Marketing Mix Modeling: What Coty, Nestlé, and Pierre & Vacances – Center Parcs do to steer their marketing performance

In a constantly shifting marketing environment, Marketing Mix Modeling (MMM) is becoming a key transformation lever to optimize investments, measure performance effectively, and demonstrate the added value of each action. More than an analytical tool, it acts as a real strategic support for marketing management.

A round table brought together Ekimetrics’ teams and the Coty, Nestlé, and Pierre & Vacances-Center Parcs groups, where an insightful exchange showed how implementing an MMM program enables the breaking down of silos, the standardization of marketing decisions, and the streamlining of resource allocation, and all of that by better coordinating the key functions and upskilling the teams on performance interpretation.

The feedback from these three companies, at different stages of deployment, reflects a common ambition: improving marketing effectiveness. What emerges from this is that MMM is not a rigid approach, but a dynamic process that progresses with the data and organizational maturity of the company, and can enable the steering of all marketing mix levers (media, promotions, pricing, etc.) into a unified vision that informs strategic decisions.

THE WINNING PRACTICES OF MARKETING MIX MODELING

Why invest in an MMM program today?

1. To structure marketing decisions around a common language, across all key functions (marketing, finance, media, sales, data)
2. To prove marketing’s value to general and financial management and control budget allocation with robust data
3. To break down organizational silos and rally teams together around common performance KPIs
4. To adapt more quickly in the face of market uncertainty, thanks to simulation and optimization tools
5. To accelerate data appropriation across the company and develop an approach based on data analysis to move away from steering “by intuition.”
6. To evolve the marketing function towards a sustainable performance culture, by integrating MMM in structural moments: plans, reviews, appraisals, RFPs, with, on average, +15% to +30% of annual ROI or incremental growth

How to successfully launch your MMM project? The example of Pierre & Vacances-Center Parcs for building the foundations of a sustainable program

At the start of the project in 2023, an assessment was drawn up by Pierre & Vacances-Center Parcs, a key player in the tourism and vacation home sector in Europe: Investment decisions were often based on the different intuitions of siloed teams—between branding and performance, digital and offline.

In this context, Carole Rey, Head of Acquisition, decided to launch a comprehensive MMM program. From the start, the message was clear: Get stakeholders fully on board, from operational teams to C-levels, including data and marketing teams. The digital data was cross-referenced with that of offline actions, from promotions to pricing. Over the course of the project, certain variables, such as occupancy rate—particularly strategic in the hospitality sector—were integrated progressively. This is the whole point of an MMM program conceived as an evolving lever: the modeling becomes richer with each cycle by integrating new settings to stay as close as possible to operational reality.

Beyond the technical implementation, it is a governance choice that shapes adoption: integrating all teams in the project’s development.

“When I arrived, there were a lot of trade-offs made by intuition, with little clarity on the actual performance. MMM transformed our way of steering. Today, we can anticipate our trade-offs, draw up more solid reviews, and above all, align people around a common plan. What’s made the difference is the collective involvement of the teams from the outset. Everyone contributed to the scoping. And, as we very quickly integrated non-media variables, the models were credible from the start. MMM brings serenity to our trade-offs. We can test, be more daring, because we know we’ll be able to measure, understand, and adjust afterwards.” Carole Rey, Head of Acquisition, Pierre & Vacances-Center Parcs Group

Today, MMM is used to:

• Allocate budgets at the start of the year thanks to scenarios
• Adjust plans as they go
• Carry out more cross-functional campaign reviews
• Challenge preconceived ideas about under- or over-estimated levers

Key learnings: A successful launch is based on three pillars:

1. Aligning key sponsors from the start
2. A holistic approach to contribution modes (on/off, branding/performance)
3. A central role in the adoption for operational teams.

Learn more about the Pierre & Vacances-Center Parcs business case: How Marketing Mix Modeling transformed the decision-making process at Pierre & Vacances-Center Parcs.

