad errors into privacy arguments.