Google ramps up machine learning to automatically optimize search, other ad formats
- Google is unveiling responsive search ads that are powered by machine learning to optimize creative assets in real-time so searchers see the best-performing ad for their query, the company announced via a blog post and on stage at yesterday's Google Marketing Live event. Advertisers using machine learning to test combinations of headlines and descriptions for text ads experienced about 15% more clicks, according to internal data.
- Another new feature is Maximum Lift for YouTube, which is a machine-learning smart-bidding strategy still in beta that allows marketers to reach viewers most likely to consider their brand after seeing a video ad. The tool automatically adjusts bids at auction to maximize the video’s impact on brand perception through a customer journey.
- Google is also introducing local campaigns, designed to drive in-store visits. Businesses provide their location and ad creative and machine learning optimizes the ad across search, YouTube, maps, websites and apps. Google is improving Smart Shopping campaigns to include store visits and new customers as goals, and machine learning will adjust bids to meet these goals. Machine learning will also be used to optimize where Shopping ads appear and which products are featured. Marketers will later be able to set up and manage Smart Shopping campaigns from Shopify and Google Ads.
The latest machine learning tools are part of Google’s broader strategy to incorporate more automation into its ad business. Optimizing for digital marketing has always been a mostly manual challenge, one that has become even more daunting as mobile has grown, bringing more possible variables like time of day and type of device. By automating much of these processes through machine learning, Google hopes to simplify campaign management so that marketers can deliver relevant experiences at scale, which could drive better results. At the same time, machine learning can free up marketers to focus more on strategy.
Some marketers see machine learning and related technology, like artificial intelligence (AI), as a solution to better target campaigns with relevant experiences and measure campaign effectiveness. Google’s internal data reveal promising results in digital engagement and in-store traffic for brands using its machine learning tools, including a 15% lift in clicks. The company also noted in its blog post that mobile searches for “near me” have grown three times in the past two years, and about 80% percent of consumers will visit a store if there’s something they want to buy immediately.
Google's new machine learning-powered tools arrive at a time when marketers are considering ways to deliver higher-quality content that can reach consumers across platforms, especially on mobile. Consumers are using their smartphones more than ever to search and shop for products. Forrester predicts that more than one-third of retail sales in 2018 will involve smartphones, according to AdWeek. This means marketers often have their work cut out for them to deliver the right content at the right time on the right platform to reach these consumers.
In February, Google launched AdSense Auto ads that use machine learning to make placement and monetization decisions for publishers. The ads are optimized to display only when they will most likely perform best and provide a good user experience. Google also announced a new automated process for AdWords that uses AI and related technology to make suggestions to optimize ad campaigns. Suggestions will be based on prior campaigns, including from ad headlines, descriptions and other relevant information.