Here’s a quietly terrifying number. In 2016, researchers Bartosz Wojdynski and Nathaniel Evans sat 246 people down in front of an article that looked like any other piece of journalism — a clean layout, a legitimate news brand’s styling, a headline that promised useful information. At the top, in small gray type, it said “Sponsored.” At the bottom, another disclosure. The experiment was not about whether people liked the article. It was about whether they knew they were reading an ad.
Only 17.8% did.
🧠 The Article That Wasn’t
The Wojdynski & Evans experiment became a landmark in advertising psychology for a simple reason: it quantified a suspicion that had been floating around the industry for years. People don’t recognize native advertising as advertising. Not because they’re stupid — because their brains never got the memo to classify it that way.
Native ads borrow the visual language of editorial content: same font, same layout, same narrative tone. When the brain encounters something that looks like a news article, it fires up the same heuristic it uses for everything else in that category: this is information, process it for understanding. The “persuasion detection” system — the cognitive machinery that says “wait, someone is trying to sell me something” — never activates.
The finding has held up across replications. A meta-analysis of 23 experiments (nearly 8,000 participants) in 2020 put the overall recognition rate at 22.4%. A direct replication in 2022 with 600 people found 19.7%. The number bounces between 18% and 22% no matter how you design the study. About four out of five people are reading an ad and calling it journalism.
🤔 The “Sponsored” Button That Nobody Sees
The really unsettling part is how little the disclosure matters — unless it’s designed to be impossible to miss.
Campbell & Evans (2018) found that putting a disclosure before the article tripled recognition — from about 15% to 30%. That’s still a 70% miss rate. Kim, Pasadeos & Barban (2019) showed that labeling something “Ad” instead of “Sponsored” doubled recognition — from 20% to 40%. Still a 60% miss rate. Amazeen & Muddiman (2018) ran a field experiment with 1,200 people and found that only 11–15% could even recall seeing a disclosure after finishing the article.
The pattern is stark: even when the disclosure is present, most people don’t see it, don’t process it, or don’t care. The brain treats a visual format as a category signal, and once the category is set (“this is an article”), the ad label becomes invisible noise.
There’s a neuroscience layer to this too. Brennan (2017) ran an fMRI study showing that native ads activate the medial prefrontal cortex — a region associated with semantic comprehension and trust. Banner ads, by contrast, light up the dorsolateral prefrontal cortex — the persuasion detection zone. Same participant, same brand, two different brain paths depending on how the ad is dressed.
🔗 What This Means for Anyone Who Makes Ads
The implications for anyone in marketing or product design are uncomfortable. Native advertising works because it bypasses the consumer’s defenses — which is exactly what makes it feel ethically slippery.
The Persuasion Knowledge Model (Friestad & Wright, 1994) describes how consumers develop a mental “immune system” against advertising over time. By the time you’re an adult, you’ve seen enough banner ads to know an upsell when you see one. But native ads don’t trigger that immune response — they look like the very thing the immune system doesn’t attack: useful information.
This doesn’t mean native advertising is inherently deceptive. It means the burden of transparency falls entirely on design. The data is clear: a gray “Sponsored” label at the bottom of the page is not a disclosure. It’s a fig leaf. An honest disclosure needs to be prominent, early, and explicit — and even then, you’ll still fool three out of four readers.
For anyone who creates marketing campaigns — and Jasmine works in this space, thinking about how brand stories land with real people — the takeaway is double-edged. On one hand, native formats are the most effective way to reach a reader who has learned to ignore banner ads. On the other, the effectiveness is partly a function of invisibility. The ethical line isn’t between native and non-native. It’s between transparent and misleading.
🎲 The “Ad” vs “Sponsored” Experiment
Here’s a party trick you can try. Next time you scroll past a native ad on a news site, look closely at the disclosure label. Is it “Ad” or “Sponsored” or “Promoted” or “Presented by”?
The research says that “Ad” performs about twice as well as “Sponsored” in terms of recognition — but the industry overwhelmingly prefers “Sponsored” because it sounds softer. Almost every native ad platform defaults to “Sponsored” or “Promoted by.” In other words, the advertising industry has collectively chosen the label that its own research proves is the least effective at informing the reader.
Think about that for a second.
The only thing that actually helps people know they’re being sold to is a big, early, bold “Ad” label. And almost nobody uses it.