Eight million pageviews later..

I signed up for Google Analytics in November of 2005. Since then, some 8 million pageviews have been tracked by their software, allowing me (and of course them), to learn a lot of useful information about page performance and popularity.

Analytics isn't entirely accurate. It depends upon a bit of Javascript imbedded in the pages, so if a visitor has Javascript disabled (very common nowadays, especially for Firefox users), Google can't count the page view. That Javascript is also usually installed near the end of the page, so if the page doesn't load completely (because the visitor quickly clicked away or because of Internet problems), that view can't be counted either. For these reasons, you can't entirely trust Analytics figures. However, you can believe that if Analytics says your page was seen 50,000 times, that's a minimum figure and it is likely real people (because spiders don't ordinarily execute Javascript).

So what have I learned in almost four years? Let's take a look at some figures. First, the Top Twenty in the past year:

I have no idea why that Google Earth post gets so many hits. Notice that the "Time on Site" isn't all that high, but then again, it's a short article - it doesn't take long to read.

The rest of the table makes sense - those are useful posts about commonly encountered problems or concerns. If you want to write popular pages, studying that list might give you some ideas.

When we look at the whole period that Analytics has tracked for me, the picture changes a little. That Google Earth post is still well into the top twenty, but it's not at the top (on the other hand, it wasn't written until June of 2006). Everything else follows the same pattern as we see in the past year: articles about common problems are the most popular.

The articles that made both lists here (current year and past four years) are broken out below, and can help show whether each article is gaining or losing popularity.

Some articles (like the one on Virtual PC) have fallen out of the top because whatever they were about is no longer of interest. You have to expect that, of course. Others disappear for no apparent reason: perhaps someone else wrote a better post on the same subject and has taken the traffic? Or maybe they've just been pushed down because other posts have become even more popular.

If we are trying to find the magic formula for popularity, we need to go beyond this. We need to look at the source of the traffic - search or links? We need to compare page length, readability and the level of expertise the articles are aimed at. We need to filter out transient spikes such as might be caused by a mention on SlashDot or Linux Today. Analytics can't give us all of that, though it can help.

For any of the authors who have posted articles here: I'm happy to provide Analytics data for your posts - just ask!

Got something to add? Send me email.

1 comment

Thu Sep 24 15:57:49 2009: 6967   TonyLawrence


By the way, I don't mean "flash" popular - getting a few tens of thousand hits on the first day of your post and then nothing. Some people want that sort of traffic, which is why they beg for Diggs and try to get mentioned on SlashDot.

99% of the time, that sort of traffic is useless - big spike and then nothing. I want long lasting posts that draw traffic year after year. YMMV, of course.

------------------------

Have you tried Searching this site?

This is a Unix/Linux resource website. It contains technical articles about Unix, Linux and general computing related subjects, opinion, news, help files, how-to's, tutorials and more.

Contact us