Digital Advertising

Criteo Enhances its Predictive Engine

unnamed-10 Jonathan Wolf is currently the Chief Product Officer at Criteo, a performance advertising company that works with Internet retailers to serve personalized online display advertisements to consumers that have previously visited the advertiser’s website. Prior to Criteo, Wolf was at Yahoo where he oversaw the restructuring and sales to Europe. The Makegood recently spoke with Wolf about Criteo’s predictive engine.

The Makegood: Congratulations on the enhancements to the Criteo engine. Could you elaborate on what the engine does?

Thank you! The Criteo Engine is our proprietary prediction and recommendation technology. Basically, it allows us to far exceed previous performance benchmarks by deciding whether to buy each ad impression based on the likelihood that a user will both click on an ad and then purchase on the advertiser’s site. This improved ability to predict conversions is a major advance over our previous engine.

This enhancement creates the opportunity for our clients to expand their reach to many more potential buyers all within their previous cost-per-sales target.

The Makegood: What kind of features does the improved engine include, and what can they do for the company and its clients?

The engine strengthens the overall performance of the platform with a 38 percent increase in sales at exactly the same cost of sale. This 38 percent lift has been measured across billions of impressions, and millions of users and clicks.

The new engine operates at lightning-fast speeds. It can process up to 15 million predictions per second and respond within 20 milliseconds. We’re now able to accurately predict when a conversion will occur. A conversion generally happens about once in every 10,000 impressions, so this capability is huge for our customers.

The new engine also introduces several new capabilities for our clients, like simplified segmentation designed to streamline the workflow for planning and management, making it easier for customers to focus on their marketing objectives.

The Makegood: How does this engine give Criteo an edge over others in the industry?

Criteo is aware of a large number of head-to-head tests that our advertisers have run this year comparing us with competing solutions. We have won 90% of those tests.

Now that the engine is performing with a 38 percent increase in sales at exactly the same cost of sale as before, we expect an even bigger edge over others in the industry.

The Makegood: Why was an engine like this a necessity for the industry? How does the engine so successfully predict behavior?

In a competitive industry like ours, customers are constantly looking for improvements in reach, accuracy and ROI. This update provides significant advances in all three, and helps to push the performance marketing industry towards its full potential and away from more traditional, less accurate options.

The engine is capable of predicting behavior so accurately by analyzing many weeks of data – about 20 terabytes – every few hours. We’re harnessing the power of Apache Hadoop to do so, and are contributing to continued expansion of the platform.

The Makegood: What do you see for the future of this engine? Do you see it improving again as drastically as it just has?

I see it as one very important component of our overall mission, which is to constantly improve our ability to help marketers and clients recapture lost value by directly reaching their most valuable consumers. Technology is at the heart of everything we do, and we’re always working to make the small improvements that eventually culminate into the really cool, big updates like this one. I see our engine getting better every single day and our customers should expect to see that reflected in their results.

The Makegood: Thank you, Jonathan.