February 9, 2026
<BackBy Joe Nguyen, Senior Strategic Advisor, H+
Let's be honest, marketing measurement has always been “a tale of two cities”. For decades, we've had the old guard and the new (digital) wave, each with its own language, metrics, and tools. On one side, the broad-stroke world of traditional media, ruled by strategic, top-down models. On the other, the granular, hyper-tracked universe of digital, driven by real-time optimization. The two rarely spoke the same language, and as a result, a massive blind spot emerged right in the middle of our marketing funnels.
But a tectonic shift is underway. With the surge in available data and the rise of advanced AI, the wall between these worlds is finally starting to come down. Advanced marketers are no longer choosing one approach over the other—they're building a unified framework that gets the best of both.
Think of it this way: Marketing Mix Modeling (MMM) is your satellite view. It shows you the entire landscape and how all your business drivers, both inside and outside of marketing, influence sales.
This top-down approach analyzes aggregated data—your total sales, your overall ad spend on TV, radio, and print, even external factors like seasonality and competitor activity. It's not concerned with individual clicks or users; it's focused on the big picture.
MMM’s power lies in its ability to decompose total sales into two key components: base sales (the natural demand driven by things like pricing, distribution, and seasonality) and incremental sales (the additional sales driven by your marketing and promotional activities). This decomposition allows you to answer strategic questions like, "What's my total return on ad spend across all channels ?" and "If I increase my TV spend by 10%, what will the impact be on my bottom line ?"
Historically, MMM has been the go-to for measuring traditional marketing because it can capture the true "halo effect" of brand advertising—that broad lift in awareness and demand you get from a TV spot that you simply can't track with a pixel. This makes it aprivacy-friendly, future-proof solution that is unaffected by cookie deprecation and data privacy regulations.
If MMM is your satellite view, then Attribution is the GPS on your dashboard. It gives you the turn-by-turn directions for individual customer journeys.
This bottom-up approach traces every digital touchpoint a customer interacts with—from a social media ad to a search query to a final email—and assigns credit to each one. This level of granularity is what allows you to make rapid, tactical decisions: "Which ad creative is driving the most clicks right now ?"or "What keywords should I double down on ?" Its value lies in its directness and its focus on the mid to lower-funnel conversion events.
Over the years,we've seen various types of attribution models: last-touch, which gives all the credit to the final interaction; first-touch, which credits the initial touchpoint; and more advanced multi-touch attribution (MTA) models that distribute credit across the entire journey. While these models offer a micro-level view perfect for day-to-day optimization, their reliance on individual user tracking makes them increasingly vulnerable in our privacy-first world.
But there's the catch: it's also highly dependent on cookies and user IDs, making it increasingly vulnerable in our privacy-first world.
We can also think of these methods as looking at the "forest" (MMM) and the "tree" (Attribution),but what about the person walking through the woods ? Ultimately, our goal isn't just to optimize a budget or a click; it's to connect with the consumer—or as our colleagues at Hakuhodo call it - sei-katsu-sha.
The sei-katsu-sha philosophy is about seeing people not just as consumers who buy things, but as whole human beings with lives, aspirations, and deep complexities.
● MMM gives you a strategic view of the broad market influencing this human being (e.g., "Our TV ad raised brand recall").
● Attribution gives you a tactical view of the digital actions of this human being (e.g., "They clicked the search ad").
However, the true 'Sei-katsu-sha Insight'—the kind developed by HILL ASEAN—requires linking these two. That person who saw your TV ad might not have clicked a banner immediately, but they might have searched for your brand two days later and eventually converted in-store. If your measurement system is siloed, you miss the crucial link between the strategic, upper-funnel influence and the lower-funnel action.
This deeper, human-centric understanding is what allows us to move beyond simple transactions and develop the scalable products and offerings that companies like H+ bring to market. To effectively productize the Seikatsusha insight, you need a measurement framework that captures their entire, fragmented journey—not just the tidy digital trail. This is the ultimate "why" behind unified measurement.
The old way of operating in silos is over. The power of modern data science and AI is allowing us to build a single, comprehensive measurement solution. Marketers no longer have to choose between a strategic, holistic view and tactical, real-time insights.
This convergence is made possible by a few key developments:
This isn't just theory—it's already being implemented. A prime example is Google Meridian, an open-source, Bayesian MMM framework that's built for this exact purpose. It's designed to provide a modern, privacy-safe way to measure the true impact of your ad spend across all channels.
Meridian goes beyond traditional MMM by incorporating modern data like Google Query Volume and metrics like Reach & Frequency for YouTube. The platform's Bayesian approach focuses on understanding causation, allowing marketers to answer "what if" questions with more confidence, such as "What would our ROI have been if we had spent 20% more on this channel?" This gives a much clearer picture of how a brand's upper-funnel efforts translate into demand. And as an open-source tool, it empowers internal data science teams to build and customize their own measurement frameworks rather than relying on black-box solutions.
To make this technology accessible, Google has partnered with leading agencies, like Hakuhodo DY ONE. By integrating Meridian into its own solutions, Hakuhodo can now help brands combine their deep expertise in TV data with Google's insights to optimize ad exposure frequency and conduct high-precision analysis on a regional level. This is the future of measurement in action.
To make this all work, you need to understand the two core types of AI at play.
● Predictive AI (Correlation): This is all about forecasting what will happen. It helps you answer questions like, "Which users are most likely to convert?" or "What's the optimal ad bid?" It's the engine of real-time optimization and the lifeblood of most modern attribution models.
● Causal AI (Causation): This is about understanding why something happened. It isolates the true, incremental impact of your marketing spend. It helps you answer the "what-if" questions that drive strategic decisions.
The best-in-class marketers are using a blend of both: predictive AI for the day-to-day tactical wins and causal AI—the engine behind modern MMM—for long-term, strategic growth.
The wall between marketing's two “cities” has been a barrier for too long. It's time to tear it down and build a unified framework that gives you acomplete, actionable picture of your marketing's true power.
All inquiries will be forwarded to and responded to by the administrator of this website or our nearest representatives from the H+ offices listed above.