How accurate are analytics stats?
-
Hi, I suppose you get this question all the time, but I’m hoping there is some data on it please.
I have managed a couple of websites for 10-15 years and used Google Analytics until I didn’t feel happy with it any more. I started using another Analytics package which seemed to give similar results to Google so I was happy with it – except it added so many tables to my database and was way more complex than I needed.
I decided to trial Burst because it looks exactly what I want. Except when comparing it to my existing analytics software over a fortnight of daily stats, Burst Pageviews & Visits are consistently 15-25% below the other. (Visits 80% with SD 0.07; pageviews 83% SD 0.09.) Since it is easier to think visits might be missed rather than false visits invented, it makes me wonder about Burst’s accuracy.
I ran a test where I visited a bunch of obscure pages on my site and counted them all, and both packages were almost exactly right (each missed one visit, I don’t know why). So that test was fine, but the fortnightly test is worrying.
I’m not so pedantic as to expect or require total accuracy, but 20% seems a largish difference. Has anyone done tests to explain what is happening here please?The page I need help with: [log in to see the link]
- The topic ‘How accurate are analytics stats?’ is closed to new replies.