Google Marketing Live 2019 Keynote - Machine Learning to Drive Google Ads Advancement

min read
Google Marketing Live 2019 Keynote - Machine Learning to Drive Google Ads Advancement

At Google Marketing Live 2019 in San Francisco, Google’s chief business officer Philipp Schindler emphasized that the major factor evolving mobile right now is machine learning, a technology that is rapidly improving due to advances in cloud computing, on-device hardware, and new neural network architectures. He also offered this advice to marketers: “Stay close to the evolution of mobile … it will be the #1 platform for consumer engagement for years to come.”

The continued domination of mobile is not exactly a mind-blowing revelation, but the new ways that machine learning (ML) will allow marketers to reach, engage and convert consumers through search campaigns are pretty exciting. “Language processing, computer vision and mobile search are all coming together,” said Schindler. “And we’re bringing these advancements to our [Google Ads] products to make them more relevant for consumers.” Google is also working to make the products “easier to use and higher-performing” for marketers, he said.

Here’s a look at just some of the machine learning-powered advancements that Google unveiled this week at Google Marketing Live 2019:

Discovery Ads: Unexpected but Relevant Results

The ability to connect with consumers right when they’re most open to discovering new products and services is gold for marketers. And research from Google and Ipsos suggests that most consumers (76%) enjoy unexpected finds while they are shopping. Of course, for that journey of discovery to end with a transaction, something unexpected must also be relevant to the consumer.

Google’s new and highly visual Discovery ad format, which will be available to marketers worldwide later this year, uses the “power of intent” to help ensure consumers encounter information about products and services that will be of interest to them. Google determines, with heavy lifting from machine learning, a consumer’s intent by connecting the dots on their past activity, including website visitation and streaming video consumption.

Prabhakar Raghavan, senior vice president, Ads and Commerce, at Google, told the Google Marketing Live audience that Discover ad campaigns will be “simple” for marketers to use. He said when marketers decide to launch a Discover ad campaign, they can upload ad copy and other creative elements along with information about their target audiences, and Google will apply machine learning to develop the best combinations of assets. The native ads appear automatically across multiple Google feeds on Google Discover, YouTube and in the Promotion and Social tabs in Gmail.

Gallery Ads: More Information, Less Work

Marketers will also be able to use Google’s Gallery ads later this year to target mobile users. The ads appear at the top of the Google Search results page in — as their name suggests — a photo gallery format, complete with captions. “We wanted to combine the intent of search with a more visual and interactive search format,” explained Sissie Hsiao, vice president, Google Apps, while introducing the new format to the Google Marketing Live audience.

Hsiao said Gallery ads offer a way for marketers to highlight brands in “a more holistic way,” while reducing the work users need to do to get to the information they want when searching. Gallery ads can feature four to eight images with 70-character taglines. They can also use up to three different headlines so that marketers can test different calls to action. Users can swipe through the images in the carousel and tap on the ads to get more information about products they’re viewing. The ads can also take users directly to a retailer or brand’s website.

According to Hsiao, testing of Gallery ads has shown that the format sees 25% more interactions, on average, compared to other search ad formats; interactions include more paid clicks and swipes at the top of the mobile search page.

Deep Linking Capabilities: Less Friction for App Users

With another innovation for Google Ads — deep linking capabilities — Google aims to create a more consistent, personalized and faster shopping experience for users accessing mobile web and mobile apps. Hsiao said that users who already have a marketer’s app installed on their phone now can be taken from Search, Shopping or Display campaigns straight to the app.

That more streamlined journey across mobile surfaces eliminates the need for users to log in to the app or to re-enter payment and other information on the mobile web that’s already stored in the mobile app — which leads to faster transactions. Hsiao said that during testing, ads with deep linking capabilities saw 2X the conversion rates of other ad experiences.

Smart Bidding: 3 New Ways to Optimize Bids

Google also announced some significant enhancements to its ML-powered automated smart bidding in Google Ads. Marketers can now use campaign-level conversion goals — for example, they can choose to optimize all bids within a search campaign toward the goal of increasing visits to a physical store. Also, marketers can create conversion action sets that can be used to optimize bids across multiple campaigns. And lastly, Google introduced seasonality adjustments, which allow marketers to maximize sales or other special events to increase conversions. Google Ads will optimize bids only for the duration of an event.

Custom Audiences: A way to Expand Your Campaign’s Reach

Another big announcement at Google Marketing Live 2019 is that Google is combining customer affinity and customer intent so that marketers can build “custom audiences.” That means marketers will be able to target prospects, based on their interests and behaviors.

Nicky Rettke, director, product management, Video Ads, at Google, said marketers also will be able to use a new audience expansion tool to create very specific or broad parameters when building custom audiences for their campaigns. (The concept is similar to lookalike audiences on Facebook; marketers can expand the reach of campaigns to include users who have similar attributes as the core audience they are trying to target.) Rettke says marketers can also use the audience expansion tool to test and compare the results of campaigns.

The above highlights are only a few of the product announcements from Google Marketing Live 2019. For marketers, one message from this event is clear: Staying close to the evolution of mobile also means getting comfortable with machine learning. Another key theme of the conference was privacy, and how users want more control over their personal information and how it is collected and used. Machine learning can help provide the context that marketers need to continue reaching and engaging audiences with relevant campaigns, even as more consumers decide to share less of their data with marketers.

Subscribe to the Invoca Blog

Get the secrets of call tracking & conversational analytics delivered to your inbox.

Secrets of call tracking & conversational analytics delivered to your inbox.