Using Custom Dimensions & Metrics (on Measure Boards) with Universal Analytics

Note: Since this was posted in mid-2013, the domain has since expired. Womp womp.

So there are these new nifty things in Universal Analytics that take the place of your typical Custom Variable. These are Custom Dimensions/Metrics. I’m not talking about the Custom Data Source. I’ll save that for when I’m actually able to use that feature. However, I did want to create a post about how I’ve used the Dimensions/Metrics thus far. If I had any amount of traffic, the results might be more interesting.

Mission: Tie together users between devices… and be able to measure how many of the total users use multiple devices to browse over just 1.

Now you might think – “Psh, well that’s easy! Just find one common key and you can do it.” That’s a start. Yes, we need to find a common key. What’s a common key? It’s the way we can match up users between sessions on completely different devices. This means that the user MUST have signed in at some point on each device. Fortunately, usernames are unique to each registered member so I used that. Now sessions between computers and devices are tied together. However, we still have the challenge of creating our wonderful breakout of each user. If these forums had 300 people, that would mean I’m going through 300 unique names to see if that person has logged in via mobile/desktop (via the Mobile reports). If I grouped them all together, there wouldn’t be a way to dedupe multiple sessions from a username that visited via mobile… so it wouldn’t give us the result I want.

I want a way to be able to create a binary chart of: If a value exists in this column, then that person used the platform. That way, I can create a giant pivot table of how many individuals use Mobile/Desktop/Both. To do this, I created 2 dimensions – Desktop Status and Mobile Status. If a status returns true, it sends an integer of 1 over to GA. I don’t know exactly how it’s calculated – whether it’s hit or session based, but that really doesn’t matter since all I want to know is if it is greater than zero:

Custom Report with Universal Analytics
Click to enlarge

That image shows a custom report I created using the custom dimension/metrics. From here, I can then get a full picture of how many people are using both platforms. I can then use this data to help guide my development priorities.

What do you think? I’m certain there is an easy way to do this, but this is just my first stab at playing around with the new features of Universal Analytics. While you’re at it – see it in action at!

Leave a Comment