The Whole Truth About Marketing Mix Models
Management Summary
What does marketing really achieve? And when does advertising become just expensive noise? Marketing mix models provide answers – if you set them up correctly. We explain what is important.
Marketing mix models – the whole truth
Marketing Mix Modeling (MMM) is rightly experiencing a revival. Data protection, cookie restrictions and the increasing pressure for efficiency in marketing departments are forcing companies to reanalyze causal relationships. MMMs measure the impact of advertising investments on sales, conversions and other KPIs – cross-channel, data protection-compliant and model-based.
But as is often the case, the truth lies in the details:
1. Without a baseline there is no understanding
MMMs help determine marketing effectiveness. The baseline is taken into account – the sales that would be achieved even without marketing measures. Factors such as brand awareness, word of mouth and seasonality influence this baseline – i.e. the natural sales flow of a company. MMMs make it possible to statistically calculate the incremental effect of marketing measures on sales.
2. Effectiveness ≠ efficiency
More budget in one channel = more success? Unfortunately no. MMMs show when a channel reaches its saturation point, whether media mix inconsistencies prevent synergies or whether users slip into reactance due to too high a frequency. The key is to recognize where more investment actually brings added value – and where it doesn’t.
3. Not an off-the-shelf solution
Contrary to what many people believe, an MMM is not a plug-and-play tool.
It needs:
- at least one year of structured online and offline data or alternatively (if the database is smaller) empirical values on channel effectiveness, e.g. from other attribution methods or experiments
- Data Ownership & Governance (In this blog article you will find more information about data ownership.)
- Strategic advice on model interpretation
The interpretation of the results and the derivation of recommendations for action are crucial. Tools like Google’s Meridian, Meta’s Robyn or our own solution BREE (Budget Recommendation Engine) help – provided you have the right approach and understanding.
Conclusion: Numbers are just the beginning
Marketing mix modeling can be a real game changer for your media budget planning – provided you approach it holistically. It takes more than just smart models: What is crucial is an interdisciplinary understanding of the objectives as well as the right combination of deep media know-how and sound data science expertise. In order to exploit the full potential, it is therefore essential to rely on strong partners who can support and think along both strategically and technically.
