There are lies, damned lies and statistics, and I'm tempted to think that internet-related statistics are among the most unreliable of all. I'm prompted to ask a question after taking a look at Blogger stats for this blog. I've read several times that it's a good thing to monitor the statistics for one's blog or website, but I must admit I'm not totally convinced. I can see that if fewer and fewer people look at the website, there may be something going wrong, but I'm rather dubious about some of the figures that appear to look very healthy, since they may be inflated by spammers or the like.
My post on Father Brown last week attracted a lot of interesting comments and when I checked the stats,it featured in my all-time top ten list of page views. But wait! I remember a couple of the figures for other posts that feature in the list - a review of Sherlock, and one about Reg Hill - from when I last looked at the stats a couple of months ago. But the number of pageviews for those posts has actually reduced. So, in one case, about 2100 all-time pageviews had reduced to about 2050. How can this be?
If anyone with a better understanding of Blogger than me can enlighten me as to how figures can actually reduce in this way, I'd be very glad, because I've become intrigued and puzzled. Mind you, it leaves me all the more dubious about statistics, and the lessons that can be learned from them.
Yet you never know - perhaps my cynicism is mistaken. The crime community includes many people with a real understanding of how to make the most effective use of technology that I only wish I could emulate. So I should also be very interested to hear from others as to whether they find statistics useful, and what lessons they think can be learned.
16 comments:
I'm with Wordpress and often wonder myself about the effectiveness of stats. My biggest post hits since I started was one announcing a Paul Newman marathon on Turner Classic Movies. It was worthless after a couple of days, yet it still receives hits and has a total of more than the next five together.
I have one post on a true crime book that happened, the crime I mean, within ten miles of home that every few weeks gets a large number of hits. That one is three years old and has reached the top five all time posts.
So I too don't know what all this means.
I'm on wordpress which has just starting splitting down views into 1)views, 2)visitors and giving you a 'view per visitor' percentage. This has been quite useful as it shows (what I always suspected) that most people visit the site for a particular review and don't really click around it that much - although there are exceptions. They are only aggregate figures so it's difficult to extract any more information than that from it.
In wordpress you do get anomalies like the one you pointed out. My highest ever page views were in May last year but this changed to July late in the year, with fewer views for no reason. I just keep a general eye on my stats but don't pay too much attention to them. In general, the more often I post, the more views I get.
I find comments much more interesting and this is what I am trying to nurture on my blog. Places such as twitter are good for brief messages but on the blog, commenting on other people's and in forums such as friendfeed we are able to have interesting discussions and I've found out so much about people's lives and interests that way.
I like Fr Brown by the way but the scheduling time is so awful I feel guilty when I sit down in front of the TV to watch it.
Well, the Blogger statistic tool is highly unreliable.
As you say, the numbers are sometimes inflated by spammers, and then again Blogger cuts these inflated numbers once in a while.
As for myself, I´m using Google Analytics which seems to be a far more reliable tool with a lot of additional informations.
By the way, your blog is one of the most enjoyable crime blogs I´ve met. So, thanks for your efforts.
Kind regards
Morgenländer
Martin - I'm not thoroughly convinced by blog statistics either. I do think they're helpful in some ways. For instance they can tell one which sites lead readers to one's blogs. They can give some interesting information about who reads the blog and so on. But I agree they don't tell the whole story.
Blogspot/Blogger's counts seem particularly hincty, indeed. I enjoy also that my "all-time" counts begin a year after I started my blog.
I use Blogger stats to monitor unwanted visits. It's the referral spam that tends to create all the inflated page hits stats with the Blogger stats. There are daily complaints about spam referral sent to Blogger Powers That Be. They have no interest in filtering it out or doing anythign for the platform to eliminate the daily maurading of spam bots. You already have comment moderating so you have the spam under control. And the lack of images on your blog is a plus as far as inaccurate page hits go. I have an image rich blog and nearly all of the most visited posts are due to Google image searches.
I have a separate statistic service I subscribe to when I'm iunterested in monitoring my readership. The stats are completely different and much more accurate.
IMO the sudden drops in numbers can be explained by the annual, twice-yearly change in time from BST to GMT and back (or whatever it is these days in the UK). When the clocks change back, your numbers may go back up again.
I'm really grateful for these very constructive comments and will respond individually.
First, Randy. I've also noticed that some topical posts seem to get a large number of hits. Film and TV are also popular - more so than book reviews, or so it seems. I'm surprised by that, but the stats seem to tell me this.
Sarah, that's a very good point about comments. I've found many of the comments I've received extremely rewarding and yet I still find it hard to predict which posts will generate debate and which will not. But 'writing tips' seem to prompt comments regularly.
Morgenlander, thanks for that.
I have started using analytics for my website, though perhaps I ought to apply it to this blog as well.
Thanks, Margot. A good point about sources of visitors. I think you've also mastered the art Sarah mentions, about generating comments with your fascinating posts.
Todd - something similar happened with this blog - glad I'm not the only one!
John, very interesting. I didn't know that about images. I often find Blogger refuses to let me post images, which is why I only do it when I have quite a bit of time to spare. But your experience makes me wonder if I should bother. Having said that, the images on your blog are wonderful, so I do hope you keep sharing them.
Dr Evangelicus, thank you. The drops seem to be a bit random, though clock changes may be another factor. I think as ML says, it may be about the deletion of spam to some extent.
Subscribe to your blog on google reader, Martin, so my views probably don't show up in your stats but I read (and enjoy) most of your posts!
Blogger stats are definitely unreliable. Has anyone experienced this scenario: You're looking under "posts" and trawling back through the entries to see which are getting hits. You eventually spot an old post with a strangely large number of hits. This post SHOULD be in your top 10, but Blogger hasn't listed it as being in there. Why not? Anyone?
Post a Comment