Here is a comparison between these three sets of data:
This shows the SBI data. There was a drop on the 11th April by a about 10% for visitors but page views went up (12,560). The figures are pretty stable, almost a straight line. These figures are for the www.pictures-of-cats.org site. This blogger site gets about 3,500 page views per day.
Sorry this image is hard to read. But importantly you can see that "yesterday" the page views figure is .000004% while for the 3 months it is .00005%. The daily figure, the first figure, is 8% of the average figure. So page views dropped by 92%!! That is clearly completely incorrect. Page views are very stable for this site and the figure large enough, surely, to be able to make them reasonably accurate.
People will say that you have to be in the top 100,000 of Alexa to be accurate but being just outside that mark should, I would have thought, allow Alexa to at least provide accuracy better than this. Particularly as it is so important to webmasters who want to progress and eventually get some good advertising.
This shows a 10% Easter drop in page views on Google Analytics. The actual figures are higher as a number of pages don't have the analytics urchin code on them so are not counted. The point is the drop is a tiny fraction of that measured by Alexa.
What is upsetting is that the Alexa figures are the ones that count! Don't get me wrong, I like and am addicted to Alexa! But isn't it possible for Alexa to work with other companies to make figures more accurate for those sites that on the margins of the 100,000 mark or worse? These are often very good sites with great content. After all to get to Alexa 118,000 (the ranking at present for this site) takes a hell of a lot of work. Don't people who work that hard deserve some degree of accuracy? The Alexa page view figures are ruining the Alexa ranking. Because of the inaccurate page view figures the Alexa rank for yesterday was 666,000 or so, a dramatic fall that will hurt the 3 month average (note: things have changed one month later. Alexa ranking is now under 100,000 and getting better for the time being, but what I say above did occur and it is indicative that something was wrong at least at that time).
If Alexa can't make things accurate people won't use them. OK the people who have benefited from the recent changes will love Alexa but the bottom line is that Alexa must be built on accuracy otherwise the whole thing lacks meaning. Is it just for the big boys, who represent a tiny fraction of one percent of the total number of internet users? That can't be correct surely?
Update: things have changed for the Pictures of Cats org website for the better! Not sure why but the Alexa ranking is climbing and page views are too. I have always had difficulty with Alexa but maybe things are improving. If so good on Alexa.
From Alexa Measures Page Views Inaccurately? to Home Page
This comment has been removed by a blog administrator.
ReplyDelete