Why Your Best Ad Turns Into Your Worst Enemy

In 1968, psychologist Robert Zajonc sat a group of subjects in front of a screen and showed them a series of Chinese characters. Some characters appeared once. Others appeared twice, five times, ten times, or twenty-five times. The subjects had never studied Chinese—the characters were pure, unfamiliar shapes to them. After the viewing, Zajonc asked them to rate each character on how much they liked it. The results seemed intuitive at first: the more often a subject had seen a character, the more they liked it. Familiarity felt good. But then something strange happened. At twenty-five exposures, the ratings didn't keep climbing. They dropped. The subjects had crossed a threshold where seeing the same shape again stopped creating warmth and started creating irritation. This is the shape most marketing teams do not want to talk about. The mere-exposure effect, which every brand manager learns in their first month, is not a straight upward line. It is an inverted U. And the peak arrives far earlier than most ad campaigns assume. In Zajonc's study, liking peaked at ten exposures and declined at twenty-five. That is roughly two weeks of moderate-frequency advertising for a national brand. By week three, the same creative that built initial preference is actively eroding it. The mechanism behind this shift matters more than the number. Early repetitions work because of processing fluency. Each time a consumer sees an ad, their brain recognizes the pattern a little faster. That speed feels good—it signals safety, familiarity, a kind of cognitive ease. The first few times you hear a new jingle, it grows on you. The first few times you see a brand's mascot, it starts to feel like an old friend. But fluency has a ceiling. Once a stimulus is fully learned, additional exposures no longer deliver that small reward of recognition. The brain switches from "I know this, I like this" to "I know this, I've seen this before, why am I seeing it again?" That is the inflection point. Berlyne's two-factor theory from 1970 described it as a race between positive habituation and boredom—and boredom eventually wins unless the creative changes. This is not a theoretical problem. In practice, it means a campaign that runs a single video asset for eight weeks on connected TV is likely generating negative returns by week four. The first two weeks build awareness and preference. Weeks three and four flatten into indifference. By weeks five through eight, every impression is a small tax on brand sentiment. The consumer does not consciously hate the brand, but they start scrolling past, muting, or feeling vaguely annoyed. That annoyance has a name in the literature: advertising wearout. Tellis's 1997 research on wearout showed that the effect is not uniform across all ad types. Informational ads—those with concrete claims or demonstrations—tend to wear out more slowly than emotional or image-based ads. A jingle with no lyrics will fatigue faster than a spot explaining a product feature. A logo placed in the corner of a video will irritate sooner than a narrative that unfolds differently each time. The mechanism is the same, but the slope changes. What this means for a marketing manager is straightforward: frequency caps and creative rotation are not optional luxuries. They are the operational tools that keep a campaign on the left side of the exposure curve. If you are running a single creative across programmatic display, social, and CTV, you are almost certainly overexposing it. The channel mix does not reset the counter in the consumer's brain—they are still seeing the same character for the twenty-fifth time. The fix is not to reduce total impressions. It is to change what those impressions contain. A brand can maintain high frequency across a campaign if it rotates multiple creative assets, each one offering a new stimulus for the brain to process. The character changes, even if the brand stays the same. That is why effective campaigns often sequence different messages—awareness, then feature education, then social proof, then offer—rather than hammering one message at full volume. There is no universal peak number. The ten-exposure result from Zajonc was a lab finding with unfamiliar symbols. Real campaigns face different stimuli, different contexts, and different audiences. A high-involvement purchase like a car might tolerate more repetition than a low-involvement purchase like a snack. A B2B audience might fatigue faster than a general consumer audience. The shape of the curve is stable, but the inflection point moves. The discipline is to find your own peak. Run small frequency tests before committing to a flight. Measure not just recall and purchase intent, but self-reported annoyance and ad avoidance. Track whether click-through rates drop after a certain impression count. The data will tell you where the curve bends. Marketers have spent decades assuming that more exposure is always better exposure. The science never supported that assumption. Zajonc's subjects told us in 1968 that familiarity becomes friction once it crosses a threshold. The only question is whether you will measure that threshold or let the campaign run until the muting starts.