Of interest to bloggers

This post probably won't be very interesting to general readers, but for my fellow bloggers, it may be useful.

I've been wondering for some time about the accuracy of my visitor count, as shown by Sitemeter (their visitor count is in the sidebar). There seems to be a general consensus among bloggers that Sitemeter undercounts visitors; but no-one seems to be able to state why, or by how much, or to scientifically verify that belief. Google Analytics is very good at identifying unique visitors, but not so good at counting repeat visits, which is, after all, an important part of blog traffic (many readers check in more than once daily).

There are alternatives, other software that counts visitors, but many of the competitors in the field suffer from their own problems. One that I used to like is The Truth Laid Bare (TTLB); but it seems to have all sorts of problems in translating traffic count to ratings. For example, this blog is rated as a 'Large Mammal' in its classification system, as borne out by this listing (taken from a screenshot when I checked TTLB's home page yesterday):




That shows this blog as the 3,968th from the top in terms of links, or well within the top 5,000 blogs tracked by TTLB. However, the display in the sidebar of my blog, which is supposed to reflect that 'Large Mammal' ranking, shows instead that the blog is classified as a 'Lowly Insect' - a category that starts at 100,739 and goes lower from there! E-mails and inquiries to TTLB as to why they can't match their internal rankings with their external displays go unanswered. I think I'll be discontinuing my TTLB ranking as a result.

I'd noticed that according to Sitemeter, from a high point in January this year (almost 90,000 visitors), my blog visit count had been steadily declining. Part of this was due to a lesser number of search engine queries pointing to articles here; that's been a trend across the Web, as the economic recession has reduced the number of people who can use an Internet connection at work or at home. However, I've not seen any reduction in comments or other feedback; in fact, there's been an increase. Also, I knew many readers were getting here via Facebook or another outlet, or using an RSS feed, and they weren't reflected in Sitemeter's figures.

I decided to check up on the number of readers I was getting via the latter sources, and signed up for Google Feedburner to get the statistics. They've been quite eye-opening. Not only have I not been losing readers, I've been gaining them! Let's look at the last week of July as an indication. Sitemeter (measuring visitors who come to my Blogspot page directly) recorded 13,208 visits from July 25-31 inclusive, for an average of 1,887 per day. Google Feedburner (measuring visitors who read this blog via an RSS feed or through another portal, but don't come here directly) showed 7,940 views during the same period, for a daily average of 1,134. Putting them together, that means a combined total of 21,148 visits, or an average of 3,021 per day. Extrapolating that over a full month of 31 days, that gives 93,651 visits for July 2010.

There's another interesting wrinkle to the Google Feedburner figures (or so I was told after e-mailing them to inquire about the distinction). They show both overall visits per day and 'subscribers' per day (i.e. unique users who visit the blog through a feed). For example, on July 31st there were 1,295 'views' of this blog from 525 'subscribers'. The two categories are not the same. The Facebook feed for this blog counts as one unique visitor, or subscriber: but there are many people reading that feed on Facebook, each of whom shows up as a 'visit' but not as a 'subscriber'. The same goes for the Livejournal RSS feed for this blog - there are multiple readers, but only one subscribed feed.

So, it looks like Sitemeter is under-counting my daily readership by a little more than a third of the combined total readership from that source and Google Feedburner. Based on that difference, my total readership counter should now be over two million, rather than the 1.4 million-odd it displays at the time of writing.

Fellow bloggers, you might want to check out your RSS feed statistics. They show an interesting picture of your readership - which might be rather larger than you think!

Peter