We use cookies. You have options. Cookies help us keep the site running smoothly and inform some of our advertising, but if you’d like to make adjustments, you can visit our Cookie Notice page for more information.
We’d like to use cookies on your device. Cookies help us keep the site running smoothly and inform some of our advertising, but how we use them is entirely up to you. Accept our recommended settings or customise them to your wishes.
×

Marketing Mix Modeling (MMM) for B2B – A Primer

Marketing Mix Modeling, commonly referred to as MMM in the advertising industry, has long been the gold standard when it comes to applying scientific and data-driven techniques to determine budget allocations for marketing and media plans. At its core, MMM involves ingesting the data for every part of the marketing mix (including external factors like inflation rates, competitor prices and beyond) and seeing how those parts influence an outcome. The “parts” can be thought of as independent variables, the “outcome” can be thought of as the dependent variable, and the analysis to determine the relationship between both is usually done through a regression analysis. In simple terms, you can think of MMM as exploring the relationships between drivers and outcomes.

A basic MMM exercise (as with most modeling) typically involves:

  • Determining the needed data points (online and offline, campaign and macroeconomic, etc.)
  • Cleaning and processing the data
  • Exploring the data via techniques like correlations and visualizing via histograms and other descriptive methods
  • Creating theories and hypotheses
  • Modeling the data (which involves applying market dynamics like ad stock and diminishing returns)
  • Reviewing coefficients (to understand the relationships between the variables in the data)
  • Reviewing the accuracy of the model’s predictions
  • Using the outcomes to establish baselines, channel contributions, and incrementality measures

The MMM outcomes also allow for the creation of response curves, which can enable a well-trained analytics or data science team to create scenario planners, explore synergies between channels, make predictions, estimate ROI, determine points of diminishing and maximum returns, and better inform key stakeholders on how to allocate their budgets across channels.

B2B Limitations with MMM

The success of an MMM, or any model for that matter, comes down to the quality, availability, and volume of data. As a default standard, an MMM typically requires 2+ years' worth of data. Yes, there are workarounds if the needed data sample is not available, but typically 2+ years is the standard. The data must also be accessible and available, and the dependent variable (aka the “outcome”) must be defined. This explains why MMM historically has had more appeal with B2C, as B2C transactions and their connected data points are highly addressable and available (with a few exceptions). But B2B marketing presents a separate set of challenges, namely:

  • Data management and data centralization – This is a common “issue” found in many large organizations. Data quality may not be standardized across departments (and in some instances even within the same department across years and changing leadership). High-value metrics like final conversions (i.e., sales) may not be easily accessible. Lack of visibility into final sales conversions across teams is a common theme in B2B marketing.
  • Dealing with multiple decision makers – B2B purchases are often at the enterprise level and range from thousands to millions of dollars. As such, B2B buying groups to support these large transactions involve multiple people, with different influence levels (influence does not necessarily correlate with seniority). There is also the maturation effect – decision makers may get new jobs or get replaced which adds to the list of factors that extend the buyer journey.
  • Long sales cycles – Most B2B marketing, advertising, and sales teams focus on some combination of equity marketing, lead generation, qualification, pitching, and closing, with integrations across sales and customer experience teams. These tactics roll into ABM (Account-Based Marketing) and ABX (Account-Based Experience) strategies and can take years to bear fruit.

These challenges are tied to key data points needed for B2B MMM, and they may seem insurmountable, which is why it is important to partner with an experienced analytics provider that thinks creatively and can provide intelligent workarounds to these limitations.

MMM has a Place in B2B Marketing Planning

The good news is that these challenges to B2B MMM initiatives can be addressed. A centralized data management solution, such as a Data Lake working in cohesion with a CDP (Customer Data Platform), with buy-in from the varied teams and stakeholders can be employed to help solve the data management roadblock. Rolling all decision makers/contracts into accounts and building MMM models at the account level can address the multiple decision maker roadblock. A lack of visibility into end sales can be addressed by using leads acquired as the dependent variable and as a directional proxy for sales. Long sales cycles can be addressed by modeling the data across longer timeframes/intervals and using decay transformations that find the best fit for each channel variable. A consultation session with a proven analytics partner can identify workarounds specific to unique business models.

Introspective Questions and Themes to Consider

So, you have made it this far and have decided that you want to build MMM capabilities within your marketing and measurement program. Below are some diagnostic questions to consider researching and answering as a first step:

  • What business questions are you trying to answer?
  • Who is your target end user?
  • What is your preferred outcome variable? Is it leads?
  • Are all the leads in your B2B program treated the same?
  • Have you ever done any valuation analysis (i.e., estimating the financial worth of the lead, tying leads to revenue, lead scoring)?
  • What are all the channels used in your B2B program?
  • Do you know and have access to all your marketing touchpoints?
  • How far back does your data go?

Answering these diagnostic questions and being available for a thorough consultation with an experienced analytics partner is a big step in laying down the foundations for accelerating your marketing optimizations using MMM.

Conclusion

MMM can and should be a part of every forward-thinking B2B marketer's arsenal. A well-developed MMM program can provide a scientific approach to allocating marketing and media budgets, all based on data. It can empower marketing and media executives when it comes to budget planning and forecasting, all the while taking into consideration marketplace dynamics. It can reveal the synergistic relationships between marketing channels. And, as a bonus, if you are considering or already have an attribution system in place for sub-channel/tactical optimizations, an MMM can help calibrate your attribution modeling efforts while controlling for external variables. With the rapid advancements in cloud-based technologies that allow for real-time access to structured and unstructured data on a global scale and innovations with machine learning algorithms and technology, this is an ideal climate to scale up your measurement and planning programs with MMM and make smart investment decisions tailored to your B2B clients and prospects.