Error: Contact form not found.
Marketing Mix Modelling (MMM) has long been used by consumer packaged goods brands for holistic marketing effectiveness measurement. Attribution gained popularity with the rise of digital marketing. However, increasing signal loss due to privacy regulations, iOS restrictions and cookie deprecation have altered the digital advertising landscape, making attribution less effective and accurate. As advertisers, especially digital natives, re-evaluate their measurement methods, there is renewed interest in MMM as a comprehensive technique that can help quantify the impact of both marketing and non-marketing activity on sales. In addition, it is privacy-friendly and highly resilient to the changes occurring in the digital advertising ecosystem.
A limitation of traditional MMM was the complete dependence of brands on large consulting firms or analytics providers, making it time-consuming, expensive, and not easily scalable. But today, with many brands having their own data science teams along with significant investments in cloud computing and data analytics, new MMM Software as a Service (SaaS) solutions aided by automation and AI/ML techniques, allow advertisers to run MMM models in-house. They can run MMM more frequently, easily scale it to cover more of their marketing budget and analyze their own data and keep it in-house.
Analytic Edge worked with advertisers in Middle East & Africa (MEA) – telecom services provider Zain, UAE’s largest real estate platform Bayut, and e-commerce company Noon – to help them adopt MMM, run the program inhouse and leverage the results for business planning and maximizing marketing Return on Investment (ROI). This paper outlines why these advertisers explored MMM, their in-housing experience, the resource requirements, challenges and finally the impact of MMM on their business decisions and strategy.
As per the advertisers, the main reason to explore MMM was that attributing performance to marketing channels has become increasingly difficult due to recent data privacy regulations and limitations on user-level data sharing. In contrast, MMM offers a new perspective on attributing performance based on incremental changes. It also provides an opportunity to measure the performance across all marketing channels including offline channels such as TV, radio, print and Out of Home (OOH).
The advertisers wanted more control and transparency during the entire MMM process – from data upload to modeling to insight analysis – rather than just relying on black-box recommendations from an outsourced MMM vendor. They were also interested in better understanding the key marketing, business and macro drivers that impacted their Key Performance Indicators (KPIs). Some participants indicated that their organizational culture emphasized self-reliance and a Do-It-Yourself approach, which made in-housing a logical choice.
Identifying the non-marketing drivers of their business was initially challenging, and Analytic Edge guided them through this process. Advertisers appreciated that just gathering the required data is not enough to run an MMM. Creating transformations such as media lag effects and diminishing returns curves, and then applying business constraints were critical to building robust models. Initially, while a lot of this seemed a bit complicated, with practice and training, they were able to build models independently. The detailed training videos and personalized live sessions helped them understand the different stages and nuances of the MMM process.
The rapid changes in the data privacy landscape and resultant loss of signals and hence incorrect attribution of results, has made MMM more relevant for measurement than ever before. Hence, advertisers understood the urgency of adopting MMMs as a means to validating their existing attribution models.
They emphasized that bringing MMM in-house requires a significant time commitment from all teams involved (e.g. Marketing, Trade, Finance, IT etc.), especially in the initial stages when they are getting familiar with the process.
Participants also underlined the importance of having the right internal data systems in place to feed into the MMM platform. Lining up all relevant data (e.g., sales, media spending, in-store promotions, pricing, seasonality, macro etc.) from multiple sources and in diverse formats can be arduous and error-prone without the right data infrastructure. A lack of automation here can sometimes stymie the progress of in-housing MMM. Getting the data feed automation right is crucial to running MMMs frequently, on-demand, and coordinated with planning cycles for maximum actionability.
The experience helped advertisers better understand how MMM modeling works, especially how the impact of diverse factors – internal and external, offline, and online, marketing, and non-marketing – on KPIs needs to be considered to build accurate models.
A key takeaway was that MMM is both an art and a science. It is important to balance the end outcome of MMM with both a business lens and statistical rigor.
