Sunday, January 05, 2014


I don’t often blog about blogging, or this blog in particular, because I think that it’s mostly other bloggers who’d be interested. But, hey: Bloggers are people too, mostly, so why not? Especially since I’m about to return to regular blogging topics.

Today I had a look at some of the metrics I use to gauge readership. Mostly, this is just interesting for the statistics geek in me, but looking at the data also shows me what topics readers find most interesting. Not that I pay much attention, but it’s interesting to me all the same.

I looked specifically at two general things about the blog in general, rather than specific posts. The first, in the chart above, is how this blog ranks among blogs with publicly available site statistics and written by New Zealanders who choose to participate in the NZ Sitemeter Rankings published by Open Parachute (something I started participating in back in March). In that chart, the lower bars are better because that means this blog was ranked higher.

In November, this blog cracked the top 100 NZ blogs for the first time—just—when it hit 99th place. I assumed that was a one-off due to a high pageload in November (about which, more in a minute). However, in December this blog was actually 90th.

Now, I know plenty of people who’d never admit to such low rankings, but considering roughly 265 blogs are ranked ever month, that means plenty are far lower than mine. I also consider it a bit of fun, and possibly a source of an occasional new reader, and nothing more. It is possible to game the system, since there’s no perfect indicator of who is accessing a blog and for how long.

Which brings me to my next chart, at left (click to embiggen). It compares pageloads against unique visits as reported by StatCounter, which is my data collector. Pageloads (the top jagged line) refers to the number of times a page was accessed. Unique Visits (the bottom jagged line) are determined by cookies and counts the total number of visits from all people—new and returning.

You’ll notice the gap between pageloads and unique visits. This is because of numerous things, for example, people whose browsers don’t accept cookies, bots trawling the Internet and also the fact that a person whose browser does accept cookies could access pages on the blog several times during their visit (they have to be away at least an hour to be counted as a unique visitor again).

I’ve also shown the median for both metrics (page loads and unique visits), the straight line running through the data. It makes the massive spike in November really obvious. That spike was because of my post on Illinois’ marriage equality. That post got around four or five times as many pageloads as is typical because it was included in “Mike’s Blog Round Up” on Crooks and Liars website. That sort of thing doesn’t happen very often. A version of it was also published by Chicago's Windy City Times, but I digress.

What all these stats show is that this blog is growing over time, and I’m glad about that. It would be nice if more readers participated, but maybe that will become more common eventually, too.

But the second chart also shows what happens when the number of my posts drops as it did especially in May, mainly due to technical problems beyond my control. It shows how fragile these statistics really are.

Finally, on New Year’s Eve I spent hours compiling a blog post I never published. I looked at every post I published in 2013, commented on some and posted links. It included posts from every month, and by the time I got to October, it was a massive beast, some 967 words. While that included the links as words, it still would have easily been 1200 words or more when completed. Even worse, it was also looking like a link farm of sorts, which is not a good thing. So, because of all that, I abandoned the post. I thought about publishing it as a downloadable PDF for anyone who was interested, but that seemed a little, er, um, too much.

Roger Green took another approach. Each January he selects a post for each month of the previous year using a random number generator. Maybe I’ll try that at the end of this year—I’ve stolen a lot of other ideas from Roger, so why not?

And that concludes my unannounced “holiday mode” on this blog, which, apart from one particular exception, meant I mostly avoided politics and pointed commentary for the past couple weeks. But now it’s time to get back to normal blogging; there are things to say, after all—which is kind of the point of having a blog, isn’t it?

A couple technical notes: I created the charts in Adobe Illustrator, and I haven't quite gotten the hang of creating borders for web graphics. Add that to my list of things to learn this year. Also, I manually entered the data onto spreadsheets, and it's possible I made mistakes. Even if I did, I doubt they're significant.


rogerogreen said...

I've been impressed to see others do annual blog retrospectives, such as you attempted, but I have ZERO desire to do it myself!

Arthur (AmeriNZ) said...

When I started, I thought, "oh, that'll be easy: Just pick the posts I like the best or that I think were the 'most important' or whatever, provide a brief comment and a link, done!" It turned out to be MUCH harder than I'd thought. If I were to do it again, I'd start at the conclusion of each calendar month—hm, there's an idea!