Google taps AI to help you bid for digital ads

Real-time bidding is an aspect of digital marketing that can seem overly complex for the average bear, so it was only a matter of time before AI entered the picture. This week, Google brought machine learning into the process to help make it easier.

Tapping some of the same artificial-intelligence technologies that have already appeared in Google Photos and AlphaGo, Smart Bidding is a new capability for conversion-based automated bidding across AdWords and DoubleClick Search to help companies determine their optimal bid for any given campaign or portfolio. It can factor in millions of signals, Google says, and continually refines models of users' conversion performance at different bid levels.

"Smart Bidding's learning capabilities quickly maximize the accuracy of your bidding models to improve how you optimize the long-tail," Anthony Chavez, Google's product management director for search ads, wrote in a blog post explaining the new service. "It evaluates patterns in your campaign structure, landing pages, ad text, product information, keyword phrases, and many more [data points] to identify more relevant similarities across bidding items."

Essentially, Smart Bidding tailors bids to each auction across Google's properties and allows buyers to factor in a wide range of contextual signals, including device and location. Focusing on device performance, for instance, advertisers can set separate cost per acquisition (CPA) goals by device. A telecom advertiser whose best leads come in via mobile, for instance, could set a higher target CPA for that platform compared with other devices.

New reporting features, meanwhile, show companies exactly how their bid strategies are performing and flag any issues requiring attention.

Current users of AdWords Smart Bidding include AliExpress, SurveyMonkey, and Capterra, Google says.

"Google is trying to maximize effectiveness of search marketing as well as its own revenue," said Greg Sterling, vice president for strategy and insights with the Local Search Association. "Bringing machine learning to bear on bidding should advance both objectives."

Google is increasingly in competition with Facebook and other channels, Sterling added, so "it needs to continually invest and improve paid search to keep marketers engaged."