Analytic Edge’s MMM in-housing program is designed to help advertisers accelerate their journey in this space by adopting an incrementality based planning process through the integration of a self-serve MMM platform. Analytic Edge used its Demand DriversTM solution as the platform for the MMM SaaS in-housing program. The program delivers the following benefits:
AutomationThe platform is designed for automatic integration into the advertiser and publisher data ecosystem, enabling faster and more frequent data updates.
SpeedThe MMM process is automated and scheduled allowing for faster turnaround and more frequent insights.
Open-Box ApproachThe analytics team has a full view of the workings of the model increasing transparency and trust.
Forward LookingThe platform provides inbuilt simulation and scenario planning tools which allow businesses to use the outcomes for future planning.
IntelligentThe integration of Machine Learning techniques reduces analyst effort and accelerates speed to insights
The SaaS platform itself was intuitive and user friendly, with all the information necessary to go through the various steps of in-house modeling in a guided manner. The SaaS model gave a head start to organizations such as theirs that have not worked on MMM before.
Advertisers highlighted that an in-depth understanding of the brand’s market and business is a key prerequisite to achieve successful outcomes from the MMM study. Domain knowledge is key as it helps the modeler test and validate multiple hypotheses regarding the expected relationships between the key business drivers and the KPIs.
Advertisers indicated that they secured the support of different department heads for the project to run smoothly. Their Performance Marketing teams worked in tandem with their Business Intelligence and Machine Learning teams. Their Commercial Data teams were closely involved in creating the necessary input factors around their products, pricing, inventory, offers etc. as well as macro factors like salary weeks, back-to-school indicators etc. All had highly integrated cross-functional teams and MMM was seen as a key strategic initiative with support from the senior leadership team.
While there were different challenges for the different advertisers, a common theme was how to interpret results and outcomes differently for short-term conversion (lower funnel) channels and campaigns versus long-term branding (upper funnel) ones. Advertisers felt that a more granular distinction of marketing channels (e.g., brand awareness, consideration, conversion, remarketing etc.) would serve them better going forward.
Running MMM helped the advertisers to measure offline advertising and branding performance. They were able to put a number to its impact, giving them more confidence to continue investing in offline channels and upper funnel campaigns in the future. They also appreciated that, unlike attribution, MMM takes a much more holistic view of the impact of multiple drivers of sales and is not restricted by only the last customer touchpoint before a purchase.
Advertisers indicated that once they have a robust in-house MMM process running, they intend to regularly update and refresh the models and use MMM as a key driver for strategic decision-making.
Zain, Bayut and Noon found that using an MMM SaaS platform like Analytic Edge’s Demand DriversTM allowed them to run MMM and other predictive analytics in-house and on-demand, using a proven, transparent, integrated, and guided process. It offered them the cost, scale, and speed advantages they need in a modern measurement framework, while providing an alternative to both attribution and traditional MMM approaches.
Ultimately, choosing the right marketing measurement technique and deployment model depends on your brand’s specific
requirements, resources, and goals. Full-service outsourced MMM solutions are wellsuited for companies with limited in-house expertise and resources seeking comprehensive support and industry insights. SaaS-based in-house solutions are more suitable for organizations with multiple brands, markets, and in-house analytics capabilities, looking for scalability and real-time data analysis. Applying these criteria and selecting the appropriate model can empower your business to harness the power of datadriven marketing, optimize budgets, and make strategic decisions to thrive in the current competitive landscape.
Analytic Edge is a global analytics company that leverages technology and advanced analytics to help companies make data-based marketing decisions. The company’s flagship platform Analytic Edge Qube offers a suite of marketing analytics solutions with a Software as a Service (SaaS) model. The solutions include DemandDriversTM for always-on Marketing Mix Modeling (MMM), SynTestTM for AI powered Test and Learn, PriceSenseTM for pricing and promotion analytics, and PowerViewTM for analytics visualization. Analytic Edge works with clients across industry verticals such as e-commerce, mobile apps, gaming, consumer packaged goods, retail, automotive and many others. The company has offices in Singapore, India, US, Mexico, Brazil, UK, Switzerland, China, Japan, South Korea, UAE and Australia.
