What does MMM mean? What is M M M? These are very common questions among brand and marketing managers in recent times.
MMM means Marketing Mix Modelling or Media Mix Modelling and is a technique used to evaluate, define and optimise the Marketing ROI.
Personally, Media or Marketing Mix Modelling is the same, there is no difference between the two.
Some practitioners argue that it is different. It just increases the complexity on an already complex topic, without adding any relevant improvement or value.
I will write a post on it, but that is not the goal of this article.
What is?
Marketing Mix Modelling is a tool that helps to evaluate marketing activities, performance and take data driven decisions.
Basically the Marketing Mix Model attributes the impact of marketing and media activities to sales.
Imagine that you realised 1,000,000 £ in sales.
A Marketing Mix Model will decompose this value into different components.
For example:
- 40% will be a constant amount of sales, it represents the amount of sales you would have without any marketing activity
- 30% will be the sales related to price and promotions
- 20% will be related to the product distribution (for digital products can be used the organic traffic, but also paid search, on that I agree with Byron Sharp on how to consider paid search)
- 10% will be the sales uplift from Paid Media Channels (YouTube, Instagram, TikTok, Facebook…)
In this trivial example the Marketing Mix Model is attributing a % of sales to different variables.
The output of the model is sales, in terms of volume or units, the inputs are all the key activities that made the sales possible.
So you will have demand seasonality, price, weighted distribution and media impression.
Why is mmm useful?
There are three big reasons for using MMMs.
- A Marketing Mix Model helps understanding which marketing activities worked and which not
- A Marketing Mix Model helps planning and optimising future marketing activities based on past results
- A Marketing Mix Model helps the marketing department have a more data oriented approach. Working side by side with an experienced analyst in marketing mix model boosts the marketing discussion from a data perspective and leaves nothing to chance
MMM Case studies
I don’t want to make this all about me and that’s why I am highlighting below how big companies like Meta and Google are working in the media mix modelling field.
Meta is heavily investing in Robyn, an open-source tool for media mix modelling, while Google often publishes new research and findings on the subject.
They also share the mmm results and targets reached providing a perspective on how they can be truly useful:
- Subway was able to maximise the impact of marketing mix modelling to drive growth. Specifically by layering bumper ads on top of TrueViews ads they doubled their ROI compared with running True Views alone.
- Subways was able to be 20% more effective in terms of Media Spend and Sales Uplift thanks to Marketing Mix Modelling
- The Hershey Company’s portfolio of brands increased their Retail ROI by 40% year over year by using Marketing Mix Modelling and optimising their Media Mix
- Frito-Lay was able to optimise their marketing budget and choose the right video channels in their Media Mix
Marketing Mix Model Tutorial and Free Resources
I wrote other articles on Media Mix Model for helping people getting a better understanding of this field with additional sources and reference:
What is a Marketing Mix Model and other 7 key questions
What are AdStock Effects and Diminishing returns?
Freelance marketing analyst for building a mmm (Marketing Mix Model)
I don’t believe in automated tools for Marketing Mix Models and fortunately you can always investigate what is behind the scenes with Robyn or Google Lightweight MMM.
There are a lot of vendors of Automatic Systems, but I truly believe in the value of a tailored work and an open discussion with the marketing department.
Of course my opinion might be biassed, because I am a freelance analyst for marketing mix model.
On the other hand building a marketing mix model requires an eye for details especially to spot anomalies in the data or weird fluctuations in sales that can be related to data quality issues..
Marketing departments are always busy with a tight schedule that doesn’t allow them to investigate all the statistical nuances of a Marketing Mix Model and for that reason it is better to invest in an external consultant specialised in Marketing Mix Model.
If you want to know more about MMM, we can schedule a free meeting with the form below.
