Retail media was one of the biggest industry conversation starters of 2023. Every retailer from Dollar General to Cars.com has launched a retail media network to capitalize on their existing first-party data.

As both a new channel and data layer empowering existing channels, it has the potential to transform the advertising industry. But one question stands out: How can marketers measure it all?

Industry surveys have shown that marketers are not confident in the measurement of retail media, and the primary reason is due to a lack of standardization. Particularly as retail media moves beyond search into off-site media, CTV and in-store, measurement standardization is critical to ensure that advertising outcomes are properly captured.

Today, the vast majority of retail media is measured through last-touch attribution using lookback windows—the length of time after an impression or click that an ad can take credit for a sale. Lookback windows vary by platform and ad type. A lot of focus has been on standardizing these lookback windows so they are comparable across platforms. While this is a step in the right direction, it oversimplifies the problem of capturing advertising’s impact.

Lookback periods should vary by platform and channel. For example, a 14-day lookback window may overestimate the impact of a search ad but underestimate the impact of a CTV ad given the difference in how these ad forms impact consumer perceptions and behavior. Ultimately, lookback windows are based on assumptions, not answers. 

The adstock effect

Rather than taking a one-size-fits-all approach, marketers should consider a core component of measurement: how and when the impact of an ad occurs. The concept of adstock offers a dynamic approach to solving this problem that takes cross-channel differences into account and has long been incorporated into measurement in other forms of advertising outside of retail media. 

Adstock is used to estimate the effect of advertising on consumer behavior and determine the saturation point or point of diminishing returns. This captures the cumulative effect of multiple exposures and the decaying effect as the impression of those ads fades with time.