What’s Old Is New Again: Marketers Adopt Marketing Mix Modeling

Share this:

As signal loss complicates granular measurement frameworks like multi-touch attribution, marketers are reconsidering more holistic techniques like marketing mix modeling, or MMM. 

Street Fight checked in with Hugo Loriot, partner at fifty-five, a global martech consultancy that specializes in data, to learn why marketers are giving MMM another look and how to make the methodology work best.

Why did marketers move away from MMM, or why has its reputation declined over the last 10 years (if it has)?

Because MMM is based on aggregate data, it doesn’t provide the type of granular data that other advanced measurement solutions like MTA do.  Marketers like specifics, as they help with targeting and attribution, but MMM’s purpose is to help marketers understand how various marketing activities drive the business metrics of a product or service. 

Marketers have complained that MMM has too much of a lag, and this is partly because MMM is a statistical model that requires collaboration between data scientists and marketers to ingest and unsilo data. The rise of biddable media and audience planning has also raised expectations for getting a granular, near-real-time feedback loop that traditional MMM studies are not designed for. Marketers have had a hard time applying the insights MMM provides to a consumer level,  leading marketers to believe marketing mix modeling is “dead.”

Why is now the right time for a resurgence of MMM?

As Google phases out third-party cookies, consumer-level insights will be difficult to attribute regardless of what advanced measurement solution a marketer uses. The marketing industry is in flux with the rise of the privacy-forward consumer, and marketers are still learning how to function in today’s privacy-centric era. Tools that work now may not be as effective after third-party cookies disappear. With the dominance of walled gardens and introduction of key privacy regulations, there is a growing need for greater control over all steps of the marketing decision-making process — and a refreshed, faster, more granular and transparent version of MMM can offer marketers that.

Advanced measurement solutions like MMM offer clear benefits, especially for large global bands that operate across linear TV and digital media, online and offline channels. Technological capabilities are advancing and data teams are getting increasingly more power within companies. Now is the best time for marketing teams to reconsider their approach to MMM (many companies already use some form of MMM, though it may be a traditional or outdated model) in order to gain greater control over the marketing decision process, provide pragmatic recommendations, and accelerate decision making. However, it’s important to remember that like all marketing tools, MMM should not be the end-all-be-all of advanced marketing measurement, but an additional tool to help marketers make decisions that will increase ROI.

How should MMM be updated, if at all?

Advanced marketing measurement solutions have been successful and widespread in the market, but it’s difficult to bridge the gap between what’s promised with the actual functionality of the solutions. As third-party cookies fall by the wayside and data platform technologies rise, traditional MMM must evolve to become an always-on and accurate tool that provides real benefits to marketing teams. Companies can hire third parties to run marketing modeling solutions, but they are often costly, black-box and not an ideal long-term solution. Instead, brands need to start internalizing MMM to gain control of their decision-making. An internalized approach to MMM puts the responsibility of collecting and processing data on internal teams so data can be properly interpreted by those closest to the business and it creates a one-stop-shop to support internal teams in all strategic and tactical marketing decisions.

Data is only getting more compartmentalized, so developing an internal data platform can help marketers maintain control in the long run. This process, though time consuming, is an important investment as it allows the autonomous measurement of marketing effectiveness and gives more control over the consistency and quality of the data. An internalized and automated MMM program allows results to be updated at the pace of decisions, be it biannually or even quarterly. This approach allows marketers to refine results through continuous testing and a deep understanding of business objectives.

Does MMM work with MTA and other measurement practices? How should it be implemented?

From MMM and MTA to the new generation of attribution models, controlled tests, and geo-experiments, the range of tools for measuring and optimizing marketing effectiveness has become abundant. MMM is just one of the advanced measurement tools marketers should be leveraging and investing in, especially after Google extended the deadline for third-party cookie depreciation. MMM should not be approached as a standalone discipline, but rather as an overarching measurement approach that needs to be complemented with in-market testing. Whether using geo-based experiments, ghost bidding or ad experiments, running randomized-control tests across biddable media platforms help recalibrate MMM coefficients and increase measurement accuracy over time.

Joe Zappa is the Managing Editor of Street Fight. He has spearheaded the newsroom's editorial operations since 2018. Joe is an ad/martech veteran who has covered the space since 2015. You can contact him at [email protected]