Marketing-Mix Modeling: Leaping Over The Walled Gardens’ Gates

"Data-Driven Thinking" is written by members of the media community and contains fresh ideas on the digital revolution in media.

Today’s column is written by Matt Voda, CEO at OptiMine.

In the last six months, both Facebook and Google have launched formal, public marketing-mix modeling (MMM) partnership programs. Participating vendors are allowed special access to more detailed campaign data for use in their marketing-mix models.

Because these models do not use consumer-level data to measure marketing performance, on the surface, they appear to be somewhat incongruous with the mainstream mindset toward tracking individual consumers.

This is especially true when considering the fact that both Facebook’s and Google’s measurement platforms rely heavily on their own consumer tracking to report campaign performance. Google even doubled down on consumer tracking by launching a free version of its own attribution platform, which makes the MMM partner program announcement seem even more of a head-scratcher.

So, what is going on here?

Some may argue that the driving force behind these partnership programs is advertiser pressure on the walled gardens to provide more transparency into their measurement methodologies. The transparency demands in the advertising community have grown louder with each announcement – and admission – of Facebook and Google’s measurement system errors, bugs and other issues. But while valid, the cries of “How can we trust the walled gardens to tell us about their performance?” may actually have nothing to do with the new MMM programs. The likely reality is that this is simply (finally?) an acknowledgment from the walled gardens that MMM methodologies, which have been around for 40 to 50 years, have their place in modern digital marketing measurement.

As brands spend more of their budgets on digital and seek to understand how these investments drive consumer behaviors across all touchpoints – online and offline – the digital tracking and attribution systems don’t provide all the answers required.

The realization by brands that it isn’t possible to track every consumer, across every device interacting with all of a brand’s ads and touchpoints, has been evolving for some time. Many brands initially adopted attribution systems to help them understand digital paths to purchase and track how consumers engage digitally with the brand. But these early methods, which rely on digital clicks and tracking, inherently miss all offline activity, such as in-store traffic and sales, call centers, sales agents and more, which is where much of the consumer activity still occurs.

With an MMM approach, the brand benefits from being able to measure all campaigns across both online and offline channels. Brands can also extend their understanding by using MMM to measure how other channels, such as TV, radio, print and OOH, contribute to and lift their paid social, display and paid search investments. With advances in machine learning, cloud-based scaling and automation, modern MMM approaches can also measure granular digital campaigns with more detailed advertising data, which is precisely what both Facebook and Google are offering.

This gets us to an important final point: Both Facebook and Google stand to benefit when they help brands understand how their ads are driving all customer activity – not just digital engagement. If an MMM can show a brand that a specific Google keyword is driving sales of their products in physical stores, both the brand and Google win. Google’s ad is now more valuable, and the brand has a new level of understanding on how to manage their marketing investments.

This new mindset toward MMM is a good thing, as it expands marketers’ understanding of the power of their digital investments and allows them to leap over the gates of the walled gardens’ own measurement systems for a more complete, comprehensive and independent view.

Follow Optimine (@OptiMineInc) and AdExchanger (@adexchanger) on Twitter.

 

Add a comment

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>