Customer Case Study: Marketing Mix Modeling
A leading Mobile App company sees a projected reduction of 5% in annual marketing spend using decisionOS
Customer A large Mobile App Company
Industry Consumer Apps
Use Case Marketing Mix Modeling
Value 15x ROI, 5% reduction in marketing spend
The Challenge
Understanding and quantifying which marketing channels drive installations and recurring paying customers is key to determining how to allocate future marketing resources to maximize revenue growth. However, the client’s existing approach to marketing mix modeling (MMM) had some limitations.
The client needed a causal approach to attribute the causal effects of conversion, simulate how these would differ in varying scenarios, and then optimize their marketing budget allocation.
-
1
ExplainabilityThe marketing team struggled to quantify which campaigns were driving uplift
-
2
Learning from correlationsLeading to the budget being allocated to the wrong channels.
-
3
Focus on predictionsThe client wanted to go beyond predictions and have the ability to simulate differing scenarios.
The solution
Understanding the limitations of traditional MMM approaches, the marketing team turned to causaLens to help build trusted attribution models. Using decisionOS, the client was able to build:
- Explainable and transparent causal models that provide an intuitive description of the interactions between different marketing channels
- Models that do not rely on historical correlations but cause and effect relationships, built using a combination of domain expertise and algorithmic approaches and therefore generalize better
- Models that go beyond predictions, allow “what-if” scenario analysis and recommend a set of actions to optimize their marketing spend
- Business-friendly decision applications that allow traditional users to interrogate and interact with the models.
Results and Benefits
The initial calculations show a total reduction of 5% in annual marketing spend, while maintaining the same amount of new installations, returners and paying users. This led to a projected return on investment in decisionOS, causaLens’ data science platform, of 15x. This was achieved by improving marketing outcomes by identifying combination of channels that best drove their internal KPIs allowing them to allocate their budget more efficiently
Using decisionOS, the client is able to understand its customers on a more granular level by understanding the effects of:
• different platforms, such as iOS and Android, and
• different channels, like social media or digital.
These lead to more tailored campaigns that reduce overall spending and improve the ad experience for customers.
