Privacy-first marketing measurement means building on data designed for privacy from the start, not patching old data later. This matters because multi-touch attribution relies on user-level data that’s grown incomplete since iOS 14, and no algorithm fixes missing data. Without a privacy-first foundation, brands scramble whenever tracking tightens, while attribution quietly gets less reliable. This video explains how geo-based experiments and an agile mix model replace last-click and multi-touch attribution.
Why Privacy-First Marketing Measurement Couldn’t Just Patch Multi-Touch Attribution
Privacy-first marketing measurement couldn’t just patch multi-touch attribution, because the underlying user-level data multi-touch models depend on was already unreliable, and no new algorithm can fix data with holes in it. In this video, John Wallace, CEO of LiftLab, speaks from direct experience, having built multi-touch attribution algorithms at a previous company, before concluding that the data feeding those models was, in his words, Swiss cheese, with holes that only got bigger over time. That realization is what led to LiftLab: a deliberate effort to build new algorithms on new, privacy-compliant data sets from the ground up.
That foundation paid off when iOS 14 hit. Wallace recalls a nervous call from a customer’s leadership asking how LiftLab would adapt, and being able to say the answer was doing nothing differently, because the system was already built to work with fully privacy-compliant data. The forward-looking piece of that foundation is combining geo-based experiments with an agile mixed model, an approach Wallace says LiftLab was first to bring to market, giving marketers a real alternative to last-click: a way to see incremental revenue without depending on individual user tracking at all.
In this video, you’ll learn:
Why traditional multi-touch attribution models are becoming obsolete
How privacy changes impact existing marketing data quality
The importance of building systems with privacy compliance as a foundation
Why patching flawed data with more algorithms fails to solve measurement issues
How geo-based experiments and mixed models offer a better alternative for marketers
TIME STAMPFull Video Transcript: Privacy-First Marketing Measurement Explained
+–
LiftLab’s CEO John Wallace explains why multi-touch attribution couldn’t be patched for privacy changes, and how LiftLab was built privacy-first from day one.
[0:00 – 0:24] The Limitations of Multi-Touch Attribution
There’s a lot of brands out there that thought that the alternative to last click attribution would be multi-touch attribution, and there was a lot of work that went into that, and in my own case, I designed some algorithms for multi-touch attribution at my last company. My conclusion is that the data going into MTA was problematic, and we like to say, kind of, it’s Swiss cheese, and the holes are getting bigger, and you can’t fix that data with more algorithms.
[0:24 – 0:54] The Origin and Privacy-First Approach of LiftLab
That was the nexus for LiftLab. You know, how do we work on this problem with new data sets and new algorithms? We’ve been working on this for four years, and so about three years ago, one of the bosses’ bosses jumped onto a weekly LiftLab status call. We were kind of surprised, kind of said in a panic voice, “You know, hey, there’s all these changes to iOS 14. How are you ready to handle this? How are you going to?” We got to smile and say, “You know, we’re going to do absolutely nothing. We built this from the ground up to work with completely privacy-compliant data sets.”
[0:54 – 1:31] Future-Proofing With Geo-Based Experiments and Mixed Models
So that changes, and I feel even overdue changes in privacy that need to work their way through the industry, we need to work with data sets that are anticipating them. That’s been our North Star. Our answer to that came out to market with geo-based experiments and augmented that with an agile mix model. We’re first to market to have those two algorithms working on these data sets, and that is our foundation for getting an alternative to last click, right? In our case, incremental revenue into the hands of smart marketers.
Key Lessons: Building Privacy-First Marketing Measurement
The core lesson is that privacy-first measurement has to be a foundation set before a regulation changes, not a patch applied after.
Attribution limitations: Multi-touch attribution data is often compared to Swiss cheese, where increasing holes make it impossible to fix with more algorithms.
Privacy evolution: Anticipating regulatory shifts is essential, and LiftLab chose to build a privacy-first platform rather than reactively adapting to changes.
Innovative methodology: Combining geo-based experiments with an agile mixed model provides a superior, reliable foundation for measuring marketing performance.
Incremental value: The primary goal for modern marketers should be identifying incremental revenue rather than relying on legacy last-click attribution metrics.
