Clocking at 44 minutes, the latest episode of AdExchanger Talks with Criteo CEO Eric Eichmann is our longest yet. Pardon our loquaciousness, but there was just so much to talk about.
Criteo is one of the high flyers in the data-driven marketing space, with a $2.7 billion valuation that’s higher than any other pure-play ad platform stock. The company has accomplished that by focusing purely on data.
“The company didn’t start out thinking about advertising,” according to Eichmann. “The company always started by thinking, ‘How can we use the data that users leave on sites to drive sales?’”
Initially, that founding principle led it to experiment with machine-learning-driven product recommendation. That turned out to be a hard business model. So it transitioned into ads, about which its founders knew so little that they had to Google “ad server.” True story.
The company won those budgets by excelling both at publisher development – basically, seeing more ad inventory than anyone else – and algorithmic optimization. As it ramped up spend on the platform, the latter engineering advantage only became easier to sustain.
“The bigger the data set you have, the faster you can innovate on the algorithm that you build. Machine-learning algorithms are not static things. They evolve over time,” Eichmann said. “The faster we improved the algorithm, the more clients we could get in … and the more data we would have.
“It got to the point where we were far ahead, and the network effects we got both on the publisher side and the advertiser side made it very hard to catch up with us.”
Also in this episode: Retail industry pain, risks to Criteo’s business and the future of last-click.
This episode of AdExchanger Talks is sponsored by Nucleus.