Using Google Analytics Audiences

Google Analytics Audiences is a powerful real-time audience tool, when using the full google stack. 1st party on-boarding audience strategies are delayed because of matching. Cookie/floodlight audience strategies are becoming more and more limited due to ITP, and don’t allow for the depth of behavioral segmentation (time on page, number of pages viewed, etc.) Leaving GA audiences a nice real-time and behaviorally deep compliment to your overall audience segmentation strategy.

The google stack:

DoubleClick Campaign Manager | Display Video 360 | Search Ads 360 | Google Analytics 360 

The power of Google Analytics audiences is that they can be shared across the google ad stack, allowing you to adjust audience segmentation strategies across search, video (including Trueview) and programmatic display. Dynamic creatives can be assigned based on site behavior, and all the audience segmentations can be setup through logic parameters within the GA audience tab in admin. The rules run in the background assigning any traffic that meets segmentation criteria automatically to that audience.

DCM and GA 360 have a “click-on” integration too – porting view-through ad impression data (for site visitors only) into Google Analytics, allowing you to segment audiences by which ads they’ve been served, along with site behavior. 

So what does this all mean? 

Let’s take a look at a common acquisition funnel, and three different possible abandoner segmentations.

Deeper Behavioral Segmentations

This is a typical three audience bidding strategy, commonly implemented through tag implementation on the site. From a traditional tagging perspective, the segmentations are made based on how many “steps” the abandoner completed. But it’s harder to target behavior beyond that,.

For example, how long did the abandoner spend on the site before reaching the cart and not creating an account? Wouldn’t we want to more aggressively bid to retarget someone who spent 10 minutes on the site, as opposed to someone who spent 2 minutes on the site? 

Return visitors vs new visitors. Cart abandoners who came in on organic search vs those who came in from paid search vs those who came in direct to the site. Abandoners who were served a video ad then came in through paid search vs abandoners who came in through search without being served a previous ad. By consolidating so much of a users ad and site behavior in Google Analytics, there’s a myriad of deeper segmentation possibilities. 

All of these segmentations can be pushed into DV360 and SA360 as audiences to be individually targeted, bid adjusted and budgeted – allowing marketers to reach the highest value segmentations, and trickling spend to other audiences from there.

Which Segmentations Are Most Valuable?

Start with general higher level segmentations and don’t apply bid adjustments yet. For example, you could set up a simple the three above audience segmentations without any individual bid adjustments. After a week or two of running, confirm if those who have created account convert higher when retargeted then those who did not create an account. 

If that’s what the results confirm, then set higher bids against account create audiences. You could then split those who didn’t create an account into two sub-segments: those who didn’t create an account and spent under 5 minutes on site vs. those who didn’t create an account and spent 5+ minutes on site. Let those two segments run without mid modifications for two weeks and see which has a higher conversion rate. Apply bid modifications, and then dig deeper into another segment.

By repeating this process you’re separating the more-likely-to-convert behavioral segmentations from those that are least likely to convert, allowing you to spend your retargeting budgets where they have the most likely impact. 

Audiences are available in the free version of GA, but GA 360 gives you a higher allowance of audiences and also allows integration of DCM data into GA. (Along with offline and 3rd party integrations to import audience targetable data from other sources.) 

I’m beginning to think of GA 360 as a DMP-lite within the Google Stack. While it certainly doesn’t replace, or fulfill, all of our audience needs (1st party data is still our primary audience source) – it’s becoming a bigger component of our overall audience targeting strategy.