The Fatal Flaw in Your MMM: Why Marketing Mix Models Must Include Price, Promotions, and Economic Data?
“Other providers we spoke to don’t include wider business factors.” That comment from a recent conversation with a client really shocked me.
If you’re providing Marketing Mix Modelling, then by definition you should be capturing all key business drivers in your analysis: price, promotions, distribution, product changes, seasonality, and many more.
Yet, too often, what’s being sold as an “incremental-focused Marketing Mix Model” is actually just a media-only attribution model. Not all products with the same name offer the same service.
What should I look for when buying or setting up a Marketing Mix Model?
- Includes all marketing factors
This appears to be a more common problem arising. In my experience, when setting up 100s of MMM programmes, invariably, there needs to be some compromise on the access to data to feed the models.
That said, not including key factors such as price, product, promotions, seasonality, economic or market factors is not an MMM model. These factors are crucial for capturing the baseline of sales and are clear contributors of purchase decisions.
Not including them will significantly overestimate the impact of your media. Meaning that you are not measuring the incremental media impact.
- A focus on accuracy & speed
These two factors shouldn’t be separate components. Taking longer doesn’t necessarily mean greater accuracy. However, there will always be some compromise between these two elements.
For example, if brand health or economic data is an important driver of your sales. This data is only available on a monthly basis. As such, you need to have a conversation as to how best you are going to capture this in your model.
- Transparent access to models & confidence intervals
MMM is a statistical technique. This means that each variable (e.g. Meta, TV, Price or promotion) in a model has a confidence level around them. Some variables will demonstrate high statistical confidence, and some will show a low level.
However, when MMM teams are reporting a single number, that level of confidence that backs this up is often overlooked. Buyers should seek to get access to this level of information.
- People who say MMM will solve every marketing measurement question
Yes, MMM is a great approach, but it is not the best approach for answering all measurement questions. Here at Linea, we work on the basis of matching your business question with the right measurement tool. See a more detailed overview here.
MMM Buyers Beware
Is it faster to not include wider business factors? Yes
Is it inaccurate? Yes.
Does it make marketing measurement easier for brands to understand? Absolutely not.
Your MMM partner should be fast, but never at the expense of accuracy or completeness.
At Linea, our focus will always be on providing transparent measurement & advice: full access to our models, outputs, and tools to help teams take action from measurement.
Don’t be swayed by flashy dashboards. What truly matters is what feeds them. The quality, breadth and openness of the analytics.