The AI and machine learning powering the interfaces we use, the advertising that we build and optimize, and the way that we measure performance are all setting the foundation for big data to radically alter the ways we work and live.
Google’s annual Marketing Next conference represented a push forward on “AI-first world” themes, many of which were advances or iterations on existing offerings — and those big bets are starting to gain real traction for the tech giant. Google’s Senior Vice President of Ads and Commerce Sridhar Ramaswamy even mentioned in his keynote that “technology is becoming the enabler it was always meant to be.” Below are the key takeaways marketers and advertisers should note from the conference:
Officially rolled out, this free version of the enterprise Attribution 360 attributes performance across channels measured in AdWords, Analytics, and DoubleClick Search. Google aims to solve attribution in 2017, but the immediate outcome is a continued shift away from last-click attribution within silos and into a cross-channel, cross-device view that includes their dynamic attribution tool, Data-Driven Attribution.
- Store Visits and Sales Enhancements
Google is officially rolling out store visit measurement to YouTube TrueView ads via location extensions, and store sales measurement via third-party credit card partnerships like Visa in late 2017. The move is intended to easily connect to credit card transaction data and see offline sales in-store. Store Visits will use Smart Bidding to maximize visits.
- Unique Reach for YouTube & GDN
De-duplicated reach stats across devices are available now in AdWords and will be available soon in DoubleClick.
- “One Step, One Second” and Accelerated Mobile Pages (AMP)
With most mobile site visitors dropping off pages that take longer than a few seconds to load, Google emphasized “one step, one second,” a push toward sites and apps that load lightning-fast with minimal steps. Google is continuing to push AMP pages generally but also for landing pages in search and AMP display ads.
- Virtual Reality (VR), and Augmented Reality (AR):
Google aims to make VR and AR ubiquitous on YouTube for brands and through Daydream VR platform and products. Google showed an example of a footstool that could perfectly match size and dimensions, and showed how various shirts would fit on a mannequin in the shopper’s size.
Machine Learning in AdWords
- Universal Shopping Campaigns
Like Universal App Campaigns, except Google’s machine learning capabilities can maximize assets (feeds, product data, creative, images) and placements (search results, YT, Maps, etc.) at a breadth not feasible manually.
- “Target ROAS” bid rule at the product level:
Moving this rule down to the product level means that in conjunction with Universal campaigns, manually-developed and curated structures for Shopping campaigns become less vital or unnecessary.
- Asset Ads
Like the dynamic campaign formats above, you simply input several phrases or snippets and Google will auto-test variations of ads built on those phrases to find optimal combinations.
Google Audience Segment Rollouts
Some of these are new, some we’ve heard of before.
- Consumer Patterns/Life Event segments for Gmail/YouTube
Google is using search behavior data to build on interest segments and life events (like moving, graduating, etc.) that can be used as audience targets for YouTube and Gmail campaigns.
- In-Market Audiences for Search
These are the same in-market signals that could be used on GDN (in-market for a car, for example) and applied to search campaigns. There is value for general non-brand terms, competitive terms, or even tangentially-related terms to provide further validation of the searcher’s interest.
Also of note:
Budget planning, forecasting and pacing tool. This looked promising to help manage across lines-of-business and accounts at scale, and granularly.
- Product Placement Ads
Google’s version of product ads that can be promoted on a retailer’s search results pages. Any manufacturer brand can join now, and they are building out the scale of retail sites included in the network on an ongoing basis.
Overall, Google continues to push the boundaries by grounding its innovations in machine learning capabilities. Announcements like the release of Attribution, increased in-store visit and sales measurement and expanded audience targeting capabilities on Google segments are exciting. However, there still lacks a common connector between the “walled garden” platforms and devices (from a user perspective) that acknowledges the complexity of the customer journey within certain platforms and sites. This is where some of the fundamental limitations become obvious in these steps forward and what we look to overcome in the near future. Resolution continues to partner with Google, adopting and testing these key feature sets and capabilities, and prioritizing customer experience, measurement, and the adoption of smart technology paired with marketing expertise.
For more information on the conference, check out Google’s Marketing Next 2017 one sheeter here.