When it comes to running a Marketing Mix Model (MMM) at your scale, there are two real philosophies I've heard discussed. The first is to go all-in on a sophisticated, top-down MMM as your central source of truth. The second is a more hybrid approach that uses MMM for high-level strategy but combines it with ground-level incrementality testing.
Camp A: The Top-Down Purists
This camp argues that modern marketing is too complex for anything other than a comprehensive MMM. Michael True, CEO of Prescient AI, made a compelling case for this on an episode of a href='/show/limited-supply'>Limited Supply</a>, explaining how MMM has evolved from a backward-looking report card into a forward-looking strategic tool. Proponents like True and Dave Levett, who appeared on the a href='/show/ecommerce-australia'>eCommerce Australia</a> podcast, see it as the only way to escape the trap of last-click ROAS and measure what's actually growing your business.
A pure MMM approach is designed to capture everything that influences sales, not just digital ad clicks. It accounts for TV or out-of-home ads, seasonality, competitor actions, and the 'halo effect' between channels. The idea is that this holistic view allows you to make confident, top-down a href='/topic/budget-allocation'>budget allocation</a> decisions that you can't get from platform-reported numbers. It's about measuring the true, overall business impact rather than getting lost in misleading channel-specific metrics.
Camp B: The Hybrid Pragmatists
This school of thought, which really comes to life in discussions on the a href='/show/ecommerce-playbook'>Ecommerce Playbook</a> podcast, doesn't refute the value of MMM. Instead, it positions it as one part of a larger measurement system. As the hosts explained in their episode a href='/episode/why-perfect-attribution-is-a-lie-and-what-actually-matters'>Why “Perfect Attribution” Is a Lie</a>, you can use MMM for its broad, long-term perspective while using a href='/topic/incrementality-testing'>incrementality testing</a> to get granular, short-term insights. You get the best of both worlds: a strategic compass and a tactical map.
In this model, the MMM might tell you to shift 10% of your budget from paid social to YouTube for the next quarter. That's your high-level strategy. You'd then run a geo-holdout or a lift test on that specific YouTube spend to validate the model's recommendation on the ground. This combination of top-down modeling and bottom-up testing creates what the Playbook hosts call a 'shared reality' for marketing data that is both strategic and empirically verified. It builds trust in the numbers and the decisions that follow.
My Take
I personally land firmly in the hybrid camp, especially for a brand of your size. While a pure MMM is powerful, it can feel like a black box, and its insights can be slow to update. For an eight-figure DTC brand that needs to be agile, the hybrid model is more resilient and practical. You need both the long-term strategic view and the confidence to know your campaigns are working right now.
The real goal here isn't to find 'perfect attribution' because, as many hosts will tell you, it doesn't exist. The goal is to build a system for a href='/topic/data-driven-decision-making'>data-driven decision-making</a> that gives you directional confidence. An integrated approach, using a top-down model and bottom-up testing, is the most robust way to achieve that.
If you're just getting started on this journey, you could begin with a more basic MMM to guide your high-level budgeting. As you grow and your marketing mix becomes more complex, you can layer in more sophisticated a href='/topic/incrementality-testing'>incrementality testing</a>. But if you're already spending across multiple online and offline channels, I'd argue the hybrid model isn't just a 'nice to have.' It's essential for sustainable growth.