As a consequence of one of the previous posts on the redesign of this blog I had a question on statistics from someone who reads this blog (at least occasionally). I take the liberty to repeat the question and my comments here. I thought it could be interesting for more people:
Spurred on by your blog post where you compared WP and Blogger and mentioned this, I signed up to StatCounter and have even paid them for extra logs (on [the main site the writer operates] there are 1000s of pages hence it’s needed!). I’m finding the stats quite different to Google Analytics, for my main site it’s about 1/3rd higher on Unique visitors on average though not directly linear. On the other hand, for my WP blog it’s about 50% lower that the stats WP provides (but that doesn’t specifically mention Unique Visitors I don’t think). What’s your experience? Which is more accurate/quotable on ‘unique users’?
I use both StatCounter and Google Analytics. StatCounter since very long. It’s easy to use and to understand. Much easier to understand than Google Analytics. Perhaps Google Analytics gets more use friendly once you’ve spent days and days digging into it. One tip: if you want to track some specific pages, create a special StatCounter project for it, with its own code, and put that on the specific page(s) you want to track (in addition to the general code – but I assume you put that in the footer).
Google Analytics and StatCounter does give different stats. I’ve read various comments on it but nothing conclusive. The basic problem is that there is no single and reliable way to collect statistics. Both StatCounter and GA use cookies, but they use them differently. Hence they get different results. Some comments say that the difference may be due to Google Analytics being more effective in NOT counting visits by spiders / web crawlers than StatCounter. Counting spiders inflates the stats (it’s not real people). Not sure if that’s true because some say that Google Analytics may give higher stats in some occasions.
I don’t think it’s much of an issue. I would say both are accurate. As accurate as you can get. Stats is in any case a rather nebulous area. E.g. what is “a unique visitor”? If I remember right StatCounter defines it by a cookie that has a 30 minute life time (you can chose to change that). If the same visitor comes back more than 30 minutes later it counts as a new visitor. I am not sure exactly what Google Analytics defines it as but I would think it’s similar.
The most important thing with stats is not the absolute numbers (albeit those are good for marketing purposes) but rather:
– Trends: growth and fall in traffic. (Light-hearted example: a web site, not to be named, claimed that they had a lot of professional visitors since the traffic was higher in weekdays than on weekends… Hmmm. More likely, people surf for private things at work.)
– Where visitors come from (originating sites)
– Popular pages: which are those? Why do you think those particular pages are popular?
– Keywords: which keywords do people search on in search engines when they find your site
– And more
All of that and much more is in both StatCounter and Google Analytics. Use both and then decide which you like best. Or continue using both. It’s not expensive, but very informative.