6 questions to ask before launching an MMM project

1. What business objective am I trying to achieve? → Optimizing budget allocation? Increasing top-line growth? Gaining market share? Better trade-offs between channels and brands?
   Before discussing data, the strategic aim must guide the scope of the MMM project.
2. For whom and for which investments is the MMM program designed? → For marketing management, financial management, the brand manager, the media planner? For digital, promotions, TV, distribution?
   Identifying the business use cases and the trade-offs to be quantified is essential for organizing the construction of the models.
3. Do I have access to reliable, structured, and exploitable data for modeling the key levers (media, promotions, pricing, distribution, etc.)? → And am I prepared to invest in structuring them if this is not already the case?
4. Who will sponsor the program? → Which C-level profile (marketing, finance, sales, data) will carry out the process and justify it to the other sponsors?
5. How to involve the key functions in the development of the program? → Data, marketing, media, finance… MMM works when it is designed collaboratively and not seen as a top-down initiative.
6. How to bring out a sustainable dynamic and avoid a ‘one-shot’ program? → Integrate the MMM deliverables into key business moments: budget, re-forecast, reviews, RFPs. And make MMM a continuous lever for learning and adjustment.

How to fit MMM into the marketing decision-making process? How Coty built bridges between global strategy and local tactics

At Coty, a key player in the beauty industry, the complexity is structural: the company operates in two divisions (consumer beauty and prestige), has nearly 30 brands, and serves markets with very different maturity levels.

The challenge was not to produce a model, but to ensure its organizational adoption. To respond to the challenge, Coty created an interdepartmental “MMM cockpit” bringing together the data, media, finance, transformation, and business teams. This cockpit meets weekly to coordinate the updates, resolve friction points, and adapt the analyses to strategic needs.

“MMM is a guarantee of consistency and credibility. It allows us to secure our trade-offs and anchor each decision into a shared base”, emphasizes Alexandra Selly, Global Senior Director Data, Analytics & Measurement. “Thanks to MMM, we all started speaking the same language. It has allowed us to ask the same questions, use the same definitions, and use the same figures at the right moment. Whether the local or the global market, we look at performance through the same lens”.

Today, the internal teams present the MMM results at strategic committee meetings. Coty has even put in place:

• A unified definition of levers between countries (TV, social, promotions, retail media, influencers, CTV, etc.),
• A unified database
• Presentations adapted to maturity level (simple graph, customized slides, impact newsletter, etc.)

“MMM’s asset has been bringing everyone around the same table. Finance, sales, marketing, global, and local. It’s a cross-functional exercise that encourages dialogue. And today, in our boards, we no longer discuss figures or data, but the decisions to be made. That changes everything.” explains Alice Foglierini, Digital Transformation Program Director.

Key learning: When fully integrated into the process, MMM becomes a governance tool: planned in scheduling, used during trade-off and scoping meetings, understood by everyone, and designed with unified data.

The 4 key stages of a successful MMM

1. Set a clear and shared ambition, with a sustainable vision. Involve C-level sponsors from the first discussions to ensure strategic alignment. It is essential to gather the marketing, data, finance, and sales teams around the table right from the start.
2. Collect, clean, and structure the data in a common format. Alignment on definitions (TV, social, retail media, promotions, stocks, etc.) creates a shared language that is essential for harmonization.
3. Design an initial pilot model integrating several levers—online, offline, promotions, and pricing, for example—as well as internal variables (stocks, offers) and external variables (such as seasonality, competition, market trends), to measure business indicators like ROI, sales, or attribution.
4. Deploy at scale: More than an analytical tool, MMM becomes a steering pillar integrated into performance reviews, tenders, budget screens, or annual plans. Automation becomes key to ensure continued use.

How to deploy an MMM at the scale of an international group? How Nestlé succeeded in industrializing its media measurement in several countries and for several brands and categories.

At Nestlé, a worldwide food industry group, a unified, industrialized, European-scale MMM program was launched four years ago, starting with France, to model the effects of media investments on business performance.

“We wanted to validate the relevance of the solution on one market and then deploy it at a European scale,” explains Charles-Antoine Dreyfus, Senior Product Manager, Zone Europe Data Team. Today, our media and marketing teams in Europe have an analytical platform powered by MMMs, allowing them to analyze the performance of their media investments. The long-term ambition is to go beyond media and integrate other levers, such as pricing, promotions or assortment, into the analysis process. Moving from “tactical decision-making” to “strategic trade-offs” in a way. This provides a holistic view of performance.”

