Archive: Google Analytics

This week I helped out Peter John Meiklem on a story for the Sunday Herald about the future of journalism. Amid all the job losses in the Scottish media, the piece looks at whether bloggers can step up to the plate and begin to supplant traditional media outlets.

My view has long been that blogging is best consumed as a complement to professional journalism, as you’ll see if you read the piece. But of coruse there are other views out there, which the article also represents.

There are a few points I thought I’d mention, just to expand on or clarify a few points. It’s worth remembering what we mean when we talk about web stats. It’s a thorny area, and there isn’t really a good way to accurately estimate how many unique visitors a website has. I tend to look at visits rather than unique visitors because I think it’s more unambiguous.

I was a bit vague on the telephone about how many visitors this blog gets. I knew that all of my blogs put together get over 10,000 visits per month (this is the number I keep in my head because it’s nice and round, and it’s also sufficiently large to sound relatively impressive). In this case, it’s bad luck that the number of unique visitors to this blog in the month in question was 8,465. Not quite the >10,000 mentioned in the article, though if you throw in the numbers for my F1 blog vee8 it nudges above 10,000 unique visitors.

Apparently 10,000 per month is a similar readership to many local newspapers. I don’t know if this refers to the circulation of the hard copy or the figures for a local paper’s website.

Certain bloggers, who regularly post indulgent stat pr0n posts, get quite excited about how many visitors they get. But it’s worth remembering just how meaningless most visits actually are. Okay, this blog gets roughly 10,000 visitors a month. But it would be pure delusion to believe that there are 10,000 people out there who just can’t wait to read what I have to say.

Only an eighth of those visitors came here directly (i.e. on purpose). Over two-thirds of the visitors to this blog come from search engines. Of these, 94% have never come across this blog before. And 86% of search engine visitors to this blog look at one page and leave, spending on average a paltry 40 seconds here. They will probably never come back again, no doubt having failed to find what they were looking for. All-in-all, only 13% of this blog’s readers are returning for a second visit. Kudos to the 1% who have visited 100 times or more.

Of course, as always, these statistics come with all sorts of health warnings. Then there is the fact that many people are able to read blogs without ever having to visit, thanks to the magic of RSS. For what it’s worth, according to Feedburner, 270 people are subscribed to this blog.

Partly because of all the problems of getting accurate figures, I don’t get as hung up on stats as I used to. I like to know where traffic is coming from if someone has linked to this blog, but the numbers don’t excite me as much any more.

It’s come a long way though. I remember when I started out blogging, I used to be a bit freaked out when I saw the blog had had 60 visits in a day. That must have meant that I had (accidentally) said something too controversial or someone had ripped me to shreds and linked to it. Given that I was so young when I started blogging, they were probably right to do so. Eventually, getting 60 a day was the norm. Now 300 a day is a disappointment.

In the piece I am quoted as saying, “The average age of a blogger is around 40.” I don’t think that’s quite what I said (and I certainly didn’t intend to say that). I think the average age of the readers of political blogs is 40. My impression — it’s just a guess — is that the bloggers themselves are generally younger than that. But the point about blogging is that it can be — and is — done by people from all sorts of backgrounds. The eclecticism of the blogosphere is, of course, one of its biggest attractions.

As for the idea that the average reader of a political blog is aged 40, this is something I heard or read a long time ago and the source is long lost. I do like to pluck it out from time to time though to illustrate that blogging is not just a young person’s game. A quick search has yielded this study (PDF) which found that the median age of a political blog reader in the USA in 2006 was 49.

Another thing I wanted to mention was that the piece says that Guido Fawkes broke the story about North Lanarkshire Council’s head of communications job. In fact, my post about it was published about an hour before Guido’s, though I understand if more people came to learn about it through Guido. It’s also true that Gudio went a lot further, by actually naming the people involved, which I was reluctant to do.

Anyway, it’s great to have been quoted so much in the Sunday Herald this morning. I don’t mean to come across as sniping — inaccuracies are always bound to creep in, and you certainly couldn’t say that bloggers are much less error-prone.

One of the great things about having a blog though is that it allows me to clarify a couple of things which I said when I was working from the top of my head. That is one area where the blogosphere definitely has the upper hand over traditional media. On an open blog, some pedant like me will soon be along to point out the mistakes in the comments section. But the newspaper will never be corrected.

