• by Laurie Sullivan , Staff Writer @lauriesullivan, October 27, 2016

A London-based company has developed technology that tells marketers the revenue they're missing by not targeting a specific market segment.

Qubit Wednesday introduced technology that mines data and looks for sets of customers, or audience segments, that are under-monetized. Bud Goswami, head of data science at the company, calls it "opportunity mining," which the company built on top of its Qubit ML engine.

Marketers have predictive analytics, programmatic technology, and tons and tons of data to target experiences. So what? The advertising industry has proved it can collect the data, but few marketers know why they need it or what to do with it. Machine learning without the ability to determine why the brand should target a specific segment remains worthless.

Goswami says opportunity mining gives marketers a reason to target experiences to a specific type of consumer they might not have thought to reach because machines can do what humans cannot. They can troll through petabytes of data to uncover behavioral patterns, so marketers can mine and integrate the insights with demographics and use the combined facts to improve the quality of the experience on their Web sites or target experiences.

Qubit's opportunity mining technology compares the observed, the predicted and the difference in statistics like conversion rates, revenue per visitor and bounce rates.

The knowledge lets marketers target consumer segments with a specific experience such as recommendations for products or abandonment recovery.

The technology, which becomes available in November, took about a year to build, Goswami said. The challenge has been to scale the product for clients as big as Amazon, he said.