Why (most) advertisers should be running Marketing Mix Modelling (M3)

MMM has been around for decades, but the release of several open-source solutions and the deterioration of digital attribution is heralding a new golden era for this field of marketing measurement.

The current marketing measurement landscape

The barrier to entry on M3 is lower than it’s ever been. Today, any individual or team with a decent understanding of a programming language like R or python *and* a good grasp of marketing effectiveness theory can run an end-to-end M3 project.

In the last 2 years the release of several open-source (free to use) libraries, including those developed by tech giants like Meta, Google and Uber, has revolutionised the field of M3. Even without utilising these software packages, it’s never been easier for analysts to build bespoke econometric solutions across a range of tools (e.g. the STAN library in R). And the algorithms that underpin these solutions sit at the cutting edge of data science and machine learning.

Off the back of these developments, there’s been a notable rise in the number of commissioned M3 projects from companies sitting across a wide range of industries, from omnichannel retail to apps, and a mix of both B2C and B2B brands (see Figure 1 below). This wider adoption has also coincided, and no doubt been influenced by, the deterioration of digital attribution we’ve seen in recent years. The introduction of Apple’s ATT in 2021, alongside other initiatives aimed at giving individuals more control over their personal data, is undermining the validity of bottom-up digital attribution. Digital campaigns are recording fewer conversions, with increasing uncertainty from marketers around how efficacious digital channels actually are.

Multi-touch Attribution (MTA), which enjoyed widespread use and acclaim in the 2010s, is particularly at risk. And whilst M3 isn’t without its own limitations, its holistic, “top-down” approach makes it generally more insightful, actionable and (ultimately) impactful than forms of digital attribution, especially for clients investing across a range of offline and online channels. It remains the only marketing measurement solution that allows advertisers to compare all channels like-for-like on critical business metrics like ROI and CPA. In addition, M3 can be used to quantify the contribution of non-marketing or external factors on business metrics (e.g. macro-economic trends, promotions and competitors ad spend).

The future of marketing measurement

Many of the biggest providers or researchers of M3, including Kantar and Google, are introducing what they call “Unified Marketing Measurement” (UMM) in 2023. Rather than referring to one specific measurement framework, each firm has their own definition of UMM. But the one unifier is M3, which lies at the heart of any UMM proposal. It’s the synthesis of M3 with some combination of other measurement solutions like digital attribution, experimental results (e.g. conversion lift studies on Meta) and brand equity studies that makes UMM “unified”. The glue that holds these disparate measurement solutions together is typically bayesian inference - a statistical model itself. So, in effect, UMM is an “ensemble” approach, a collection of sub-models interacting within a grand model.

Whilst the development of UMM is fascinating, it’s incredibly nascent technology. There is no one “best approach”, and most providers are charging vast sums to deliver UMM projects. Our suggestion for any advertiser looking to advance the way they measure the impact of their campaigns is to begin with M3, and build from there. At Cream we offer competitively priced end-to-end M3 solutions for any client that meets the required criteria:

At least 2 years of historical data
A good volume of conversions (ideally no weeks over the 2 year period without sales / leads / installs)
Access to media plans and ad accounts

You can reach out to me at jamie@creamuk.com or neil@creamuk.com if you’d like to learn more about how M3 can work for your business

Welcome to the golden age of marketing measurement