While perusing the stats for my blogs, I noticed that one of the referrers was this URL: http://www.charlesgordonmsp.com/stats/usage_200902.html. I clicked through to see what it was all about. To my astonishment, I was taken directly to the Webalizer stats for Charlie Gordon’s website.

This seems quite unusual to me. To access these stats for my websites, I need to log in with a password. Surely most other people do for their websites as well. But for the most expensive website for an MSP, such basic security measures do not seem to be in place. When you consider the possibility that search logs may contain constituents’ sensitive information, it seems to be quite an oversight.

There is one upside though. This free access to Charlie Gordon’s stats does give us the ability to calculate just how much value for money the taxpayer is getting out of his website.

For those who missed it, last month the Scottish Parliament released MSPs’ expense claims. The Scottish Parliament website allows you to search for expense claims by category. One of the categories is ‘Website Costs’, giving us the ability to see just what MSPs are spending on their websites.

It made the news that Charlie Gordon’s website was the most expensive of all the MSPs — by a very long way. Duncan Cumming conducted a full analysis. Charlie Gordon claimed £12,822.62 in website costs for the financial year 2007–2008. The next largest claim was by John Wilson, who claimed £2,291.25 — less than a fifth of what Charlie Gordon claimed.

For what it’s worth, Charlie Gordon released a statement on his website. There is no permalink for it, so you will have to scroll down — it’s (erroneously) dated 23 January 2008. It says: “My website costs for 2007/08 were around £1,700; not £12,900 as stated erroneously on the Scottish Parliament’s website!”

The MSP claims that only 20% of the costs outlined by the Scottish Parliament went on the website itself, the rest being paid for “call handling”. It is worth pointing out that even if we take Charlie Gordon’s claims at face value, a £1,700 claim would still make his website the second most expensive MSP’s website.

Here are the full details of the expense claims as laid out by the Scottish Parliament:

Claim Month: November 2007
Payee: GMG SOLUTIONS
Amount: £1,709.38
Additional Info: ADMINISTRATION AND WEBSITE MAINTENANCE

Claim Month: October 2007
Payee: GMG SOLUTIONS
Amount: £1,441.61
Additional Info: ADMINISTRATION AND WEBSITE MAINTENANCE

Claim Month: November 2007
Payee: QUEENS PARK FC
Amount: £11.80

Claim Month: September 2007
Payee: GMG SOLUTIONS
Amount: £1,200.00
Additional Info: ADMINISTRATION AND WEBSITE MAINTENANCE

Claim Month: August 2007
Payee: GMG SOLUTIONS
Amount: £1,932.00

Claim Month: July 2007
Payee: QUEENS PARK FC
Amount: £11.80

Claim Month: June 2007
Payee: GMG SOLUTIONS
Amount: £1,152.00

Claim Month: July 2007
Payee: GMG SOLUTIONS
Amount: £1,032.00

Claim Month: April 2007
Payee: QUEENS PARK FC
Amount: £11.80

Claim Month: April 2007
Payee: GMG SOLUTIONS
Amount: £804.00

Claim Month: May 2007
Payee: GMG SOLUTIONS
Amount: £708.00

Claim Month: March 2008
Payee: GMG SOLUTIONS
Amount: £144.00

Claim Month: January 2008
Payee: GMG SOLUTIONS
Amount: £1,464.00

Claim Month: December 2007
Payee: GMG SOLUTIONS
Amount: £1,044.43

Claim Month: September 2007
Payee: QUEENS PARK FC
Amount: £11.80

The pongy whiff intensifies when you read the press reports which noted that GMG Solutions is in fact run by Charlie Gordon’s son, Gavin. As Heather from Idea15 noted, GMG Solutions “does not have a web site, a portfolio, or any basic contact information, and from that we can infer that they do not exist.”

Heather was none too impressed by the website:

It’s done in table layout, its base colour is flamingo pink, it uses Flash for basic navigation buttons, and it has 45 basic coding errors. Worryingly, there are no analytics counters or codes, which means the MSP neither knows nor cares why people might be reading his site.

Moreover, the poor design of the website means that it actually may be in breach of the Disability Discrimination Act.

