MMM & Incrementality

“Half the money I spend on advertising is wasted; the trouble is I don’t know which half.”

This famous quote by John Wanamaker, a successful American merchant and a pioneer of marketing, was spoken over a century ago. Yet, for many companies, the situation hasn’t changed much.

On the contrary, new regulations on privacy and data transfer, coupled with increasing competition, risk exacerbating the problem.

A recent development is the new agreement between the EU and the US on transatlantic data transfers, which came into effect on July 11. This essentially establishes a third Privacy Shield, resolving a contentious issue that had been unresolved for over a year.

While this agreement provides some peace of mind for marketers and digital advertisers running campaigns in international markets—as long as its compliance is not challenged by digital rights activists—the cookieless web and the emphasis on user rights remain central.

How to Manage and Measure Advertising in a Privacy-First World

As users demand greater control and transparency over their data, we are witnessing a significant shift in regulatory landscapes. Frameworks such as GDPR, CCPA, and ePrivacy introduce diverse solutions to protect users and manage data sharing responsibly.

Additionally, browser updates from Safari, Firefox, Edge, and Opera have blocked cookies, limiting measurement protocols.

In this scenario, attribution tools are less accurate, leading to business impacts such as:

• Unclear performance metrics
• Rising CPA (Cost Per Acquisition)
• Difficult-to-analyze ROI

This makes it challenging for marketers to answer two fundamental questions:

• How many conversions would have occurred without any marketing activity?
• How many conversions can truly be attributed to each channel?

Answering such questions would grant full control over marketing activities.

What Is Marketing Mix Modeling (MMM) and How Does It Work?

Marketing Mix Modeling (MMM) offers a breakthrough in this context, allowing businesses to analyze the incremental impact of marketing and non-marketing activities on sales using a privacy-friendly, data-driven approach.

This model calculates the baseline—all sales that would have occurred without marketing—and attributes conversions to specific channels.

The algorithm relies solely on aggregated data, such as impressions, sessions, conversions, costs, and revenue. These first-party data points can be collected without identifying or profiling users, ensuring they remain available despite privacy regulations or institutional decisions.

To account for macroeconomic trends, these data are integrated with metrics related to competition and demand for the product/service, such as average product costs, inflation data, employment rates, or even mobility data from Google Maps. These additional insights proved invaluable during the pandemic.

Robyn’s “Analyst’s Guide to MMM,” an open-source MMM package by Meta, provides a useful metaphor: “Think of Marketing Mix Modeling as reverse-engineering a cake. Imagine ordering a delicious dessert at a restaurant and wanting to replicate it at home. By uncovering the recipe, you can successfully recreate the cake and experiment with ingredients to achieve even better results.”

In essence, MMM helps businesses understand the factors (ingredients) that influence their key performance indicators (the cake) over time, offering flexibility across their strategy.

Key Benefits of MMM: Revenue Contribution and Budget Control

MMM provides two fundamental insights:

1. Revenue contribution: Understanding the impact of each channel on overall revenue.
2. Budget control: Optimizing resource allocation by identifying where to increase or reduce investments.

An End-to-End Solution for Implementing MMM

Navla’s MMM solutions offers rapid execution, a robust first-party data foundation, and predictive capabilities for future scenarios. Tailored to each company’s specific needs, it consists of six main phases:

1. Business Understanding

This phase involves collaboration between our team and the client’s team to define:

• Business objectives
• KPIs to monitor
• Active channels
• Business-impacting variables such as seasonality and promotions

This helps identify the most relevant and accurate input data for statistical and descriptive analyses.

2. Automated Data Pipelines

The second phase involves collecting data from all available sources and centralizing it in a single Data Warehouse. This unlocks critical insights, such as contribution margins, profit analysis, and accurate last-click attribution.

3. 100% Data Integrity

This phase ensures the integrity of raw data by comparing samples directly from the source. Once verified, the process is repeated for aggregated data.

4. Data Exploration & Modeling

During this phase, Navla’s partners platforms explores and learns from the company’s specific data, creating and testing a tailored model.

5. Budget Allocation

Sophisticated optimization algorithms determine the best budget distribution across campaign types to maximize conversions or sales. These algorithms consider various factors such as upcoming holidays, seasonality, and channel saturation points.

6. Centralized Data Visualization

All data are cleaned, organized, and aggregated before being displayed on dashboards that offer a comprehensive view of marketing performance and growth. These insights enable the creation of strategic actions and tactical plans for data activation.

Data Empowerment Through MMM

Navla’s MMM approach can be integrated into a broader Data Empowerment framework, enabling a comprehensive strategy for collecting, enriching, and activating business data.