But scaling entails overcoming several challenges:

• Ensuring reliability of data access by guaranteeing the definition of a common reference framework to standardize the granularity and the protocol of sales and media data collection. This work must be combined with the creation of tools to centralize data and ensure a sufficient level of quality and frequent updates.
• Finding the right balance between a common, standardized approach and consideration of local characteristics such as temperature for water, the calendar of public holidays for chocolate, and inflation in countries like Turkey.
• Developing a modular approach to meet different needs, between granularity and strategic altitude, and covering market, category, and product specialties at the same time, allowing for a holistic and coordinated understanding of the issues.

“What we’ve developed around MMM requires more reliable, industrialized data collection and provision. This is what now allows us to accelerate the provision of results, and to provide concrete input for our business discussions.”

Today, the program is deployed in five markets and around fifteen categories in an industrialized approach. About a hundred end users can access results from a centralized tool to cross-reference insights from various media levers and steer their performance marketing.

A dual governance plan has been put in place to orchestrate everything:

• A central scoping that defines the benchmarks, roles, data flow, and analytical tools
• Localized steering with representatives trained in processing results and adapting the models according to their market context

Key learning: Integrating other levers (pricing, promotions, assortment) enables centralization of the analyticzal work and the breaking down of strategic trade-off silos.

What MMM leaders have in common: Key success factors and lessons to learn

Three requirements come back systematically throughout these business cases

1. Create a common language between professions and make internal alignment a compass: The most transformative programs are those co-developed with the field, media, finance, marketing, and data teams. Work with them, not for them, right from the start.
2. Anchor insights in the key decision-making moments to encourage adoption: The more MMM is used in the key business events (quarterly agile steering, 1-year plan, 3-year plan), the more it spreads as a governance tool, beyond marketing.
3. Support a progressive but structured transformation:
   - Do not underestimate the effort of collecting and structuring data at the outset: one of the most structuring investments, but also one of the most complicated. Automation is essential in the medium term for a smooth adoption.
   - Prioritize simplicity in the deliverables: making access available to all thanks to decision-making tools that are understandable, personalized, and compatible with the teams’ different maturity levels (from the media planner to the CMO).
   - Take small steps with a clear goal: The starting point does not have to be perfect. Start with a priority scope, measure the initial effects, then broaden.

The tangible benefits of a successful MMM deployment

• Better budget allocation: The company optimizes its budget allocation, with the possibility of simulating and adjusting plans according to business objectives. The trade-offs are quantified and shared between departments. On average, companies observe between 15% and 30% ROI efficiency gains and up to +2% to +3% incremental topline growth.
• Team maturity: The MMM program acts like an upskilling accelerator for marketing, media, and data teams. Each presentation becomes a learning tool to understand the real performance levers.
• Breaking down organizational silos: Marketing decisions are coordinated between the local and global sales, data, and media teams… Coordination is strengthened—resulting in improved overall effectiveness.
• Business cadence in line with data: MMM enables adaptation of the decision-making pace to market developments (seasonality, stock shortages, push media, channel mix changes, etc.) The program gives access to actionable insights every quarter.
• Long- and short-term vision: 3-month models (short-term ROI) are supplemented by 18-month models, allowing steering of brand equity or deferred sales. Useful for long cycles (vacations, food, cosmetics, etc.) where conversion is not immediate.
• Operational time-saving: Once structured, the data becomes a valuable production tool and not an irritant. Time saved with automation enables concentration on the essential: trade-offs.
• Performance culture: MMM installs a continuous improvement mindset at the heart of the process. It equips the plans, reviews, and RFPs, and becomes a structuring element of governance marketing.

MMM TRENDS: AI, AUGMENTED STEERING, AND CREATIVITY

Toward an MMM augmented by agentic AI

Agentic artificial intelligence (agentic AI) is transforming the nature of decision-support platforms—and Marketing Mix Modeling is no exception

Unlike previous generations of AI, agentic AI does not merely interpret or generate content on demand; it takes action, stimulating, planning, recommending, and connecting systems that were previously disconnected—based on business goals set by humans.

What does this fundamentally change for MMM?

Firstly, the capacity to connect to multiple sources (internal reports, models, CRM, measurement tools, external benchmarks, etc.) to produce a report or recommendations in a few clicks. Secondly, the agent can break down a complex question (for example: “How can I adjust my plan in case of a budgetary cut of 10% of the marketing budget?”) into concrete actions, while documenting its reasoning. Finally, when coupled with generative AI, agentic AI enables natural interaction with the platform in a common language, making the learnings accessible to everyone, from data scientists to marketing directors.

3 use cases in development at Ekimetrics:

1. Get Insights - from insights to impact: This new agent facilitates access to MMM results and enables:
   - Understanding of the context, audience, and user goal
   - Instant identification of the most relevant insights
   - Result summaries at the right level of granularity
   - Generation of ready-to-use visualizations
   ‍
   The agent puts the MMM within reach—transforming complex data into accessible and actionable results. It saves time, strengthens trust in the data narrative, and upskills with each use.
   ‍
2. Scenario Planning: This intelligent agent generates customized scenarios for a persona (CMO, CFO, e-commerce director, etc.), an objective (e.g., traffic generation, awareness, profitability), and a constraint (budget, channel, seasonality, etc.). Useful in re-forecasting, budget planning, or tactical arbitration phases, it allows teams to:
   - Create and compare several scenarios in a few seconds
   - Visualize the expected impact on the key KPIs of each adjustment
   - Understand the compromises to be made between channels, investments, and results
   
   It puts the power of modeling directly in the hands of business teams: to test hypotheses, quickly adjust plans, and make informed decisions… without losing the link to performance.
   ‍
3. Battle Plan: This module enables moving from scenario to action. It issues recommendations to develop a realistic and actionable operational plan (“How should I allocate my 500 radio GRP?”) by combining MMM insights, external benchmarks, and the accumulated expertise in the platform. This feature is particularly useful for communicating effectively with media agencies or local teams.

A single priority: transparency

Each recommendation is explainable, sourced, and challengeable. The user can follow the reasoning, or even the data source, and request alternatives. No reasoning is operated outside the scope of validated data, and professional safeguards prevent drift or hallucinations.

Agentic AI is not there to replace a job, but to leverage it. As summarized by Thibault Labarre, Deputy General Manager at Ekimetrics: “Agentic AI acts as a strategic copilot. It’s not intended to replace but to improve insights and accelerate decision-making. Eventually, it will significantly increase adoption and the value generated by MMM.”

Collaboration with partners—notably, agencies—remains essential: Agentic AI fits into a value chain. It does not replace it. Better yet, it becomes a catalyst for strengthening exchanges between advertisers and partners by reducing friction in understanding recommendations and aligning on multi-channel initiatives.

When the creativity of advertising videos becomes the best performance lever—an exclusive Ekimetrics x Google study on the impact of creative video execution on sales

In this joint study by Ekimetrics and Google, conducted on four key sectors and two continents (Europe and United States), a key learning emerges: the quality of the creative execution of videos—that is, the way in which the creative elements are arranged in the advertisement (structure, tempo, message)—represents on average 60 % of the impact of a YouTube video on sales—more than traditional drivers such as coverage, frequency, or targeted audience.

The objective of this study is to go beyond creative intuitions and identify the 20 highest priority elements (out of 73 analyzed via Google’s ABCD framework) that actually contribute to business performance.

Key results:

• Directional elements (calls-to-action, product benefits, offers) are the most under-utilized… even though they generate the most impact.
• A gap of x2.2 was observed between the average ROI of a video whose creative execution is judged “unsatisfactory” and that of an advertisement considered “excellent" according to our proprietary scoring system (Creative Execution Score).
• Only 15% of European advertisements are considered “excellent” compared to 42% in the United States: a high potential for optimization.

This study also provides a reproducible methodology for brands:

• Measure the creative effectiveness of videos via granular MMM models
• Classify creations according to actual observed execution (presence/absence of key drivers)
• Steer the quality of execution from the creative conception stage, as with other marketing mix levers.

In short, the creative execution of advertising videos is not a “black box”. It is a lever that is quantifiable, measurable… and optimizable. A business issue as much as a marketing one.

👉 Learn more about our Marketing Mix Modeling solution.