Damien just posted ComScore’s Top 60 most viewed sites in Ireland for July 2007. I noticed that the previous data, Websites that Irish people spend most time on, missing at least one significant site – boards.ie with 11 million pageviews a month and time on site in the 15 minute range[1] didn’t even make this Top 50 (it should have been 7th) .
The latest data confirms this – in the “Most viewed” category boards.ie scrapes in at No. 57, whereas the actual figures of ~685,000 absolute unique visitors in the 1st-31st July ’07 period should place it at No. 3!
This leads me to believe that the demographic that this data is pulled from is not as broad reaching as one might hope and should bear major caveats. I’m not denying that the data might still be useful in comparing sites, but be aware of potential omissions, particularly in niche areas and communities.
Update: I see from Damien’s comments that John McCormac and others have flagged other not-so-niche sites like Ireland.com and the Independent missing the party too… I too would love to know more about their methodology.
Update 2: Damien mentioned on his blog that they may have used a dataset of 2000 users. I’d like to know more about the demographics of the set.
Aside: I’m quite familiar with boards.ie as I’ve been a registered user for nearly 10 years now (yikes, I’m getting old)! I’m friends with several of the admin team, and I do not claim to speak for them, so hopefully Damien can get an interview with one of the team to get their official word on the data!
RSS Feed
By Email
Twitter
{ 1 trackback }
{ 3 comments… read them below or add one }
It looks rather strange that high profile Irish sites are missing. This may point to weaknesses in their methodology or a very limited set of data. Getting accurate stats without actual webserver logfile analysis is not easy. From the way they have listed the sites, it does not look like they have access to actual logfiles.
The ComScore figures don’t make sense but the chances are that the press release recyclers in the Irish tech media will run them without question or analysis. These figures really need clarification before being trusted.
Hey John,
I think it’s a strong possibility that they’re using opt-in client side browser snooping on a small subset and extrapolating the data from there. That might explain the serious inconsistencies, especially if they’ve got the demographics wrong.
Cheers,
Alastair.
All rankings of how many visitors sites get should be taken with a gain of salt. They are typically not very accurate. Even log files are not clear-cut, since not everyeone agrees on what a visit is. I’d not worry too much about those things.