|Google's recent changes and their affect on analytics|
Https search and browser add-ons
| 5:28 pm on May 26, 2010 (gmt 0)|
This week Google announced two new features, one that was expected and another that caught me completely unprepared. Both have the potential to significantly detract from our ability as webmasters to gather meaningful information through analytics.
The first was yesterday’s announcement of new add-ons for IE, Firefox, and Chrome that would allow end users to opt-out of Google Analytics.
Now this only affects Google Analytics users and not other paid solutions, so if it becomes widely used one can always move to another platform to avoid the loss of data.
My opinion on this change is that Google has unnecessarily put the cat amongst the pigeons. Most web users have no idea about what data is being collected about them when they browse online. If someone (Google) informs them that data is being collected and there is a really easy way to stop it, then they will do just that.
However, what the average user fails to understand (and what Google fails to mention in their blog post) is that the tracking of this information provides webmasters with critical information that we use to improve the functionality of our sites. Google already does a terrific job of anonymizing this data. Trust me, I’ve tried to connect my analytics data to individual sessions before and if you have even a moderate amount of traffic and it is at worst extremely difficult and at best impossible.
In my opinion, this must have been a preemptive move in regard to the privacy pushback against Facebook’s recent changes and even Google’s own Buzz fiasco. However, unlike both of those situations, I don’t believe that Google was truly invading anyone’s privacy by allowing sites to gather and analyze aggregated usage data. I think this was a mistake on Google’s behalf, and the first real argument that I’ve seen for a move to a paid analytics platform.
The second announcement has the potential for much farther reaching ramifications to the analytics world. Last Friday, Google launched their much-awaited secure search function that allows users to perform searches under an https connection.
On the surface, this looks like a good option for privacy-centric searchers. Again, most people have no idea the number of ways that 3rd parties can see what they’re searching for, whether it’s through an open WiFi connection or even records stored by an ISP. Unlike the browser add-ons, I think secure search is a good thing for the web. The downside is that it strips all referrer data when the user moves from https back to http. This isn’t any fault of Google, just the way that browsers intrinsically handle the https/http handoff.
Any analytics solution that uses referrer data to track source information will now view this traffic as direct key. As far as I know, any tracking done by URL parameters will be unaffected by this change, but I’d like to hear others opinions of that.
How widespread will secure search become? Who knows, but it is already the default Gmail users.
What do these changes mean for the future of the web analytics industry? Are we heading back to analyzing log files, and if so will even those be as reliable as they used to be? Is this the first inkling of a push back against the collection of basic usage data?
I have nightmares of focus groups dancing in my head…
| 12:10 pm on Jun 7, 2010 (gmt 0)|
google analytics updates which you have provided when did this occur.....?
| 3:58 am on Jun 10, 2010 (gmt 0)|
May 25th for the first, May 21st for the second.
Maybe I was overreacting a bit. I guess I was more disturbed by the sentiment behind the changes (an analytics provider that was actively making it easier to not be tracked by their tool) than I was by the actual changes. From other discussions I've had, neither of these changes are likely to have any significant impact on any site's analytical data.
However, its a slippery slope that all users of GA should be keeping an eye on.