Why you keep seeing ads for things you already bought (and what it reveals)
Everyone has experienced it: you buy a mattress, and for the next three weeks the entire internet tries to sell you the exact mattress you already own. It's mildly annoying as a shopper. But if you run ads, it's a symptom worth paying attention to — because that wasted ad isn't just irritating a customer, it's revealing a gap in what the advertiser can actually see. Let's look at what's really going on, from both sides of the screen.
Why the ads keep coming after you've bought
The simple version: the ad system serving you those mattress ads doesn't know you bought the mattress. It knows you looked at it — that signal fired loud and clear when you browsed. But the purchase either happened somewhere the ad platform couldn't see, or the "you bought it" signal never made it back to the system doing the targeting. So as far as the ad platform is concerned, you're still a hot prospect who looked and didn't buy. It keeps spending money chasing a sale that already happened.
That gap between "saw you look" and "didn't see you buy" is the whole story. And it's not a rare glitch — it's one of the most common and most visible symptoms of broken measurement, playing out millions of times a day.
Frequency and ad fatigue: when repetition backfires
Beyond the already-bought problem, there's the sheer repetition. Ad frequency is how many times the same person sees the same ad, and when it climbs too high, you get ad fatigue — the point where seeing an ad again does nothing, or actively annoys, or breeds resentment toward the brand.
For advertisers, this is a double waste: you're paying to show ads that have stopped working, and you may be damaging how people feel about you in the process. The customer who's seen your ad twenty times isn't more likely to buy on the twenty-first; they're more likely to be irritated. Frequency capping — deliberately limiting how often someone sees an ad — exists precisely because past a certain point, more exposure is negative return. Yet a huge amount of ad spend goes to impressions well past that point, because the advertiser can't tell they've already saturated that person.
Poor targeting vs. data gaps: two different problems that look identical
Here's the distinction that matters for anyone running ads. When your ads land badly — served to people who already bought, shown too many times, or aimed at the wrong audience entirely — there are two very different possible causes, and they look the same from the outside.
The first is genuinely poor targeting: you chose the wrong audience settings. That's a strategy problem, and you fix it by changing your targeting. The second, and more common, is a data gap: your targeting logic is fine, but the system doesn't have the information it needs to apply it — it doesn't know this person bought, or that these two devices are the same human, or that this "new prospect" is actually an existing customer. That's not a strategy problem. It's a measurement problem, and no amount of tweaking your audience settings will fix it, because the settings were never the issue. The data feeding them was.
Confusing these two is expensive. Businesses spend months "optimizing targeting" that was never broken, when the real issue was that the system couldn't see purchases, couldn't connect devices, and couldn't tell repeat customers from strangers. You can't target well on data that can't tell who someone is.
What the annoying ad is really telling you
- The already-bought ad means a purchase signal isn't reaching your targeting. The system still thinks the sale hasn't happened.
- High frequency and fatigue mean you're paying to annoy. Past a point, more exposure is negative return.
- Bad-looking targeting is often a data gap, not a strategy error. Fix the data before you blame the settings.
- All of it traces to one question: does the system know who this person is? Purchases, devices, and repeat customers all depend on identity.
That last point is the thread running through every one of these annoyances. The reason you keep getting served a mattress you own is that the ad system can't connect "the person who looked" to "the person who bought" — because to it, they might be two different browsers, two different devices, two unconnected records. Close that gap — resolve the person so the system knows the sale happened — and the wasted impressions stop. That's exactly what Chapter is built to do: resolve each customer to one identity so purchase signals, devices, and repeat visits all connect, and your ads stop chasing sales that already closed. For the bigger picture of measuring marketing properly, see our guide to marketing measurement strategy.
Frequently asked questions
Why do I keep seeing ads for something I already bought?
What is ad fatigue?
What is ad frequency and frequency capping?
Is bad ad targeting a settings problem or a data problem?
How do advertisers stop showing ads to people who already bought?
Related reading
- Measurement StrategyHow to measure marketing properly: a plain-language frameworkDashboards everywhere, still no answer to 'is this working?' A plain-language marketing measurement framework: which metrics matter and how to connect them.
- Data & TrackingHow ads follow you around the internet (explained by people who run them)You looked at a pair of shoes once, and now they follow you everywhere. Retargeting, cookies and cross-site tracking explained plainly by people who run ads, without the spooky mythology.
- Data & TrackingIdentity resolution: how one customer stops looking like fiveOne person on a phone at lunch, a laptop that night, an email on Thursday — and your analytics counts four strangers. Identity resolution is the work of recognizing they're one, and it's the floor every other metric stands on.