Privacy Policy

We have updated our Privacy Notice for this website. Please click here to review.

Changing The Internet Audience Measurement Standard

Joint Winner (Media), WPP Atticus Awards 2005

This paper describes a completely new way of measuring Internet audience behavior. By combining a low tech TGI Survey with a high tech user centric panel measurement and a site centric electronic measurement system it allows us to see the surf patterns of a panel of which thousands of target group variables are already known and furthermore the ability to optimize advertisement exposure electronically.


The great thing about the Internet is that almost everything is measurable. Unfortunately the dilemma is precisely that – that everything is measurable. This has at times been in the way of the development of new and more flexible measurement systems.

Media fragmentation, advertising avoidance, technology development and the need for an understanding of ROI are four of the most important driving forces in the media world today. More heterogenic consumers are also pressing the need for more target group data to target and describe consumers.

Media fragmentation forces advertisers to use more media vehicles than before to reach the same result. Advertising avoidance forces advertisers to meet the customers when and where the customer at any moment accepts to receive the advertising. Technology development leads to the rise of digital media and changes in the way people consume media. The pressure on marketing managers to produce ROI figures also brings out the necessity for mixed media planning.

In Sweden, as in many countries, competing ways of measuring Internet audience behaviour, based on different techniques and presenting very different results, have confused the online advertisement market since the birth of Internet. In fact the closest thing to an industry standard in recent years has been the electronic traffic measurements of the total number of unique web browsers visiting any given website during a given week or month. Though such figures may have been better than nothing it is long since recognized that there are at least two aspects limiting their usefulness: 1) they account for the number of computers visiting a website rather than the number of individuals; and 2) they tell us nothing about who is consuming what, since they lack target group information.

The solution to the questions arising from the driving forces noted above is that the media industry needs to address these issues by creating mixed media databases that include a vast amount of target group data and also includes ‘new media’ such as the Internet.

In Sweden there already exists a working single source survey (50,000 respondents) covering all the major media (television, radio, direct mail, print, cinema). Until recently, however, the Internet was not adequately included in the model. This paper will cover the work and show the results on the experimental work that has been done to include Internet in the mixed media model and also turn it into a commercially acceptable product, called ORVESTO Internet.

Download the full report (pdf)

Tools Print page E-mail page Reading Room Get Acrobat Reader


Full report (pdf)
About the authors
  • Peter Callius
    Senior Advisor, Research International, Sweden

  • Anders Lithner
    Research Director, Research International, Sweden

  • Stefan Svanfeldt
    General Manager, Research International, Sweden