Three-year old Cognitiv on Thursday introduced a deep-learning search optimization platform that uses artificial neural networks to optimize paid-search campaigns based on internal and external triggers.

The product, Cognitiv Search, launches with a feature called WeatherIntent that allows marketers to optimize bids based on current, recent or forecast weather conditions tied to specific locations.

With large amounts of data becoming available to marketers, performance marketing is ready for deep learning, the same technology that powers Siri, Alexa and sell-driving cars, according to Jeremy Fain, CEO and co-founder at Cognitiv, and former head of North America accounts at Rubicon.

The mission to predict consumer behavior across the entire marketing and advertising funnel led company engineers to move from programmatic, video and display to search.

A brand may want to sell shorts when it’s “hot” outside, but what’s “hot” in Florida is different than “hot” in Vermont. Rather than manually changing the bidding, the platform will automatically optimize the bid for that area.

“Google doesn’t give you the same one-to-one experience that’s found in display,” he said. “In Search optimization you’re pulling different levers like time of day and location. You can experiment with audiences as a segment, but cannot tell if the person was shown the ad and clicked to make a purchase because Google doesn’t allow that data to get out.”

Fain said clients were looking for performance metrics, with an optimization and automated tool, so the company built Cognitiv Search.  

Kenneth Hamner, VP of search engine marketing at Edelman, worked with Cognitiv for about nine months to build out the self-service interface in the platform prior to today’s launch.

“We are seeing better click-through rates and targeting based on a number of factors,” he said. Marketers have always been focused on keywords and devices, but now they have the ability to adjust bids around weather, serving copy in the ads and adjusting bids in real-time based on triggers such as sunshine, rain, snow, extreme cold or heat.

Edelman focuses this type of optimization on ads that serve up in metropolitan areas such as Manhattan and Los Angeles.

The plan is to add additional optimization capabilities that automatically change bids based on keywords, devices, times of day, week, month, locations, as well as custom first-party inputs.