Quantifying both the short-term activation and long-term brand-building effects of advertising is a central part of commercial marketing mix modelling (MMM). To address this issue, we recently introduced a new peer-reviewed modelling framework for estimating long-term effects, building on our previously published research.
Given the growing adoption of this approach by industry vendors, we have produced a peer-reviewed prequel paper showcasing why more common Adstock-based approaches to long-term measurement are so flawed and the implications for the marketing industry.
The Problem
The Adstock concept is ubiquitous in commercial MMM and attempts to capture the total (decayed) effects of marketing on sales over time. Short-retention rates are typically used to represent activation effects, whereas longer retention rates are often used to capture brand-building.
On face value, this makes intuitive sense. It is natural to model short-term activation with stationary Adstock variables - perhaps with variations on simple geometric decay such as polynomial distributed lags. It is also tempting to use high-retention Adstocks to model brand-building, where long decay rates coupled with repeated exposure can mimic a strong sales growth effect.
However, trying to measure brand-building in this way is inconsistent and inherently flawed. Not only does it characterize each effect in terms of the same type of distributed lag process, it approximates long-term advertising effects with an ad hoc drift term. This can lead to spurious correlation issues and tells us nothing about the dynamics of long-term brand-building.
The Solution
The only credible way to separate short-term activation and long-term brand building effects is to first identify whether a permanent (long-term) component is actually present in the sales data. If so, observed sales can be meaningfully decomposed into short-term variation and long-term (base) evolution. Brand-building effects then equate to the impact of advertising on base evolution, whereas activation effects correspond to transitory variation around the baseline.
To illustrate the point, we present some common approaches in the academic literature, culminating with an overview of our 2022 paper - Modeling short-and long-term effects in the consumer purchase journey. Here, we demonstrate how advertising builds long-term ‘memory structures’ through positive emotional brand perceptions, leading to a persistent increase in long-term purchase probability. However, initial long-term effects are negligible and take time to build up due to the average purchase cycle. The result is a more realistic and accurate representation of long-term advertising effects in marketing mix analysis.
Implications
The implications for the marketing analytics community are twofold.
- We demonstrate both theoretically and empirically how a popular industry approach can lead to spurious claims of long-term effects. This highlights the need for more credible measurement techniques in practical MMM work: serious estimation of brand-building effects needs to be based on realistic models of underlying consumer behaviour rather than simple extensions of the short-term.
- We show how long-term effects can mean different things depending on the time series properties of the data. Consequently, if commercial MMM practice is to prove the real value of marketing investments to the C-suite, and provide meaningful industry generalisations, brand managers need to question how long-term effects are defined and measured.
