MMM: Were the right variables included in the model?
Understand if the foundations are right.
For an MMM to work properly, it must reflect everything that genuinely influences sales - not just media.
Sales rise and fall for many reasons: price changes, promotions, distribution shifts, seasonality, competitor activity, weather, PR spikes. If these drivers aren’t included, the model has to “guess” - and that guess can often wrong.
This matters for Out of Home because if a major sales driver is missing, the model may wrongly credit another channel that happened to be live at the same time. OOH, often running alongside other media, can easily lose or gain credit unfairly.
Can you see a full list of model variables - and are you confident they reflect everything that drives sales in your category?
YES!
I have a comprehensive list of variables and they reflect my business reality.
Good - your model has the building blocks it needs to isolate media impact properly.
That means credit is less likely to be misattributed simply because something important was missing.
NO!
I suspect important sales drivers are missing.
If key variables aren’t included, such as price changes, distribution shifts, competitor activity or seasonality, the model must still explain sales movement. And it will do so using whatever data it has available.
That often means media gets too much credit… or too little.
To understand what should be included and how to spot the gaps: