I had a client once who paused all advertising for two months during a budget crunch. They expected sales to fall off a cliff. Instead, sales declined gradually, maybe 15% in month one and another 10% in month two. By month three, the decline accelerated sharply. What they were witnessing, without knowing the term, was the carryover effect.
The carryover effect is one of those concepts that should fundamentally change how you think about marketing budgets, but most marketers either don't know about it or treat it as an academic curiosity. I think that's a mistake, because understanding carryover is the difference between measuring advertising correctly and measuring it very wrong.
The carryover effect (also called adstock or advertising decay) refers to the lingering impact of advertising on consumer behavior beyond the period in which the ad was delivered. When you run an ad campaign this month, some of its effect carries over into next month, and the month after that, diminishing over time but not disappearing instantly.
The concept was first formalized by Simon Broadbent in 1979, who introduced the term "adstock" to describe this phenomenon. His core insight was simple but powerful: advertising doesn't work like a light switch. It works more like a dimmer that fades slowly after you turn it off.
Monash University's marketing dictionary defines the carryover effect as "the rate at which the effectiveness of an advertising campaign diminishes with the passing of time." A carryover rate of 0.50 means that the impact of this month's advertising retains 50% of its effect next month.
Here's why I think every marketer needs to understand this concept.
If you're using last-touch attribution or even multi-touch attribution without accounting for time decay, you're crediting the wrong campaigns for your results. Some of your "organic" conversions this month are actually the carryover effect of paid campaigns from last month. Some of your "brand" lift is residual advertising awareness from campaigns that ended weeks ago.
When you cut advertising spend and sales don't immediately drop, it's tempting to conclude that advertising wasn't working. But you might just be spending down your adstock. The real impact of the budget cut won't show up for weeks or months. Conversely, when you increase spend and don't see immediate results, you might be building adstock that will pay off in future periods.
Brand-building campaigns have longer carryover effects than direct-response campaigns. This is a core finding from Les Binet and Peter Field's IPA research, and it explains why brand advertising often looks "inefficient" in short-term measurement frameworks but outperforms in the long run.
The standard adstock model uses a decay function. The simplest version is geometric decay:
Adstock(t) = Advertising(t) + λ × Adstock(t-1)
Where λ (lambda) is the decay rate, a value between 0 and 1. A higher λ means the effect persists longer.
The most intuitive way to think about carryover is through the half-life: how long it takes for the advertising effect to decay to 50% of its peak.