Evaluating retail media and the questionable cause fallacy

The evaluation of retail media is one of the larger issues for buyers today and there are a variety of effects that contribute to the confusion surrounding it. Here we look at one common mistake that actually applies to advertising outside of retail media, but which may suggest a better way forward specifically for retail media.

Confusing cause and effect is a specific example of the Questionable Cause Fallacy . The QFC has the general form:

- A and B occur together.
- Therefore A is the cause of B or vice versa.

In fact, there may be no common cause that links A and B.

This fallacy is made when it is assumed that one event must cause another just because the events occur together: the conclusion is drawn, without adequate justification, that A is the cause of B simply because A and B are in concurrence. It is possible that cause and effect are both present but the mistake is in choosing which is which, or it is possible that both are caused by a third event. When people put up umbrellas cars generally become get wet, but putting up umbrellas does not cause cars to become wet, the rain causes both. A typical mistaken causal effect might be:

- People wear shorts when it is sunny.
- People wearing shorts causes sunny weather.

To bring this closer to the reality of retail media, Rosser Reeves, who fifty years ago was head of the Bates advertising agency in the US, demonstrated what became known as the 'Rosser Reeves Fallacy'. He showed that consumers aware of his client’s ads were more favourably disposed towards those clients' brands across a range of measures, including likelihood to buy. Did this mean his advertising was more effective?

No. Generally, brand purchasers have a higher awareness of a brand's advertising than non-purchasers. We tend to remember better those things we like or are familiar with. So rather than those aware of the ads being more likely to buy the brand, it's actually those who buy the brand being more aware of the ads.

Is there a direct lesson for retail media here? Yes. It means that when we look at shopper buying behaviour, we must not jump to conclusions that seem obvious and right lest they in fact prove to be a fallacy. This is part of the key challenge to retail media. In-store advertising suffers the same plight as other forms of advertising: namely that proxy measures such as brand and advertising awareness are used to measure effectiveness.

There is not one single piece of proof that demonstrates that increased awareness is a measure of advertising's ability to build sales. Perhaps the only measure of success for retail media that has any validity is the ability of adverts on retail media to build sales. You could argue that building brand is important, but surely this is true only because in the longer term it adds to sales?

Given the success of loyalty schemes in measuring sales, down to at least a household level, this proxy measures should not be necessary. It is possible to do better and retail media, though being sold in conjunction with the retailer has the advantage of being able, at least in theory, to take advantage of sufficient information to create a truly valid measure of success. We will discuss how greater information about purchasing behaviour may lead the way to better metrics in a later article.