Twitter’s 140 character limit brought URL shorteners (like bit.ly, ow,ly, etc.) into the limelight. Now, URL shorteners are commonplace – many services and applications embed their functionality (TweetDeck, for example), and even Google & Facebook have come out with their own shorteners (goo.gl and fb.me). A lot of websites and analytics programs rely on variables included in the URL to know more information about where a person came from (what site referred them). When an URL is shortened and then shared across social media sites, does that mess up the statistics? Does Google Analytics, for example, know where someone came from when a URL which includes extra information is shared on Twitter?
I do a lot of reading from my RSS reader – Google Reader. When I click to a story, I noticed that the URL has a lot of extra variables on it to denote where the person came from (the referring website/service).
In this case, it’s saying that the source of the content is feedburner. There are also variables showing that the “campaign” (Google Analytics campaign) is “Feed: Sazbean (Sazbean)” and something about Google Reader. So, I wondered what Google Analytics thought of all this. What happens if someone shares the page from this special URL? Does Google get confused as to where the traffic is actually coming from?
I send my RSS feed from my blog to HootSuite for it to automatically tweet when I post. HootSuite has it’s own URL shortener (ow.ly), that it applies to feed URLs so that it can track when they were clicked on, etc. (HootSuite has pretty good analytics in terms of this, by the way). So, here’s how the URL looks in my Twitter feed:
The URL is now shortened to http://ow.ly/169ra7. HootSuite says that it’s been clicked on 5 times. When you click on the link and close the ow.ly toolbar, the URL is the same as the one from the feed (with all the extra variables). So what does Google Analytics think about all this?
The only entrance sources (referrers) listed for the page are Feedburner. Now, HootSuite says that the link was clicked on 5 times, but Google Analytics only shows 4 pageviews with an entrance source (referrer) for yesterday and today (only a fraction of the pageviews for the page). This may be due to a delay in Google’s processing (which is usually delayed at least a few hours) and/or a difference in how the 2 services determine “legitimate” clicks (meaning perhaps 1 of the clicks on HootSuite was a bot from Google’s point of view). It looks like Google may be attributing the traffic from Feedburner when at least some of it may have come from Twitter. Google may be smart enough to figure this out (a lot of sites send the referrer information along in the background), in which case the Twitter visits may be showing up as something else – although Twitter isn’t listed as a referring source during that time period.
I searched around the web to see what I could find about this – most of the information about URL shorteners is just reviews. Google apparently changed how it was sending some of it’s variables back in February 2009, but that seems unrelated (and they said it wouldn’t affect Google Analytics). So, it may be that Google’s statistics in terms of referring sources is a bit skewed (not that you should take analytics at face value – they all have to do some type of estimating).
Does anyone know how Google Analytics handles variables in shortened and then shared URLs? Do your statistics seem to be correct? I’m going to do some more testing of my own….
(photo by AliceJamieson @ FlickrCC)
Liked this post? Consider subscribing to our RSS feed or our free email updates or following us on Twitter.
Technorati tags: social media, Google Analytics, analytics, marketing, statistics,social networking, business