Attribution in Marketing: How to complement your attribution model with Marketing Mix Modeling (MMM) and Experiments?

In the world of digital marketing, attribution allows us to understand which tactics and channels truly drive conversions. However, traditional attribution models such as last-click attribution and multi-touch attribution (MTA) have proven to have limitations in the face of changes in privacy, the elimination of third-party cookies, and the fragmentation of the user journey. In this context, measuring the impact of each channel is more challenging than ever. In this article, we explore new ways to improve measurement in digital marketing. What is Marketing Mix Modeling (MMM)? Within Marketing Science, marketing mix modeling (MMM) statistically analyzes the impact of different variables in a strategy, such as advertising, promotions, distribution, and pricing, on a company's commercial results. Its main objective is to identify how to allocate and optimize resources to maximize ROI. MMM was popularized before the Internet and continues to be used by major brands due to its aggregation approach, which makes it less dependent on user-level data. Advantages of MMM: ✅ It does not rely on cookies or individual user tracking. ✅ It is compatible with current privacy regulations (GDPR, CCPA, etc.). ✅ It provides insight into the effectiveness of multiple channels (TV, digital, and offline). ✅ It allows for evaluating the long-term impact of campaigns. ✅ It aids in budget planning. Disadvantages of MMM: ❌ It usually has slow processes in data selection and normalization. ❌ It requires large volumes of historical data. ❌ It requires expertise and experience to avoid overrepresenting reality. ❌ It requires expertise and experience to avoid being affected by biases or external factors. ❌ It does not adequately measure interactions between digital and offline channels, making it difficult to accurately measure the impact of performance marketing. What are Incrementality Studies? Incrementality studies are tests designed to measure the real impact of a marketing campaign by comparing results between an exposed group and a control group. Unlike MMM and MTA, they do not attempt to infer a channel's contribution from historical data or probabilistic models, but rather use real experiments to assess causality. An incrementality study divides your audience into two groups: a treatment group, composed of users exposed to the advertising campaign, and a control group, composed of users who either do not see the campaign or are exposed to an alternative message. Using this methodology, these studies measure the difference between the two groups: if the conversion rate of the exposed group is significantly higher than that of the control group, the campaign is considered to have generated an incremental impact. Advantages of incrementality studies: ✅ It measures causality, not correlation, eliminating erroneous assumptions. ✅ It does not rely on cookies or individual user tracking, making it more robust to privacy restrictions. ✅ It can be applied across multiple channels (digital and offline), allowing you to understand the real impact of different marketing strategies. ✅ It allows you to validate the effectiveness of strategic changes, such as reducing or eliminating a specific campaign. Disadvantages of incrementality studies: ❌ It is not real-time, as it requires a testing period before obtaining results. ❌ It does not offer continuous insights, as it measures impacts at specific points in time, which can cause the data to lose relevance over time. How do Marketing Mix Modeling (MMM) and incrementality studies complement your attribution? How to choose the complement to your way of attributing? Complement with MMM if... You want a macro analysis of the impact of all marketing investments over time. Your brand invests in traditional media (TV, radio, OOH) and digital media, and you need to measure them together. You need a tool to strategically allocate budgets across different channels. Complete with Incrementality Studies if... You want to accurately measure the real impact of your campaigns and ensure that your media investment is generating value. You need to test new strategies or evaluate the profitability of specific media before scaling up your investment. You are concerned that your Paid Media budget is capturing existing demand instead of generating new demand. slice1 Conclusion: Towards Unified Marketing Measurement (UMM) Although multi-touch attribution will continue to be a valuable tool for its ability to analyze digital events in real time, it will increasingly struggle to accurately reflect reality. Therefore, it is necessary to complement this type of analysis with marketing mix modeling and incrementality studies to obtain more precise, causality-based measurements. For an effective attribution strategy in 2025: Use MMM for strategic decisions and budget planning. Apply MTA to optimize digital campaigns in real time. Implement incrementality studies to validate the effectiveness of each channel.