Evolving your Marketing Mix Modeling Program for Recovery 2021
It’s easy to lose touch in times like these. Consumer behaviors have changed. Some of this will be short-term while undoubtedly some changes will be longer-term. But brands will go on. Consumer preferences and loyalties will persist. How brands behave in these challenging times will play a large role in that equation. But how should brands measure and drive success in order to come out on the other side stronger?
The need for unified marketing measurement and optimization in times like these has never been stronger. What is unified marketing measurement and optimization? It’s essentially three things that are integrated to provide one “unified” set of answers:
- MMM (Marketing Mix Modeling) – marketing mix modeling of marketing, operations, promotional, discounting, competitive, macro-economic to determine the value/ROI and impact of each business driver
- MTA (Multi-Touch Attribution) – multi-touch attribution, or the real-time measurement of digital campaigns
- In-Market Testing – real-time, in-market testing capabilities that enact and test multiple campaign strategies efficiently, effectively and in a manner that provides directional reads against new ideas – quickly.
Bringing these three capabilities together in a unified approach establishes a feedback loop for capturing dynamic changes in business drivers and media effectiveness, re-estimating response curves based on recent and granular data, and validating results on an ongoing basis.
Under normal business conditions each of these is important in the role they play. In the conditions that COVID-19 has created it’s essential that each of these levers be used in concert to reset marketing strategies in order to accurately represent changes in market recovery, consumer demand and dynamic changes in the effectiveness of media and commercial investments on a more frequent cadence (monthly, weekly).
The Evolution of Marketing Mix Modeling (MMM)
Marketing Mix Models have evolved considerably in the past 5 years. Analysis that used to be conducted with a few dozen megabytes or gigs of data on an annual or semi-annual basis, as companies move to a unified approach, terabytes of data are being used. Models that were refreshed annually or semi-annually, are now being refreshed quarterly and even monthly. Changes in performance and consumption patterns can be read dynamically and corresponding adjustments made to recalibrate campaigns to target changing marketplace conditions and customer segments. The base analysis that is MMM is now a more rapidly evolving analysis of the brand and its consumers and is capable of providing the foundation throughout the year to read and adjust marketing, operational and promotional campaigns, while accounting for competition, economic conditions, weather and consumer trends. Because these models now include a holistic range of business drivers, many of our clients have re-branded their MMM program as Commercial Effectiveness or Commercial Driver Optimization. They incorporate new market signals using advanced techniques, capturing synergies and calibrating media response curves by time period, location, sales channels and consumer segments to optimize for and predict dynamic shifts in consumption and engagement patterns. Now more than ever, these advanced incarnations of MMM are critically important for business planning.
The Evolution of Multi Touch Attribution (MTA)
What has historically been called Multi Touch Attribution (MTA), also known as Unified Customer Attribution is the second essential part to the unified marketing measurement program. Unified Marketing Measurement and Optimization enables companies to dynamically target customers with specific campaigns, touchpoints, offers at a local level in real time. In a world where various markets and consumers will be in different stages of recovering, by leveraging the MMM roadmap and applying a combination of MTA and testing, companies will be uniquely positioned to capitalize on what is anticipated to be a consistently varying and fluctuating environment.
Many companies have already established a roadmap for moving toward a Unified Marketing Measurement and Optimization capability that integrates MMM + MTA + Testing to move to real-time, person or targeted segment-level media optimization. I cannot understate how important this will be to enabling companies to drive not only better financially viable decisions in the short-term, but decisions that may likely have an altering impact on long-term brand health. At a minimum, the next 18-months will present brands with perhaps the most significant challenges of our lifetime. How quickly and proactively brands decide to activate may play a governing role in their success for years to come. Establishing a fully integrated, unified platform that includes a robust data strategy, the deployment and validation of a real-time, unified capability leveraging person-level data typically takes several months depending on the current data and ad-tech situation at a company. Understanding how much to invest, where to invest, when to invest and how to optimize the mix of investment to successfully navigate the current environment while driving a successful back-half of 2020 and 2021 can’t wait months and in some cases can’t even wait weeks — so time is of the essence.
Many executives are asking: “How do we evolve our current marketing measurement program to address the current business challenges presented by the COVID-19 situation while we are in the process of establishing a Unified Marketing Measurement and Optimization capability?”
Below is a playbook for addressing this question over the short-term, mid-term and long-term.
The playbook focuses on taking advantage of your current MMM program to address the current COVID-19 situation, while building a Unified Marketing Measurement and Optimization capability.
Immediate (present): Enhanced Marketing Mix Modeling, adaptive to COVID-19
- Engage in “Smart” scenario planning – Optimization and simulation scenarios should always be created through a combination of predictive analytics and the right mix of business knowledge and consensus on external factors. Across most advertisers, while the ratio of effectiveness of some marketing tactics has remained constant, there are other tactics that can’t be used or must be used differently. By applying “smart constraints” and assumptions related to both marketing and non-marketing factors, executives are finding MMM to be a powerful capability during this time period.
- Add COVID-19 variables – By expanding the range data inputs to capture a broader set of people, health and economic related variables to represent the COVID-19 environment, we have found that models are able to better explain and predict the “base business” using dynamic estimation techniques.
- Increase the frequency of updates – Advertisers are moving to a Rapid Refresh Process (typically on a monthly basis) to capture changes in variable relationships and recalibrate response curves to reflect the latest changes in marketing strategy, business performance and market trends.
- Take advantage of intermediate KPIs – While it has always been a best practice to include intermediate metrics such as consideration, search query volume, site traffic, and other measures of consumer engagement, these metrics are proving to become even more valuable during this disruption. Understanding the impact of marketing on these metrics provides a view into leading business indicators as marketers are looking to harvest pent up demand and convert this demand as states enter their recovery stages.
- Test, test, test– Leverage testing to experiment with new touch points, creative, audiences, channel shifting strategies and geo-level targeting of customers while establishing a continuous feedback loop for refinement and validation of the MMM results.
Mid-term (2-5 months)
- Expand and integrate test and learn – Leverage and expand early testing to establish a continuous in-market testing program to align with the agile MMM approaches to ensure incrementality assessments account for current market dynamics. Feed testing results into MMM to ensure consistency, update models to reflect new media consumption and purchasing habits. It also provides the opportunity to try new ideas quickly with specific campaigns targeting key customer segments.
- Increase speed – Automate data sources to enable increased speed to insight and activation across agency partners, internal stakeholders, and technology platforms to enhance targeting based on message, audience, and media channel mix.
- Accelerate your unified measurement data strategy – Establish individual-level and micro-segment-level integrations with key players, begin onboarding of ads data to owned channels and CRM through cookie less technologies, integrate site activities with conversions, ensure accurate projectability of matched samples to the total business universe, build and validate cost workflows across the all partners.
Longer-term (6+ months)
- Deploy a Unified Marketing Measurement and Optimization (MMM + MTA + Testing) capability – Real-time optimization of marketing investments and targeting of consumers to drive omni-channel business performance through the integration of first- and third-party data sources, tools to drive activation, ongoing testing and activation feedback loops.
- Balance the optimization of brand and conversion media – Measure impact on brand towards Q4 ensuring base business recovery is benefiting from any revised marketing strategies. Maintain a balance of brand vs. performance media tactics leading up to holiday campaign windows to build and convert demand across response channels.