SimilarWeb Reality-Check for Stuart Campbell

For about the billionth time, Stuart Campbell has posted a comparison of the alleged readership numbers of the top pro-independence blogs and websites, supposedly to demonstrate what a miserable state the Yes-supporting New Media is in, but in reality to make yet another unsubtle boast about how his allegedly "closed" website is still leading the pack. (It's bloody peculiar that a self-styled "closed" website is regularly updated multiple times per month. Some might describe that as, well, y'know, rather an "open" sort of website on the whole.) I'm actually quite happy when he does this, because it helpfully confirms that Scot Goes Pop remains one of the top sites (and is competitive with multi-author sites such as Bella Caledonia, for example), but nevertheless I think the time has come to warn people that the data Mr Campbell is praying in aid is simply not reliable. 

I of course have access to my own private stats, both on Analytics and the Blogger platform itself, and they aren't even remotely reconcilable with Mr Campbell's numbers, which are taken from SimilarWeb. October actually saw the highest readership numbers for Scot Goes Pop since December of last year, but if you were to believe SimilarWeb, you'd wrongly think traffic had dipped to its lowest level for several months. Even more bizarrely, SimilarWeb claims that traffic was much, much higher in August than it was in September or October, which makes no sense at all, because I was barely posting in August.  My own stats show traffic was in fact 80% higher in October than in August.

The problem, of course, is that SimilarWeb does not have access to the real numbers.  It simply extrapolates from the browsing behaviour of its panellists, which will constitute only a tiny percentage of each site's overall readership.  If that by any chance produces figures that are "within the right ballpark", it can only be a very, very wide ballpark indeed.  Admittedly, some specific aspects of the SimilarWeb rankings seem intuitively plausible - I can imagine Wings may well be in first place and Wee Ginger Dug may well be in a strong second, but if we knew the real numbers, I wouldn't be at all surprised if some of the other rankings were very much wide of the mark.

I'd also be interested to know how SimilarWeb are counting "total visits", because if that isn't supposed to be synonymous with what we used to call "absolute uniques", ie. only counting each reader once per month, it may be giving a wildly distorted impression.  Wings has practically morphed into a readers' chat forum since its "closure", which means the same people are repeatedly "visiting" the same page over and over again to see if more comments have been posted.  To a lesser extent the same is true of Wee Ginger Dug.  That could be leading to a significant overestimate of the gap between Wings, WGD and the rest of the pack. (You might remember that back in the day Political Betting / Stormfront Lite was able to claim to be the UK's "most-read" political site due to its number of page views, even though Iain Dale had three or four times as many actual readers.)

I'd also just wryly note that Mr Campbell seems to have posted the SimilarWeb numbers for October almost instantly after they were published.  Is this the day he lives for every month?

UPDATE: I see SimilarWeb also claims that Wings Over Scotland has between 11 and 50 employees, is based in Glasgow (as opposed to, say, Bath), and produces annual revenues of between $2 million and $5 million.  Oh-kaaaaaaay...

*  *  *

If you'd like to help Scot Goes Pop continue in some form, donations are welcome HERE.