ThePickards Site Stats
Someone asked me the other day “how many people read your blog?”.
As I felt my answer (”er… I dunno, really”) lacked that certain something, I decided that it was worth having a poke around in my site statistics to see what I can find out.
My site statistics program was defaulting to show all statistics for the last month — from the period of 7th November through to the 6th December 2008. Statistics for the final day would be a little out, as at the time I snapped the images, we were only about 75% of the way through the day, so if the final day shows as a little low, that’s probably why.
Nevertheless, it’s quite clear from the image — assuming of course you can read the numbers clearly enough — that I generally have somewhere around 2000 visits per day.
The obvious questions that this provokes are:
- Does this include bots and spiders?
- Okay, that’s how many visits, but how many visitors do I have?
The new and returning visitors graph starts to tell us something interesting. Obviously, for the purposes of day 1, it treats every person as new, and as you go through the month the number of new visitors in that month will tend to decline (because anyone who visits the blog on a semi-regular basis may already have visited, and therefore would be counted as a returning visitor).
If I then go and access the actual figures for the ‘new’ visitors for each day, and add them together, I will get a total of how many individual visitors I have in a given month. And the answer is apparently 20,688. Twenty thousand!
I was a bit gobsmacked by this. Granted, it’s not exactly like I’m competing with the BBC, but the idea that twenty thousand people in any given month want to read what I have to say (or at very least, end up reading what I have to say) is somewhat humbling. Although it does beg the question “why don’t more of you leave a comment from time to time?”
The next question is to determine what proportion of these visitors could be bots. Now, I think these ‘visitors’ stats have attempted to remove those identified as spiders or bots, but assuming they haven’t, we can have a go at estimating the proportion of them by looking at visits broken down by browser type.
What the breakdown by browser tells us, is that even if we assume all of the “unknown” browsers are spiders, that around half would still be people. Now it’s difficult to assume a 1:1 relationship between visits and visitors, because this rather depends on how often certain bots visit the site, but even if the ‘visitors’ graph did include bots, this would still mean that around 10,000 different people would visit my site in a month.
It was also interesting to look at a geographic breakdown, because last time I looked at this sort of thing I discovered that I had visits from Libya, and I was wondering if they were still reading the site. It turns out they were: while the majority of African visitors were from South Africa, I also had a visit from Rwanda.
I might have to look at doing some sort of “country breakdown” at some point to monitor my visitors until such point as I have had a visitor from every country…
And finally, no look at any set of site statistics is complete without trawling through the search requests to find out what people are searching for when they end up on the site. This month seems to have had an awful lot of searches for Football Manager 2009 ending up here, presumably because it just came out last month and I wrote a review of it.
Somewhat surprisingly, people are still ending up here by typing “find madeleine” into search engines, long after the media circus surrounding Madeleine McCann’s disappearance has died down. The Wife-fattening searchees are because I referenced a BBC article about said practice in Mauritania; people searching for Sunderland Fc and Ipswich Town blogs were presumably sadly disappointed.
I was also interested in the “I’m not a racist but…” searches. Was it someone searching for evidence that the statement following the but almost invariably contradicts the statement before it? Given the fact that the searches were virtually identical, this would suggest presumably that it was one person (or maybe a team), but then you’d have to ask why would they visit my site six times as a result of this search? Surely once they’ve read the article in question once, there’s not really much need to read it again just because they have typed the search terms in slightly differently?
So, all ten-to-twenty thousand of you, apologies for the introspection and self-indulgence of this post, and we will return to the normal drivel in the next post…
Hello. I am 
Have you looked at/tried Google Analytics?
As well as producing very nice and accessible traffic stats, being Javascript-based it does a mean job of filtering out most of the non-human visits to your site. Also has pretty good reverse DNS when it comes to looking at visitor demographics. Shouldn’t replace server logs but provides a useful complement to your repertoire!.