Error: Contact form not found.
Market Mix Modelling (MMM) has become a vital analytical construct as it enables the business to understand key drivers, establish marketing return on investment and do “more with less” in a privacy-first world.
In a recent conversation with analytical gurus, they agreed that we have the data (in fact, an abundance of it) and we have the tools and technology to create robust MMMs.
However, crafting a compelling narrative around the results and insights is critical to unleashing the power of MMMs.
Numerous theories, best practices, and examples delve into crafting a captivating narrative from data that practitioners can use to craft a compelling narrative. In our experience, three distinct pillars are critical to crafting a narrative are:
Create unwavering belief in the model People will often dispute the results of the MMM, especially when the results are less than favourable. These disputes are often couched as it’s backward-looking, it’s a black
box, is this correlation or causality and the list goes on. To satisfactorily address such concerns and prevent this line of questioning, it is important to do three things.
Explain what a MMM can and cannot do. The “cannot do” is more important than the “can do” as it contextualizes the ability of the MMM. We know for example, that it can help determine the ROAS of different media channels and help optimize marketing spending. But since MMM relies on analysis of historical data it cannot, for example, forecast how a new campaign will perform. If we haven’t done something in the past, MMM cannot forecast how it will do in the future. Similarly, MMM cannot forecast the impact of large, unforeseen events e.g. Covid. And importantly, MMM cannot provide accurate answers if we don’t have the necessary data at the right levels of detail.
Stress the importance of using the MMM as a starting point and overlay the results with acumen. While MMM will produce its results, these should be combined with acumen or common sense to ensure accurate results and takeaways.
Capture the organization’s prior knowledge into the modeling. Clients have a wealth of experience and learnings about their brands. It is not just important but also a best practice to encourage clients to collaborate and engage in the modelling process and incorporate these Bayesian priors into the MMM to test their learnings and hypotheses.
Use signposting to enhance engagement Presenting an MMM to various stakeholders simultaneously poses a significant challenge due to their diverse backgrounds and differing priorities. For instance, while a data scientist may prioritize model robustness, a finance professional might focus on the accuracy of the ROI. Compounding this challenge is the limited time available for presenting findings.
To tackle this issue effectively, we propose the use of signposting. Signposting involves strategically incorporating specific keywords, phrases, or a structured approach in the presentation to guide the audience through the thesis. Elements such as highlighting important words or phrases, calling out key takeaways for different audiences or using impactful charts instead of text are useful techniques. Such techniques help stakeholders quickly grasp the key points without expending unnecessary mental effort.
The primary advantage of signposting lies in its ability to enhance the readability of the presentation. By making the content more skimmable, signposting facilitates a quicker understanding of the material. Consequently, longer documents feel more digestible and fast-paced, thereby increasing the likelihood of audience engagement and desired actions.
Leverage the “linearity” to tell a story MMM analyzes historical data to give a view into the future. Data is often presented in a linear pattern and often it’s not an up-and-down line. Build your narrative around these variations. Tell a compelling story of why sales move up or down – start from the beginning, explain the variations or ‘twists’ in the middle, and conclude with a satisfying end.
For example, a campaign started airing in late April but why did sales peak only in June? This possibly indicates 5-6 weeks from awareness to purchase. Or why did sales decrease on either side of a large promotion (e.g. Black Friday)? The likely reason is consumers pulled forward or deferred purchases to take advantage of the sale.
In conclusion, to help brands better understand and operationalize MMM insights, it is important to craft a compelling narrative around the results, that clients buy into. While building confidence in the model is the first step, other techniques such as signposting and leveraging linearity to tell a compelling story can truly help clients unleash the power of MMMs.
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, China, Japan, South Korea, UAE and Australia.
