5

So I work for a Media company and we use Google Analytics for tracking our journalists and contributors performance (between many other things we track).

The tracking code

To do so we use Google Analytics custom dimensions as follows:

ga('set', 'dimension2', 'john.doe');
  • dimension2 being named Author in our profile properties.
  • john.doe the username of the writer of the articles.
  • This instruction is made before ga('send', 'pageview');

The dashboard

Segment Creation

I created a segment for a single journalist.

Segment creation

Segment results

As you can see in the screenshot below. This segment says that this contributor had ~75K pageviews during Saturday, 31th October: Segment results

The custom dashboard

The custom dashboard creation

As you can see in the screenshot below. I'm using the same dimension with the same conditions :

enter image description here

Results

The custom dashboard results As you can see in the screenshot below. This dashboard says that this contributor had ~27K pageviews (!!!!) during Saturday, 31th October:

My question

Why the pageviews values are completely different, although we are using the same dimensions?

1

I'm going to be completely honest right off the bat, I've combed this question over several times and it's quite perplexing, so kudos to you for the challenge. Now, assuming you've double-checked your work and don't have any conflicting filters, errors in your segments, etc., here's what I have to offer. About a year ago, I had an issue similar to this where I was getting double the amount of data when I checked Google Analytics within 48 hours of the day in question, as opposed to after 48 hours (that's a just a rough estimate but two days seemed about the norm). It sort of made sense when taking the reporting lag into account but I still found it odd that I was getting half the data after the fact, so I decided to alter both of the existing custom dimensions by attaching them to two separate events. To ensure it wouldn't affect my bounce rate, I made them non-interaction events. This worked for me, I was getting the correct information at any given point when filtering by the event category and event action. This has led me to believe that custom metrics and dimensions have a tendency to overlap, if you will, and create mismatched data initially. This could be due to the asynchronous nature of the tracking script or perhaps it's another factor. And perhaps there's a filtering mechanism on the platform side that's able to sort this out, or maybe it was just an anomaly. There's no real way for me to know for certain.

In conclusion, I realize that this is largely an anecdotal answer but my suggestion would be to give this a shot. I'm obviously not familiar with how your tracking script is currently set up but adding a unique event to each dimension will help give you an extra layer of durability, through which you can filter each author and event.

| improve this answer | |
1

Be careful to compare the same data.

In your segment you are asking GA to show you all sessions that include that Author. This will include all the page views in these sessions, i.e. if someone visits more than that one page during their session, this will also be shown as a hit. The 75K value is for these total page views.

Your other report on the dashboard shows just the page views on the specific pages with the specified Author. For these, the value, accurately, is 27K.

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.