Media mix modeling: The comeback kid of advertising analysis
Media mix modeling (MMM), also known as marketing mix modeling is regaining prominence in mobile advertising as the industry shifts back towards aggregated data. Let’s take a look at what MMM is in digital marketing and how it differs from attribution modeling.
What is media mix modeling?
Media mix modeling is one approach to analyzing marketing efforts by studying past sales to determine what contributed to those sales. The mix refers to a combined evaluation of product, price, place, and promotion. In essence, MMM tells marketers how effective their advertising was so that they can apply their learnings to future campaigns and get a better return on investment (ROI). Invented before the likes of cookies and attribution, MMM was one of the initial ways marketers were able to take the guesswork out of which channels were most effective for user acquisition (UA).
Media mix modeling vs attribution modeling
With the growth of digital marketing came attribution modeling. Attribution modeling focuses on user-level interactions with an app. We introduced more relevant, personalized push notifications this week and saw an instant reduction in churn among our targeted cohort. User-level data is still a crucial part of any mobile marketing strategy, with ATT opted-in users and all Android campaigns measurable as before, but working with SKAdNetwork post-iOS 14.5 means aggregated data is essential to success.
In contrast to the highly-detailed, specific data offered at user-level, MMM gives a more macro-level view of marketing performance. Our revenue increased this quarter while we ran this campaign. From here, it’s about leveraging tools and techniques that allow you to glean as much information as possible from this kind of overview.
Opportunities and limitations of media mix modeling
Now that we’ve covered the basics, let’s get back to the heart of why MMM use dropped off, and why it’s now making a comeback. Introduced in the ‘50s by the retail industry, media mix modeling became popular in the ‘60s and ‘70s. It was championed by marketers for its ability to combine first-party data like leads, marketing spend, and revenue with external factors like market conditions, cultural trends, competitor activity, and seasonality.
However, MMM lacked the immediacy, granularity, and transparency that mobile UA managers need to make real-time decisions regarding budgets and campaigns. It required years of data and a marketing budget of tens of millions to adequately provide what was only a high-level view of advertising effectiveness.
Now, this more general picture of performance is becoming a valuable way to paint a picture of campaign performance. MMM can be helpful in measuring connected TV (CTV) and billboard performance, as well as Android and iOS. B In fact, the introduction of App Tracking Transparency (ATT) with iOS 14.5 has been a catalyst in re-transitioning back to MMM.
Previously long-held best practices for iOS campaign measurement were turned on their head. Marketers were now faced with complexity and ambiguity when regarding campaign performance and insight on iOS. This meant working to develop new strategies and technical solutions using a combination of opted-in user-level data and aggregated SKAdNetwork (SKAN) data.
Taking MMM to the next level
Now, some mobile marketers are shifting toward a hybrid use of MMM and attribution models, with the help of a mobile measurement partner (MMP) like Adjust. Used in this complimentary way, MMM can help to provide context to the insights attribution presents to marketers. Higher adoption of MMM is encouraging industry-leading mobile marketing companies like Meta to develop their own media mix models like Robyn that are more relevant to our digitally-focused era than the original Bayesian model of the 50s.
Adjust, too, is building our MMM tool ‘Budget Planner’. Currently in beta testing, this tool will allow clients to delve deeper into the analysis of their past performance gained from attribution data. We aim to enlighten marketers with a more accurate view of the potential future impact of their marketing channels. Armed with this knowledge, app marketers will know how to best allocate their marketing budget and easily visualize the probable outcome of spend across channels in order to make the strongest decisions possible.
We’re continuing to work on our MMM solution to give you another way to add value and insight to your attribution data. For now, learn more about Adjust’s current product offerings.