Moving Beyond the Marketing-Mix Model

 

The marketing-mix model (MMM) has penetrated large marketing organizations to the degree that it has become a primary tool for assessing program effectiveness and tactical productivity, and an important element in determining overall payback on marketing investment.

Today, more and more companies are incorporating MMM directly into their shareholder value creation processes, extending its role beyond just the traditional view of periodic resource allocation to include guiding daily operations, building the links between marketing and corporate strategic goals, and tracking intangible asset creation.

Operational Guidance
Many performance management-oriented marketers are using “rapid-cycle” updates of their MMM to allocate incremental mix elements. Soft-drink marketers, for example, refresh their models monthly, allowing for week-to-week changes in marketing support planning. For example:

Moving dollars like this on a quick-turn basis is only possible when the model has been broadly stakeholdered and senior executives trust that incremental or unallocated funds will go to the areas of greatest need. Absent that confidence in the tools, the politics of reallocation is likely to delay decisions beyond the acceptable timeframe for action.

Linking Marketing to Strategic Goals
Mix models are increasingly being used as a strategy support tool to help set revenue and margin goals tied to an optimal level of marketing investment. Instead of simply asking how to optimize the budget allocated to achieve the business goals, MMM is now being used to set the goals.

Granted, there are still many market and environmental risks that may weigh heavily on specifying those targets. But the impact of these effects can be more reliably simulated when they sit atop an already proven and carefully parsed assessment of the quantitative business drivers. In short, companies running effective MMMs have the ability to move far beyond the traditional forecasting approach of straight-line projections of historical top-line performance.

An added benefit to this approach is that business targets are inextricably linked to the investments required to achieve them. The result is a far lower likelihood to cut spending mid-stream to achieve bottom-line goals.

Building the Asset Base
MMMs are commonly used to monitor the impact of programs on generating incremental revenue over the “baseline” — that prominent portion of the business that is attributable to all the pent-up brand and customer equity value, and which would presumably continue to flow in even if all marketing were to cease for a brief period of time. However, mix models are evolving beyond their short-term blinders and emerging to be better tools for determining exactly how that baseline is developing.

By more closely examining the extent to which the benefits of established brands, loyal customers, and well-developed distribution networks work independently or together to create a predictable stream of future revenues, MMM can identify clear pathways to increase both the magnitude and sustainability of them. Valuing this baseline and tracking changes over time can help clearly demonstrate whether the company’s marketing activity is creating asset value or just buying near-term results by tapping into its asset base.

MMM Cautions
Successfully extending the role of the MMM as described above requires an understanding of the traditional shortfalls of mix-modeling and some actions you can take to avoid them.

First, the assumptions and limitations of the MMM need to be well understood by the organization as a whole. As a black-box tool, it will be accepted only as long as the results do not suggest significant changes from the historical norm. The moment the MMM starts forcing decisions that result in reallocation of resources across political boundaries, questions about methodology, assumptions, and analytical robustness are sure to flare up.

Laying the acceptance groundwork is as important (and challenging) a part of MMM as building the algorithms or collecting the data. Having key stakeholders and decision-makers (e.g., finance, sales, business units, distributors, etc.) participate in building the inevitable set of assumptions underlying the model may require more time and heartache in the near term, but will dramatically reduce the infighting down the road when the real important decisions need to be made fast.

Second, MMM is historically much more accurate at measuring the correlation between what goes in (spending) and what comes out (revenues or profits) than it is at breaking apart what actually happens in between in the marketplace. As a result, these complex models sometimes “break” when the external environment changes significantly in ways that were not anticipated nor programmed into the inputs and assumptions. Changes in the competitive environment (new entrants, new products, changes in pricing, or competitive communications) can disrupt the historical relationship between spend and response.

However, if the dynamics of the environment are embraced as key elements of the process and not strictly isolated statistically or ignored, “broken” models can generate great insights. These marketplace events are natural experiments that enable marketers to see how relationships between the variables work in extreme contexts, making the model more useful and trusted going forward. The presumption here is, of course, that the expectations for the MMM were properly set and agreed amongst all key constituents beforehand.

This gets to the third caution regarding use of MMMs. The model will, on occasion, fail. Expect it. Marketers who treat the model like an oracle, without understanding its limitations or regularly question its output, leave themselves vulnerable when results do not match predictions. This is especially important in organizations where the MMM has been deeply ingrained into other processes as discussed above. Those who are too wedded to a single tool, regardless of its utility, will eventually be seen as one-trick ponies. Then, when their trick fails, their usefulness immediately drops and their credibility may suffer irreparably.

 

A Slice, Not the Whole Pie
Today’s marketing measurement toolkit needs to be much broader than just MMM. Risk — both to the manager who “owns” the MMM and to the organization as a whole — is magnified by over-reliance on a single tool. Even more important for the organization, however, is the perception bias that comes from over-dependence. Someone who only has a screwdriver in their toolkit thinks every problem gets solved with a screwdriver — though a hammer, chisel, or crowbar would have been the better tool for the job at hand. The same goes for MMMs.

Often a well-designed MMM becomes the focal tool of choice, crowding out investment in other useful measures. Analytical tools focusing in areas of the business like sales-funnel progression or pricing elasticity, which may be built on data sets that are less mature or assumptions without third-party benchmarks, will struggle to compete with the MMM, at least initially. Deep understanding of brand drivers, customer behavior and value, loyalty triggers, innovation, and so on also depend on measurement and analysis outside of the mix model. These measures often need to be derived with non-econometric techniques like customer surveys, market research, mining of transactional data, and experimental designs linked to well-crafted and understood assumptions.

Trying to answer these questions and measure marketing’s impact on them takes time and dedicated effort. Understanding the key questions facing marketing, then developing a stakeholdered and resourced action plan, is the first step. Building such a measurement framework is evolutionary, with some areas advancing more quickly than others based upon the availability of data and the skill sets present in house. With some areas, like MMM, the use of external resources can help accelerate maturation, but other areas will require careful tending over long periods of time before full benefits can be reaped. Quick wins are essential to maintaining momentum and proving value early on, but be careful not to let them be the only wins.