Math: The Frontier of Attribution with Michael Kaushansky
Michael Kaushansky, former President of Havas Helia, on his path from mathematics to measurement and what holds companies back from adopting attribution.
Accurate measurement tells a business what its marketing is doing to consumer behavior and sales, and few people have spent more time on that problem than Michael Kaushansky. He is the former Chief Data Officer and President of Havas Helia, and at the time of this interview was on the Decision Science team at Walmart Connect.
In this podcast interview, Michael talks about how he came to measurement, what his team built, and what keeps businesses from adopting attribution.
From mathematics to measurement
Michael's path into measurement began with mathematics. He holds a bachelor's degree in mathematics and a master's in applied mathematics and operational research, and he started his career at GE, Target and GlaxoSmithKline, working on data analytics and business measurement.
His move into digital advertising came after the 2008 financial crisis. He saw that the data sets available on the web, covering ads, searches and consumer behavior, opened up a new way to measure what marketing was doing.
Building path-to-purchase tracking
At Havas Helia, Michael and his team of mathematicians built what became known as path-to-purchase tracking. The approach credits marketing activities for how well they move consumers through the sales funnel.
When a business can see which touchpoints actually advance people toward a purchase, it can allocate budget more accurately, make decisions faster and adjust strategy as results come in.
Michael also pointed out that consumers spend more and more time on their devices, even when they end up buying in a physical store. That makes digital measurement important for understanding offline sales and the marketing that drives them, not just online ones.
Three barriers to adoption
Despite its value, attribution isn't universally used. Michael named three main barriers:
- Complexity. Advanced analytics and attribution can seem daunting to businesses that don't work with data analysis every day.
- Lack of awareness. Many teams don't know enough about what attribution can do for them.
- Fear of change. Concern about failure, and plain resistance to new ways of working, keep businesses on familiar measurement practices.
How to get past them
Michael's advice was to be proactive. He encouraged advertisers and agencies to ask for sample data sets and test different attribution models to see how each one performs in their own business.
He also noted that large companies were increasingly adopting attribution, and that new practices tend to spread in a familiar progression: the big players move first, and the rest of the market follows once the value is proven.
Why it still matters
Michael's career shows how much of modern marketing measurement comes down to applied mathematics, and how much of the challenge is organizational rather than technical. The math to credit marketing more accurately exists. The harder step is getting companies to test it, trust it and change how they make budget decisions.
Frequently asked questions
Who is Michael Kaushansky?
He is a measurement and data analytics leader with degrees in mathematics and applied mathematics. He was Chief Data Officer and President of Havas Helia, and at the time of this interview was on the Decision Science team at Walmart Connect.
What is path-to-purchase tracking?
It is an approach Kaushansky's team at Havas Helia built to credit marketing activities for how well they move consumers through the sales funnel, so budgets can follow the touchpoints that actually advance a sale.
What stops companies from adopting attribution?
Kaushansky named three barriers. Advanced analytics can feel too complex, many teams don't know enough about its benefits, and fear of failure makes organizations resist changing how they measure.
How can a business get started with attribution?
Kaushansky suggested asking for sample data sets and testing different attribution models to see how they perform in your own business before committing to one.