There’s something strange happening on my website, at least according to Google’s Universal Analytics (UA) tool.
Since our review when UA went into open beta a couple of months, I’ve been monitoring it to see if there are any differences in what’s being reported. There are. A quick comparison between my UA profile and the regular Google Analytics profile shows that UA is reporting 40 more visitors than GA.
While that won’t make a big difference to my overall analytics stats, I set out to see what other differences there were, looking at the audience, traffic sources and content reports for both profiles.
Care to join me in some detective work?
The first thing I discovered was that the changes weren’t stable. In other words, while my dashboard showed 40 extra visitors, delving into the audience report changed the number slightly.
My regular GA profile showed 1872 unique visitors and 1.89 pages per visit while my UA profile showed 1907 visitors (35 more) and 1.78 pages per visit.
At first I wondered whether UA was counting my own visits, but that would still only account for 1 extra visitor, 2 at most if mobile access was counted separately, so that couldn’t be the case. Still mystified, I hovered the mouse pointer over the chart at the top of the dashboard and discovered that the UA profile was showing between 2 and 4 extra people every day.
Looking into the audience demographics, language and location showed much the same pattern in both analytics profiles, with most visitors coming from the US, followed by the UK. However, there were some minor differences.
Looking at US visitor demographics, UA showed 14 more visitors from California and 1 fewer from New Jersey compared with the GA figures. Confusingly, UA also showed fewer page views, making the average page view figure lower.
UA is supposed to offer more features for reporting on mobile devices, so I wondered if this section of the report might provide more data on my 40 extra visitors.
It didn’t, though I noted that reporting on the use of desktops, tablets and smartphones is much improved since I last look at Analytics. The mobile report in UA differed from the GA counterpart, showing 45 more desktop visits and 5 fewer mobile visits.
Analyzing Traffic Sources
A look at the traffic sources section revealed a similar pattern to the audience analytics section, though again there were some reporting differences.
For example, the UA report did not show campaign traffic, while the GA report did. And I was unable to make comparisons for the SEO report because UA is not yet a part of my Google Webmaster Tools account. (As my site’s GA profile is already set up in Webmaster Tools, I experienced some difficulty in trying to add a UA profile too.)
Dipping into the social overview report revealed that UA was still showing 40 extra visitors and had the same basic pattern as in previous reports. However, there were minor differences in the percentages of visits reported for each social network.
This also applied to the data hub activity view. However, it was interesting to note the increased level of detail in both reports about posts, social sharing and trackbacks. This is a useful feature for marketers, bloggers and website owners.
The content report showed a difference in the number of page views, bounce rate, exits and the time visitors spent on page, though the latter was only three seconds. The top ten performing pages remained the same in both instances.
Looking at page load speed, both reports confirmed what I already knew—that my site has slowed down (perhaps because of the different code snippets I’m testing). However, the UA report showed sites loading slight faster (half a second), with that small speed increase reflected in page lookup and server connection time.
Finally, since my site has no ecommerce goals (so no conversion tracking) I moved back up to the intelligence events section of the GA and UA profiles. There were minor differences in the events and automatic alerts reported.
What’s the Answer?
So what did I learn from all this sleuthing—and where did those 40 extra people come from? Even after digging into the metrics, I still have no idea. Some of the possibilities are:
- That GA has been under-reporting visitor numbers and the UA figures are correct.
- That UA is over-reporting visitor numbers and the GA figures are correct.
- That there’s a conflict with some other bit of code on my site, resulting in a reporting anomaly.
- That there’s another perfectly reasonable explanation I haven’t thought of yet.
Until UA is more widely used and there’s more insight into setting it up, perhaps the best approach is to take Google’s own advice from the Analytics Help section:
“We generally believe that the best way to think of metrics across different web analytics programs is to think in terms of trends, as opposed to numbers by themselves.”
Since the trends on my site are largely the same, maybe I don’t need to worry about those phantom visitors after all!
What about you? Have you experimented with Universal Analytics? What are your findings?