Marketing mix modeling: A marketer’s guide

With increased pressure to prove the value of marketing, it’s time to revisit the MMM approach.

Boards and the C-suite expect CMOs to lead the way to profitable growth in 2023 despite various macroeconomic pressures manifesting into the “triple squeeze,” making everything more expensive.

And although marketing budgets, as a share of revenue, rebounded last year to more than 9%, according to the 2022 Gartner CMO Spend and Strategy Survey, they are still lower than they were in 2020, forcing CMOs to achieve more with less. 

With increasing pressure to prove the value of marketing, smart CMOs are turning to marketing mix modeling, or MMM, to improve media performance and quantify their impact. 

What is marketing mix modeling?

Marketers still struggle to answer foundational questions about the impact of marketing on the business, such as “How effective is my digital marketing at driving in-store sales?” or “How will a 10% change to upper-funnel media impact bookings?”

Marketing mix modeling (MMM) can help answer them.

The goal of MMM is to measure the impact of advertising and promotions across channels while controlling for external factors outside of a brand’s control, such as inflation or consumer sentiment.  

The outputs from MMM are used in three ways:

  1. As a scorekeeper, to show the overall incremental impact marketing investments are having on the overall business.
  2. As a forecaster, to predict the outcome that raising or lowering marketing budgets will have on marketing’s contribution to the overall budget.
  3. As a coach, to suggest shifts to current marketing investments that improve performance.

In its simplest form: MMM helps marketing leaders plan future marketing spend and measure the performance of past investments.

How impact is measured can vary; a focus on incremental revenue is common, but modeling multiple outcomes is a growing trend, such as store traffic or new account sign-ups. The details of the modeling approach differ, but all forms use aggregate (not user-level) data. This allows MMM to nicely sidestep user privacy and other digital tracking concerns, as well as consider a wide range of channels — both digital and traditional — and external influences.

Consider MMM in action: A regional bank uncovers large performance differences by channel and lowers overall marketing spend through televisions cuts, while still increasing top-line sales by investing in more effective magazine and radio placements.

MMM is a technique with a long history, and it continues to evolve. If you looked at marketing mix a decade ago and dismissed it due to “insights only at the channel level” or “results only updated quarterly,” your marketing organization may benefit from revisiting the technique.

MMM measures the financials of brand investments 

Inflation, coupled with shifting media consumption patterns among consumers, is requiring marketers to prioritize sustained brand investment now more than ever. 

MMM can help CMOs quantify one of their trickiest investments to measure: upper-funnel activities that build brand sentiment and consideration, but are not focused on driving immediate sales. With increased delivery cadences, MMM can provide monthly updates on brand metrics that are useful for filling in the gaps of less-frequent brand-tracking surveys.

Imagine a product or service with a six-month sales cycle. Sales are driven by a healthy marketing budget with investments across the marketing funnel to drive awareness, consideration and, finally, sales.

Since the average marketing mix model looks at three years of history, that means the MMM would capture, quantify or measure the vast majority of upper-funnel spend. MMM would capture an even greater portion of midfunnel impact.  It’s common to see digital video, TV and Instagram get a boost from cuts to paid search. This leads to a media mix that is more effective overall for the same budget.

MMM is a core measurement capability 

Of course, while MMM offers significant opportunity to increase returns on media investments, the models require consistent reevaluation to consistently deliver the expected benefits. Organizations who have trust in their MMM also report higher growth compared to their industry peers.

Additionally, because MMM typically provides the most holistic view of the ROI on marketing activities, it often generates findings that challenge conventional wisdom. 

Envision a compelling email which encourages a prospect to search a brand’s website the next day leading to an unfinished checkout that triggers re-targeting and an eventual sale.  Who gets credit for the sale?  Email, paid search or retargeting?  Measurement for each channel could triple count credit for the sale, while revealing no insights around relative channel contributions.   Marketing mix, since it looks holistically at the ecosystem, can partition credit for the sale across the channels.

Because these scenarios often lead to false reservations about the model itself, it’s crucial to get finance on board early on in the process and articulate how it’s essential to the company’s success among senior leaders.

So, the next time your team reviews your marketing mix model, consider the following questions:

  1. Have you prioritized the insight objectives for your MMM? Is your marketing mix model actionable — meaning, do the outputs inform adjustments to marketing activities? This often involves adjusting spending levels, but could also include shifts to ad frequency or channel mix.
  2. Are you sufficiently assessing your marketing mix model to ensure that the predictions are delivering true incremental business performance and can be trusted by executives across the organization?
  3. Are you taking full advantage of the scenario planning, optimization and simulation opportunities that your MMM provides to improve future marketing efforts?

Across interviews with marketing leaders, we heard many talk about upcoming improvements to their MMM program, such as testing new data sources to better understand an external factor.

Everyone we spoke to could answer, “What is next for your marketing mix efforts?” So at least once a year — ideally more frequently — assemble key stakeholders involved in gathering the input data, building the models and using the results to adjust media plans. Then, discuss and commit to at least one improvement that focuses on prioritization, validation or optimization of existing efforts. 

It’s crucial to recognize that improving MMM is a journey — don’t stop at your first destination.

The post Marketing mix modeling: A marketer’s guide appeared first on MarTech.


About the author

Jason McNellis is a senior director analyst in the Gartner Marketing Practice, focusing on how to enhance marketing through machine learning, artificial intelligence, marketing mix modeling, and campaign measurement.