Imagine being so detailed oriented that you play a version of SimCity that is a replica of Toronto. Not just the overall population size but each neighbourhood matches and not just numbers but on average the same number of kids, families and on and on. Imagine the map matches and when you let your Sims loose to travel back and forth to work they get stuck in the same traffic jams you and I do. While someone might use this gaming scenario to see what Toronto would look like if David Miller was never voted out and Transit City was built, it is also a solution to bridging two distinct media measurement systems, panels and census data.
A population of synthetic individuals (essentially what SimCity is) animated with “day in the life” behaviours, matched to the real population offers a strong solution for the challenges of media measurement that current systems cannot. The worlds of analogue and digital media need a common language but neither can extend their ecosystem into the other. Digital’s one to one — track everybody does not mesh with Analogue’s panel-based sampling. Panels will never extend to the whole population, tracking the whole population requires some form of permission. And so we are stuck at GRPs and Impressions when trying to combine television and digital video.
The synthetic population is the bridge. It is like a panel but at full census level, there is one synthetic person for every real person. Instead of using accurate data on an individual to represent many other individuals, the synthetic population is accurate at a small group level, a cohort. The question is how are these small groups defined to ensure the data is accurate, interoperable and rich in depth.
So that future media measurement is set on a solid foundation, the cohort definitions should be public, independent of any specific platform and interoperable with many media types. Also, they should never become obsolete due to changes in tech, legislation or consumer opt-out. The good news is that the Canadian media market has an excellent foundation to build synthetic populations in Government Census data and postal codes. In fact, this work has already been done and now the focus is no longer debating how synthetic populations should be developed, we are onto the implications on the media supply chain of their use.
Marketers are tasked with growing their brands, measurement must deliver insights that generate growth. That means focusing measurement on Incrementality as proof that marketing is delivering growth and Media Mix Modelling for insight into optimal spending to deliver growth. Working backwards from those measurement format end goals into structuring and aligning the planning, buying, delivery and reporting data should be martech’s primary objectives. Again, the quality, sustainability and interoperability of Canadian postal codes as the foundation for a unified media ecosystem is very strong. Synthetic populations built on top of this foundation open up new avenues for insight and analysis that drive brand growth. Set on this course marketers should be able to pull cross-media incrementality reports out of a DSP as easy as it is to receive a clickthrough or cost per acquisition report.