Jones Road Beauty Is Using A New Type Of MMM To Reset Its Media Measurement

Why It Matters

Grounding media‑mix decisions in causal experiments lets marketers eliminate waste, boost ROI, and gain confidence in high‑spend digital allocations.

Key Takeaways

• Causal MMM uses experiment lift as model anchors
• Tests revealed Meta overspend despite incremental returns
• Branded search proved non‑incremental, saving $3k daily
• Spend recommendations become next planned experiments
• Multicollinearity issues undermine traditional MMM accuracy

Pulse Analysis

Traditional media‑mix modeling relies on historical correlations, which can be distorted by multicollinearity—when multiple channels rise together during seasonal peaks. This statistical noise makes it difficult to isolate each channel’s true contribution, leading marketers to trust opaque recommendations that may not reflect reality. Causal MMM addresses the problem by inserting real‑world experiment results directly into the model, turning lift measurements into concrete data points that anchor the entire mix. The result is a more transparent, testable framework that reduces reliance on speculative assumptions.

Jones Road Beauty’s partnership with Haus illustrates the practical upside of this methodology. The brand ran controlled spend experiments on Google’s Demand Gen product at $7,000 and $10,000 daily, generating a lift curve that informed a calibrated recommendation to increase spend to $15,000. Parallel tests on Meta’s Advantage+ Shopping campaigns revealed that, while still incremental, the channel was overspending beyond the efficiency point, prompting a pullback. A separate test showed that $3,000 a day on branded search delivered no incremental revenue, freeing that budget for higher‑performing tactics. Each insight directly fed the causal MMM, turning recommendations into the next experiment rather than a blind bet.

For the broader advertising ecosystem, causal MMM signals a shift toward measurement that blends the scalability of MMM with the rigor of incrementality testing. Brands can now build a living roadmap where every budget adjustment is validated by prior experiment data, reducing waste and accelerating learning cycles. As AI‑driven planning tools mature, integrating causal experiment data will become a competitive differentiator, enabling marketers to justify spend in real time and adapt swiftly to market dynamics. Early adopters like Jones Road are setting a template that could redefine how the industry quantifies media effectiveness.