Haven’t used marketing mix modeling yet? Here’s 6 reasons why you will
Dynamic changes in economies have made measurement more important than ever
The industry is moving towards greater privacy, and digital platforms are adapting their technologies for advertising and measurement accordingly. This leads to less available data, which directly translates into challenges in the measurement area.
There is consensus in the industry that there is no “silver bullet” tool that will solve all those challenges, but rather businesses will need to re-create their measurement frameworks using a combination of tools and techniques. Many practitioners (Google, Meta, BCG) recommend marketing mix modeling (MMM) as part of the toolset.
This traditional approach is a statistical method that decomposes a company's business outcomes into factors which influence it. In other words, it tells you how much of each ingredient to use to bake a cake. Introduced to marketing by large Consumer Packaged Goods companies last century, MMM operates on aggregated, non-user level data which makes it an ideal fit for a privacy first world.
So why should you care now? MMM has evolved significantly during the last 24 months and here are six reasons why marketers should consider MMM as part of their toolkit.
Suitable for any business
While it originated in Consumer Packaged Goods companies, now MMM can be used by any business. During recent months we observed skyrocketing adoption among digital heavy and even digital native businesses. This adoption is connected to recent innovations within the industry that have lowered the barrier to entry. Typically seen as high cost, long-term and resource-intensive projects, MMM can now also be executed as a SaaS solution with a significantly lower price than a typical stand alone project or even for free with use of open source code that data scientists can run in-house.
Complements the measurement puzzle
Each measurement tool has its limitations. Marketers married to attribution can benefit from a more complete and accurate picture of business outcome drivers. Whilst experiments are highly accurate, they are also difficult to scale and rarely provide a more holistic cross media understanding of advertising impact. MMM offers a holistic business overview perfectly supplementing other techniques used by marketers today.
Fast and automated
Historically MMM took weeks if not months to be completed. Due to many factors (including faster data collection processes as well as the use of AI technologies to speed up some stages in the modeling process, while reducing any potential human bias), this is not the case any longer. MMM is now faster and more dynamic, being a more useful tool to support decision making in the "fast-moving" digital world.
Granularity that meets your needs
Daily data split by placement, optimization technique or any other execution layer, is broadly available for marketers within MMM. Using data with that granularity means insights can also be actionable on the channel level. More granular insights will translate into a better media plan optimized for business outcomes.
Actionable
Contemporary measurement tools are expected to provide not only strategic directions (like channel-level information) but also tactical and actionable how-to guides frequently evaluating campaign performance. Granular, omnichannel data combined with innovative methods and modeling techniques are making MMM also highly actionable to drive optimizations at the campaign level.
Accurate (despite operating on aggregated data)
Combination or calibration with other tools (especially experiments) brings precision to aggregated views within MMM. This process makes it easier to measure incremental value from each media channel.
MMM and its evolutions are here to stay. It is time for marketers to get started on their MMM measurement journey if they haven’t yet. The best way for marketers to understand advertising efficacy is to adopt a unified marketing measurement approach consisting of modern modeling, causal testing (experiments) and attribution tools.
Before the doors to individual-level data are closing indefinitely, marketers are urged to consider setting up marketing mix modeling now. It will help prepare brands for the changes to online privacy. Time and resources invested in building models now will massively pay off in the near future. Learn more about marketing mix modeling from top industry experts’ presentations during Meta’s MMM summit 2022.
Alfonso Calatrava, Gabriel Matwiejczyk, Meta