As Heather pointed out, the navigation buttons are Flash files. This is totally unnecessary, limits web accessibility and makes it more difficult for Google and other search engines to find pages.

Indeed, Charlie Gordon’s stats show that in January 2009, while the home page accounted for 1,792 hits, the eight navigation buttons (which appear on every page of the website, not just the home page) accounted for an average of just 1,324 hits. This is a clear indication that many users are unable to properly navigate through the site. This could be easily diagnosed by a quick look at the statistics (as I have just done), yet no action has been taken to remedy it.

In fairness, there are plain text links at the bottom of the page. But why should users be made to scroll all the way to the bottom of each page just to navigate through the website? The plain text links should instead be at the top of the page, where the Flash buttons currently are.

Charlie Gordon's Webalizer stats
Guess when people became interested in Charlie Gordon’s website?

January’s stats for Charlie Gordon’s website show a huge spike on 23 January, the day the expense claims were released. Traffic did not return to normal levels until the very end of the month, so I will look at the period 1 January 2009–22 January 2009.

Between those dates, Charlie Gordon’s website received an average of just 54.8 visits per day. The maximum was 80 visits on 5 January. The minimum was 0 visits, achieved on both 17 and 18 January. These are outliers, so I assume that the website was down on these days. So even with Charlie Gordon spending £13,000 £1,700 per year, he can not arrange a vaguely reliable service. Considering the website is supposed to be a valuable resource to his constituents, this is a poor show.

The statistics for the whole of December 2008 are not much better. The website received just 63.7 visits per day that month.

We don’t yet know what Charlie Gordon claimed in website expenses for January 2009 or December 2008. But we do know that his largest claim in one month for the 2007–2008 financial year was £1,932.00 in August 2007. The public also has access to his web stats for that month, allowing us to calculate just how much value for money his constituents are getting out of his website.

Charlie Gordon’s website received 561 visits throughout the month of August 2007. This translates to just 18.1 visits per day. It is worth remembering that all Webalizer stats include robots (i.e. non-human visitors) such as Googlebot. As such, all of these visitor statistics are generous estimates!

Making the calculation, we can see that Charlie Gordon spent £3.44 per visit on his website that month. Even if we accept Charlie Gordon’s assertion that the website costs were in fact 20% of what the Scottish Parliament lists, this is still 69p per visit to the website (including robots). This is quite simply extortionate.

For comparison, I will use the same methodology to analyse the costs of my websites. I actually make more money on my websites than I spend on them, but I understand that advertising may not be an option on a publicly funded MSP’s website. So I will look solely at the costs of running my websites. This, too, is slightly flawed because I don’t pay anyone any wages to maintain my websites. This is purely my blood, sweat and tears, and maybe MSPs are too busy to do that. It is, nonetheless, an interesting exercise that will bring Charlie Gordon’s figures into perspective.

For the month of December 2008 I paid £7.67 (and £1.50 of this was a charge for using my debit card) for webhosting to last me for that month. I ran six websites during that month. During that time, these websites received 75,849 visits in total according to the same Webalizer package. This translates to a cost of £0.0001 (one hundredth of a penny) per visit. This is infinitesimal compared to Charlie Gordon’s figure of £3.44 69p per visit.

Incidentally, figures provided by Webalizer are much larger than any figures provided by any other stats package which excludes robots. Google Analytics, for instance, counts only human visits. It says that my sites collectively received 11,184 visits during the month of December 2008. This is just 14.7% of the figure given to me by Webalizer.

Assuming Charlie Gordon receives the same ratio of robot visitors to human visitors as I do, this would give him just 2.7 visitors per day for August 2007. As Heather pointed out, there appear to be no analytics codes installed on Charlie Gordon’s website, so we’ll never know just how many visitors Charlie Gordon gets or got. But it really could be as few as three per day or less.

I am not an MSP who is providing a potentially vital public service to his constituents. Nor am I a professional web designer. Yet I manage to get many more visitors, and spend much less money. £1,700 per year for a website is, quite frankly, a rip off. Taxpayers would have every right to be furious — especially since the taxpayer seems to be less than attracted to his website.

I previously covered Charlie Gordon’s website on the Scotweb2 blog.