retail news in context, analysis with attitude

IRI Spotlightby Doug Brooks, Senior Vice President, Analytics and Modeling Services, Information Resources, Inc.

Content Guy's Note: Today we're introducing "Spotlight," a new sponsored feature here on MNB in which industry thought leaders will offer extended thoughts on the critical issues that face the business.

Marketing mix modeling (MMM) is a powerful approach for gaining industry-tailored insights that can directly lead to improved marketing ROI. The retail industry is a perfect fit for MMM, given the continuously evolving attitudes and behaviors of shoppers, the large amount of data available for analysis and the fierce competition that characterizes the market.
MMM is a statistical analysis linking multiple variables to changes in consumer behavior and supports the development of forward looking business simulations and optimization exercises. The power of these models is derived from the statistical analysis of two years of historical data, reflecting non-linear behaviors and including inputs that represent all significant influencers on business performance.

MMM benefits companies in the following ways:
  • Provides backward visibility into the ROI for marketing tactics
  • Helps diagnose the reason for changes in business performance over time
  • Forecasts future trends and predicts results of any changes made to the marketing mix as a part of response planning

While MMM has existed for decades, it is still misunderstood and misapplied by some, with sometimes disastrous results. To be successful, marketers must realize MMM is equal part art and science. Marketers must pick a point where the data is "good enough," realizing that improving accuracy from 95 percent to 98 percent is rarely worth the delay and expense. Marketers must recognize it is impossible for the model to explain or predict 100 percent of sales activities. The MMM model must align with the current business challenges and planning processes and should map to the analytics decision makers need-whether geographic, socio/economic, store-level, or brand-level.

Addressing these key challenges up front will provide a strong foundation and help avoid any potential failures when building a MMM that will fit the company and its goals. The following 10 recommendations provide a valuable set of tested and successful ideas to help executives create an effective marketing mix modeling program:

1. Choose and focus on the most actionable point(s) in the consumer funnel: Determine the greatest needs of company managers and hone in on solutions for that first.

2. From day one, gain cross-functional buy-in: Include key managers from all relevant teams, such as market research, finance, sales, operations, marketing, and sometimes other functions.

3. Align analytics with the latest business challenges: Ensure the analytics meet the needs of managers to address significant issues. Provide managers with reports specific to their needs.

4. Translate, integrate, and triangulate to tell the whole story: MMM cannot be developed in a silo it must be integrated with other company strategies to create a "single point of truth."

5. Make it forward-looking: Rapidly fluctuating market conditions make reviewing backward-looking insights in an attempt to make predictions a thing of the past. MMMs need to be forward-looking, providing "what-if" scenarios to stay on top market changes.

6. Incorporate and synergize digital and emerging media: Digital and social media are becoming standard parts of media research. Examine the relationships between digital and traditional media consumption to prioritize different types of media correctly. Not all media impressions are created equally.

7. Manage expectations. Reinforce strengths, limitations and appropriate application of MMMs. It's an ongoing process often requiring a 2-3 year investment to reach the desired maturity level.

8. Update analytics frequently to put the right information in the hands of the right people at the right time. Timely delivery of insights is critical. Successful organizations are updating their analytics and continuously recalibrating them in response to changing market conditions.

9. Educate, Educate, Educate. Keep cross-functional stakeholders across the company educated on progress at every phase of the project.

10. Experiment with complementary and leading-edge technologies.Test different analytical tools and approaches, such as in-market tests, consumer funnel analysis, brand equity analysis or agent-based modeling.

MMM is a critical exercise for gaining breakthrough insights that can directly lead to improved ROI. Managed well, MMM can provide significant competitive advantage to companies across almost every industry by offering new, predictive insights that can directly result in revenue and profit increases. Poorly managed MMM can point companies in the wrong direction with disastrous consequences. Executives eager to implement MMM should take the time to conduct an exhaustive study of successes and failures prior to undertaking a new MMM exercise. Companies with limited MMM experience should consider involving subject experts who can help the company's cross-functional team successfully navigate the many decisions involved.

For more information, click here.

KC's View: