How Rexona optimized its paid media investment with Marketing Mix Modeling

In an increasingly complex, fragmented media ecosystem with growing attribution challenges, brands are seeking more strategic solutions to maximize the return on their advertising investments. It is thru the partnership between Meta and Bunker that Unilever has become a standout success story in the region.

The challenge: media optimization and ROAS maximization

Rexona, one of Unilever’s leading brands in the Costa Rican market, set out to increase the impact of its paid media investment. The objective? Accurately determine how each channel contributed to sales and uncover opportunities to improve the efficiency of your advertising budget.

The solution: a Marketing Mix Modeling approach focused on results.

With the support of Meta’s Marketing Science team and the analytical expertise of Bunker DB’s data science team, a Marketing Mix Modeling (MMM) model was developed that enabled measuring each channel’s contribution to sales and uncovering clear opportunities to optimize media investment. Thus, Rexona gained the confidence to allocate more investment to Reels and Carousel formats on Facebook and Instagram, enabling it to achieve greater efficiency and long-term growth potential across all digital channels.

Case results

Thanks to this joint implementation:

• Rexona identified a potential 36% improvement in ROAS by reallocating its investment to digital media.
• Meta achieved a ROAS 2.9 times higher than the average across all channels.
• The brand gained confidence to continue investing in high-performance formats such as Reels and Carousels.
• A solid foundation was established to improve efficiency and long-term growth.
• This study was recognized and published by Meta as an official success story, available here: Meta Business Case Study - Rexona.

What is Marketing Mix Modeling (MMM)?

MMM is an econometric model that quantifies how different factors (paid media, promotions, seasonality, macroeconomic conditions, etc.) contribute to sales. Based on these models, it is possible to estimate the marginal ROAS of each channel, optimize budget allocation, and make more precise decisions about media strategies.

In times when user-level attribution is weakening (due to cookie blocking and regulatory changes like GDPR or ATT), MMM is gaining prominence because:

• It doesn’t depend on individual data, but on aggregated and anonymized information.
• It allows for strategic decision-making at the macro level (annual budgets, campaign planning).
• It is robust against fragmentation of the customer journey and the emergence of new channels.
  

Why does this case matter to other brands?

Marketing Mix Modeling allows brands to:

• Accurately quantify the impact of each channel.
• Reallocate budgets based on data, not assumptions.
• Make better decisions in your annual media planning.
• Raise the level of marketing conversations to focus on the business and its results.

In simple terms, it’s a tool for doing more with less and with greater certainty.

Key requirements for implementing MMM:

• +18 months of historical data (sales, media investment, pricing data, competition, etc.).
• Contextual information such as promotions, organic activities, and market data.

Conclusion: Marketing based on science, not assumptions.

This case shows how the combination of data, technology, and strategic alliances can drive smarter, more profitable decisions. Bunker, Meta, and Unilever demonstrated that, even in complex contexts, it is possible to take control of media performance and turn analytics into real growth.

Would you like to know how to apply MMM to your brand?