Online audience measurement

We have blogged here before about the problems of measuring an online audience. Traditionally it has been done via page views and often with the use of cookies (small tracking programs inserted on your computed by the servers of the sites you visit which track what else you look at). Interestingly the BBC has investigated the issue, covering recent reports from comScore and Nielsen NetRatings.

The traditional method of measurement has been page views which have always been a bit unbelievable and with an increasing number of PC users regularly clearing their cookies, there is every chance that these are being astronimically overstated. Another measure suggested is by taking a sample audience and scaling up their behaviour; in the same way this is done for TV and radio.

Possibly more reasonable but I would caution using it too precisely. When you are on time unmetered online access where is the onus to read only the pages which are open? I usually have a number of browser panes open at any one time, one of which is normally showing the BBC's news homepage, but I am not reading it all the time!

This is a massively important issue for media measurement and if users are to have faith in the metrics adopted then they must reflect the consumer out-take not the media output.