Marketing measurement is broken, and everyone knows it. Every week, I talk to marketing leaders who are desperate to quantify the revenue that marketing is incrementally delivering to the business. Many of them have some kind of marketing mix model (MMM) in place, yet still rely almost exclusively on last-touch attribution (LTA) to optimize their marketing investments on a daily, weekly and monthly basis.
What gives?
First off, can we agree that last-touch is an outdated, insufficient way to measure the effectiveness of your marketing if you’re doing anything but paid search? I hear a lot of, “I know we shouldn’t rely on last-touch anymore, but …”
What follows are the most common reasons I hear – as well as my rebuttal.
“MMM is too slow and expensive”
Agreed. MMM used to be painfully slow. But that’s not a technology problem; it’s a people and process problem. If your MMM still takes months, it’s likely because manual handoffs are bogging things down: data prep, ingestion, exploratory analysis, modeling, tuning, interpretation, planning. Each stage requires a different resource, and those roles are tough (and expensive) to staff.
AI has changed the game. With automation across the workflow, MMM can now deliver insights in hours, not quarters. Speed and flexibility are finally within reach, but only if your tools are built to support end-to-end workflows and can scale with your data.
“No one trusts our MMM”
Trust issues with MMM are real—and sadly, often earned. There are horror stories: models tweaked to fit expectations, missing or smoothed-out outcomes, black-box inputs hidden behind buzzwords and diagnostic metrics withheld entirely.
That kind of opacity undermines the entire discipline. Without transparency – clear inputs, accessible diagnostics and traceable outcomes – your measurement will never influence decisions. If you’re not building trust to drive decisions, you’re just building fancy reports to collect dust.
AdExchanger Daily
Get our editors’ roundup delivered to your inbox every weekday.
Daily Roundup
“It’s too hard to change our measurement”
The change management problem is real – and worth solving. We’ve seen firsthand how challenging it is to shift from surface-level metrics to truly incremental insights.
Nearly a decade ago, our own marketing team faced challenges in predicting what outcomes finance should expect from marketing efforts. We were relying on a combination of publisher-reported performance, channel-specific partner metrics and last-touch attribution from our internal analytics.
We had a 1+1 = 4 problem. The sum of the reported results from these fragmented sources didn’t match what finance was measuring.
We knew we had to get to a view of incrementality and that it would not be painless.
Fortunately, we had strong executive backing to pursue more sophisticated measurement, which gave us the foundation to engage in tough but constructive conversations with our media teams and agency partners across channels. By showing side-by-side comparisons of how our measurement would evolve – grounded in two years of rescored historical data – we made clear how our approach to strategy and execution might need to dramatically change.
Some channels (and you can probably guess which) looked amazing through a last-touch lens but faltered when measured for actual incremental impact across the funnel.
Measurement has to drive decision-making
Here’s a crazy idea: If you trust your measurement, let your partners try to game it. That’s how you’ll know it’s working.
Because great measurement isn’t about pretty curves or perfect model fit. It’s about real-world impact. If your results look stellar but nothing changes in business performance, something’s off.
That’s the mindset we push our customers toward: progress over perfection. You don’t need to wait for 99% statistical confidence to act; you just need confidence that your measurement is good enough to learn, improve and iterate. And for that to happen, your culture has to reward curiosity, not punish being wrong. That’s the real unlock for modern marketing teams.
If all of these challenges sound familiar, you’re not alone. But if you’re still relying on last-touch, ask yourself: Is it guiding smart decisions – or just easy ones? Yes, change is hard. But so is wasting millions on marketing you can’t prove worked. In a recent webinar, H&R Block shared why they’re investing in moving beyond last-touch – and how they’re making the shift work.
We’ve worked through these same challenges at Adobe, and it’s why we’ve built our own MMM solution natively on Adobe Experience Platform with automation, transparency and decision-readiness at the core. Marketers need faster answers and more actionable insights, and we believe Adobe Mix Modeler can finally deliver both.
The future belongs to marketers who measure boldly and act accordingly.
