Google Analytics is a widely acclaimed web analytics program that allowed traditional marketers to better understand the performance of their websites without paying any fees. However, there are also issues that arise from Google Analytics in terms of accuracy, reliability and other possible issues that come out as a result of it being a free offer and patronized by thousands of users.
As a user of this tool since late 2005, I’ve had my own share of problems that were construed as misunderstood metrics.
Non-real time results
Yes it’s easy to argue that this is not a problem but a feature that’s not getting a full-blown support. I usually have to wait until about 9am the next day to get some significant reports for the previous day. Yet, these report cover only up to three-fourths of the daytime. The good thing about this is that it keeps me from looking at the stats the whole day; Google Analytics is an addictive hobby for me. I have talked to representatives from Omniture and WebTrends who said providing a real-time report is a challenging offer to make. So perhaps I can accept this as a “mini version” of a web analytics tool rather than a problem itself.
Lack of accurate data
No two web analytics tools reports must be compared in anyway. It’s because each of them has its own defined set of rules in measuring data. For example, Google Analytics defines a user session as 30 minutes by default although webmasters can modify it. But what I saw was the case of referral data. A site like CNN.com that’s perceived to provide a link referral (identified as Referral) can also spit out organic search results (called Organic). Google Analytics does not distinguish the two and instead uses Referral for these sites under Visitor Segment Performance | Referring Source.
In addition, SEOMoz noted Google Analytics under-reporting of data, although I am hardly impressed by the explanation.
Update: There is a facility to define additional search engines.
Almost everyone agrees that bad publicity is still publicity, which is all good. But for a company like Google to receive flak about its subpar performance like its inability to handle a surge of registration and slow network connectivity is too much. Now that Google Analytics is open for everyone to register, those issues may be a thing in the past. Its current issue on unannounced downtimes is something that concerns me more.
Not tracking a site exclusively
What if someone stole your source code and makes changes to content, graphics and colors but FAILING to remove the Google Analytics code within that copycat site? As Terrill Dent found out, someone else could inflate your numbers without you knowing it. Google provides a workaround with this problem using the exclude filters. But what about if you haven’t figured out that your numbers are artificial and you’re not yet ready to face the reality?
One scary proposition is the fact that Google Analytics and Google Adwords belong to one company. Though it seems that Google can easily penetrate user data despite its promises of respecting user privacy, I don’t believe that it generates a big impact if they really do it. And I won’t mind either that: 1) Google Analytics offers the product free and with no conditions; it’s a risk users have even before they thought of signing up. 2) If data collected by Google would only mean more targeted ads served and lowering cost per acquisition while addressing click fraud problem, the go for it!
Incomprehensible (yet obvious) error
I tried to enter the following URL: https://www.google.com/analytics/home/report?rid=17133161291&scid=5021321
and I got the message:
Reporting Error Encountered. User does not permission to report.
All in all, benefits Google Analytics brings definitely outweigh the problems despite giving up the $199 price tag. I wouldn’t mind if Google fixes the “errors” too!