Our latest webinar brought together some of the brightest minds in marketing measurement to talk about the new era of Marketing Mix Modeling (MMM) and how it’s evolving fast to meet the challenges of a cookieless world.
The session featured Dr. Ramla Jarrar, Co-founder and CEO of MASS Analytics, Alice Sylvester, Partner at Sequent Partners, and Dr. Firas Jabloun, expert in data analytics and AI. Together, they unpacked what’s next for MMM – from granularity and segmentation to competition, privacy, and even the rise of “DIY” modeling.
Why MMM Is Back in the Spotlight
Ramla opened the session by reminding everyone how privacy laws like GDPR and CCPA changed everything. Once cookies started disappearing, many marketers were left wondering how to measure performance effectively without relying on user-level data.
That’s where MMM came back into play. Unlike attribution models, MMM relies on aggregate data, so it’s fully privacy-compliant. And thanks to advances in modeling techniques (like pooled regression, nested models, and log-linear models) MMM is now far more granular, faster, and actionable than ever before.
Ramla also shared her “granularity triangle”:
- Variables – capturing every creative, format, and campaign execution detail.
- Regions – analyzing performance across geographies to understand local differences.
- Customer Segments – recognizing that not all shoppers respond to media the same way.
The Renaissance of MMM
Alice picked up the discussion by talking about how MMM is experiencing a real renaissance. For years, the industry focused on attribution and short-term results. But as privacy tightened and data fragmentation grew, MMM’s long-term, channel-agnostic view became more relevant than ever.
Alice highlighted how today’s MMM can:
- Combine brand health and sentiment data with short-term sales.
- Run faster refreshes — from quarterly to even weekly.
- Deliver granular, segment-level insights that guide smarter decisions.
Real Case Study: How a Retailer Bounced Back After COVID
One of the most interesting parts of the webinar was a case study shared by Alice and Ramla. A retailer wanted to understand if the drop in sales after COVID was permanent or fixable.
MASS Analytics analyzed CRM data and identified two key customer groups:
- High-frequency shoppers, who represented only 8% of customers but 25% of sales.
- Event and holiday shoppers, who made up 26% of customers but only 9% of sales.
It turned out most of the post-COVID decline came from the holiday and event shoppers, heavily impacted by lockdowns and shorter store hours. Competitors who increased ad spend during this time also stole share from these same segments.
With this insight, the retailer was able to refocus promotions, extend store hours, and target the right shoppers — turning around their decline with precise, data-driven actions.
Key Takeaways from the Q&A
The Q&A session was full of valuable questions from the audience. Here are a few that stood out:
- On using different methodologies together: Ramla stressed that MMM shouldn’t work in isolation. Combining MMM with lift studies and attribution can validate findings and provide a fuller picture.
- On collinearity and copy testing: Alice and Ramla agreed that when ads run simultaneously and overlap too much, it’s best to use weighted averages or run controlled experiments to separate effects.
- On the DIY trend: Alice cautioned that while MMM tools are becoming faster and more affordable, running a model without real media or business understanding can lead to major misreads. As Ramla put it, “MMM is 70% business and consumer understanding, 30% statistics.”
- On long-term ROI: The speakers explained that with the right data structure, MMM can measure both short-term sales and long-term brand effects, especially when accounting for lagged or decayed media impact.
Final Thoughts
This session made one thing clear: MMM isn’t old-fashioned. It’s future-proof, privacy-safe, and smarter than ever. As marketing gets more complex and data gets harder to track, MMM is evolving to be faster, more flexible, and more connected to real business questions.
