Ask Not What You Can Do for Your MMM, But What MMM Can Do for Your Marketing

Econometrics (read any form of holistic marketing measurement) can be fantastic. Discussions about its outputs form some of my favourite meetings, where numbers feed narratives and rigorous analysis helps decode the mysteries of marketing effectiveness. However, I get frustrated with the industry narrative around it, which simply does not do what it should to ensure that its outputs are put to best use. Too often, we fetishize the technical components of the model, when what really matters is achieving the best possible marketing outputs.

Firstly, let’s get one thing straight: I am not an econometrician. Ask me to build or query a model, and I am completely lost. At least partly because of this, my interest is not on the model itself but on driving effectiveness improvements from it. That is why it exists in the first place, there are not many marketers paying for econometrics just for the sake of having good econometrics. MMM is a powerful tool, but only if you use it wisely. Too much discussion focuses on the technical composition of MMM rather than its practical application to marketing campaigns.

So here are my thoughts and principles for getting real value from MMM and avoiding the pitfalls that can turn a great model into an expensive distraction. And no, to be clear, the ‘perfect triangle’ model of MMM, MTA, and Testing isn’t one of them. I’m not saying these elements are inherently bad, but the idea that they combine into a single, comprehensive answer to marketing effectiveness is flawed and oversimplifies the complexity of marketing effectiveness.

1. MMM Is a Strategic Tool, Not Just a Model

At its core, MMM is an educated guess wrapped in complex mathematics. A very good guess, yes, but still a guess. MMM is like a finely tuned compass, extremely useful and generally reliable, guiding you in the right direction. But just like a compass, it's not foolproof. It can point you toward a destination, but it won’t account for every variable along the way, like a sudden change in terrain—or a detour past the Restaurant at the End of the Universe.

It is the job of all involved in its creation, dissemination and use to ensure that MMM isn’t just a theoretical model but a strategic asset embedded in marketing decision-making. Outcomes should always trump inputs, what ultimately matters is how MMM helps marketers refine their approach and drive better campaign results. Results need to be crafted and presented in a way that arms the marketing department with actionable insights, ensuring they can optimize budgets, improve media allocation, and enhance overall effectiveness. Spending weeks refining a model but only a few hours on how the results are presented is a critical mistake, presentation and interpretation are just as important as the underlying analysis.

2. Transparency Enables Better Marketing Decisions

Any business investing in MMM (or any analysis) has the right to transparency into the model. If MMM is a estimation, then providers and clients need to understand exactly how and why estimations are being made. A good MMM provider should be like a great teacher, able to explain not just what the answer is but how they arrived at it. And just like the best teachers, they should welcome questions and challenges from their students, after all, learning is a two-way street.

Importantly, this transparency ensures that insights from MMM are actionable and can be directly applied to campaign planning. Marketers need to be able to explain MMM and answer questions from senior stakeholders internally, ensuring that its role and findings are clearly understood at all levels of the business.

Transparency is good for transparency’s sake, but its real importance is in providing the equivalent of an airplane’s Black Box for your MMM, capturing all the critical details that explain how decisions were made. Ensuring as many people as possible understand these details and assumptions will drive continuous improvements.

3. Use MMM to Shape Campaign Strategy, Not Just to Report on It

Marketing and advertising effectiveness analysis should always be done in reference to business goals, not just standard outputs. For instance, if the MMM is scoped to measure short-term sales without unpicking longer-term impacts, then assessing brand-building channels requires a careful, nuanced approach. In such a case, any modeller who fails to acknowledge that key aspects of a business’s marketing aren’t properly evaluated by the model is either displaying a lack of confidence or prioritizing their own model over the client’s success.

If you ever hear the phrase, “I’m only an econometrician,” as an excuse for not thinking beyond the model, run. Fast. Your provider should be able and want to think critically about your marketing reality, not just the numbers their model spits out. MMM should not be a retrospective report, it should actively guide campaign planning.

4. Integration: Embedding MMM into Planning and Buying Processes

The real value of MMM lies not in its intellectual purity but in its practical application. A model must be rigorous, but its real purpose is to drive better decisions. Whether built in-house or via an external provider, MMM outputs should seamlessly integrate into planning tools and processes. We are not at war with the Cylons! Static PDF reports with a few graphs or reporting in isolated tools should be a thing of the past, as they are simply not helpful. Instead, integrate raw data, decompositions, response curves, and other granular insights directly into the systems teams use to plan and buy activity.

Don’t forget, with AI on the MMM horizon, data collection and QA, the building of models, and the analysis of their outputs are all going to see significant advancements over the next few years. Let’s not be MMM Luddites keeping valuable info in silos, steadfastly refusing to accept the change on the horizon. Instead, let’s use it to drive dynamic, responsive marketing strategies.

5. MMM Alone Is Never the Full Answer

MMM has its own biases, limitations, and struggles. It can’t work in isolation. Relying solely on MMM for all marketing decisions is like trusting a single witness in a courtroom trial, it provides valuable testimony, but you need corroborating evidence. Anyone presenting MMM as a complete answer rather than a key piece of the puzzle is either overestimating its power or underestimating the complexity of real-world marketing.

Additionally, try not to over-rely on its outputs as the sole reporting metric for marketing performance. We all love a measurement framework and scorecard, integrate the MMM findings into this as a critical component of performance. Explain marketing performance in the context of the PNL, be honest about what we know and what we don’t, and focus on how to improve for next time.

The Key to Success: Bringing MMM into the Heart of Marketing

So in summary, MMM isn’t perfect. While it needs to be responsible, as accurate as possible, and technically robust, that is just the foundation of what is needed. The biggest gains, and therefore where most effort should be focused, come from using it wisely. MMM is a highly effective way to bring data-driven clarity to marketing decision-making, but only when it is embedded into planning, execution, and optimization processes.

It’s time to move beyond just discussing how MMM works and start focusing on how it can work for marketing teams. So, embrace it, challenge it, and most importantly, make sure it’s working for you, not the other way around.