This will give Google Analytics users a unified view of customer behaviour across the web and mobile apps

Google Analytics has launched its first major platform rehaul since 2012, Google Analytics Four, with machine learning at its core and a focus on privacy.

Similar to recent updates to Google Ads, the new analytics platform uses machine learning to help businesses take advantage of real-time data as consumer trends evolve during the COVID-19 pandemic, while giving them a long-term, unified view of customer behaviour across channels.

Google Analytics up until this point was primarily focused on desktop web, said Russell Ketchum, group product manager at Google Analytics. With the rise and proliferation of apps, the way customers interact with businesses is fundamentally changing. We needed something new for them to keep pace with all of that.

Google Analytics users will now be able to view customer behaviour across the web and mobile apps in one view, something they previously could only see in separate platforms, giving them a unified picture of how customers are behaving. Users can also view analytics broken out in phases of the customer lifecycle, from acquisition, to conversion to retention.

It provides a fuller understanding of the customer lifecycle by bringing together app and web data at scale, Ketchum said.

The new version of the platform uses machine learning in multiple products. Predictive algorithms surface real-time business insights for marketers, alerting them when demands for certain products are spiking, or when a customer is likely to churn or make a purchase. Businesses can use that information to optimize paid media spend.

Once a customer goes dormant, it can be really expensive to get them to re-engage, Ketchum said. We’re able to look across a host of signals and predict the likelihood of that happening, he said.

The new iteration of the platform also measures customer events rather than page views, allowing it to glean insight on different types of behaviours on new channels. For example, Google Analytics can now track engaged video views on YouTube, which indicates when someone is actively watching a video.

Because there’s not a click and it’s just the duration of watching the video, [the previous version of] Google Analytics couldn’t pick that up, Kethcum said. Now we can configure that media event as part of our overall cross channel attribution.

By leaning into Google’s massive logged-in user base, the new Google Analytics aims to help advertisers fill the hole left behind by the disappearance of third-party cookies.

Google Analytics is optional for Google Ads users, and vice versa, but the products are deeply integrated, especially for performance marketing use cases like audience targeting and conversion tracking.

Beyond being able to deduplicate user activity across channels, Google Analytics can use signals from Google users that have opted in to receive personalized ads to fill in the gaps on cross-platform activity. The platform has also added more granular privacy controls for advertisers to better manage how they collect and store user data.

As data sparsity becomes the norm, the overall picture is going to be more complex for a marketer,” Ketchum said. That’s another place where we intend to use machine learning and modelling to provide insights across those gaps.

Any view of the customer using Google data, however, can’t leave Google’s ecosystem due to privacy concerns, pushing advertisers’ ability to track cross-platform activity even further behind the platform’s walls.

If there’s a log-in or hashed ID they can pass to us, that’s the most deterministic and that’s what we’ll use, Ketchum said, but if you’re a business who doesn’t have that log-in, that’s where the Google signals can augment that.