I love a good nerdy pick for a top 2025 trend. And picking a Media Mix Model (MMM) in “Think With Google” definitely hits the mark. Sure, the rest of the article has the usual suspects, beginning with AI and following on with Ethical Marketing (Yawn 😀). Still, you got me hooked when you lead off with measurement, incrementality, and attribution.

“2025 is the year to master measurement. Start by aligning on KPIs, mapping your media spend, and building an experiments calendar. Measurement is an ongoing process, requiring continuous learning and optimisation to unlock your marketing’s full potential.” Dyana Najdi

At The Home Depot, we had what most closely represented a true MMM, but we never put much faith into its accuracy. At other locations, we used season-long hold-out panels in tandem with traditional A/B testing, each with drawbacks centered around the appropriate duration and statistical significance.

“MMMs have always played a critical role here, helping brands cut through the noise to understand what’s really working in driving business impact,” says Dyana Najdi, managing director of Google Advertising UK and Ireland. In 2025, MMMs that were once thought to be overly complex will be smarter, faster, and more widely available, including “Google’s Meridian — an open-source MMM with access to both Google and YouTube data — is set for global rollout in early 2025 to meet the demand for a transparent and robust measurement solution.”

Here are three reasons why you should try a Media Mix Model in 2025 and three reasons why most marketers will continue putting it off for another year.

Three Reasons for Using a Marketing Mix Model

🟢 Quantify Marketing Impact – MMM helps businesses measure the effectiveness of various marketing channels (TV, digital, radio, etc.) and campaigns. It attributes sales, revenue, or other key outcomes to specific marketing efforts, allowing data-driven decision-making.

🟢 Optimize Marketing Spend – By understanding the ROI of each marketing channel, businesses can reallocate budgets to maximize efficiency. For example, MMM can identify underperforming channels or show where increased investment could yield higher returns.

🟢 Evaluate Long-Term Effects – MMM accounts for delayed or indirect impacts of marketing (e.g., brand building through TV ads). Unlike digital-only models, it captures the influence of traditional channels and how they interact with short-term tactics like promotions.

Three Reasons for NOT Using a Marketing Mix Model

🛑 Complexity and Cost: MMM requires significant resources, advanced statistical expertise, and ongoing analysis and hold-outs with a large opportunity cost. “You mean I need to take 10% of my customer base and stop talking to them?”

🛑 Data Limitations – MMM relies heavily on high-quality, granular data. The results may be inaccurate or misleading if data on marketing spend, sales, and external factors (e.g., seasonality, economic trends) is incomplete or inconsistent.

🛑 Lag in Real-Time Decision-Making – MMM is retrospective, analyzing past data to provide insights. It lacks the real-time responsiveness of digital attribution models, making it less suitable for businesses that need quick insights for agile marketing adjustments.

If you try an MMM in 2025, here are tips on how to integrate it effectively:

1. Establish Clear Objectives – Be very specific when defining what you want to measure.
2. Use MMM Alongside Other Tools – Don’t abandon what you already do. Pair MMM with Multi-Touch Attribution for granular digital insights or incrementality testing for tactical adjustments.
3. Data, Data, Data – Ensure historical data is clean, consistent, and comprehensive across channels.
4. Keep it Simple – Translate MMM outputs into actionable insights for marketing teams and leadership.