The approach – believed to be one of the first of its kind – is aimed at providing online advertisers with a fast and accurate evaluation of the success of online advertising and the composite effect of cross-media campaigns. In an unique one-step process that does not rely on ‘pop-ups’ appearing on the website, a survey of those who have been exposed to an online ad is served on exit from the site hosting the advert. In addition, a number of visitors who have not been exposed to the ad are also interviewed as a comparison. Visitors can choose to ignore the survey. However, if they respond, their feedback is based on the certain knowledge that they have seen the banner advertisement. This, in turn, increases the integrity of the data and gives it increased validity as a measuring tool. According to Arno Hummerston, Business Development Director, TNSi UK, "Measurement of the effectiveness of banner advertising to date has been a relatively hit and miss affair. Our approach and methodology are unique in that they ensure that responses on specific adverts come only from those visitors who have actually been served the specific banner ad. This greatly increases the integrity of the feedback for advertisers and no longer relies on the pre- and post-testing methods of the past. This service will help give online advertisers – and online advertising sales executives – far greater confidence in their planning and much more faith in the feedback they receive. The evaluation of online advertising, and the resultant effect on the advertiser’s brand, is a complex and largely undeveloped area. The insight into the opinions and attitudes of the actual audience of adverts that this service provides is something we believe will restore much of the confidence that has been eroded by low click-through rate publicity in recent months." In addition to this quantitative research methodology, the service also allows qualitative, "focus group" style questions to be asked, and post-advert exposure behaviour tracking that allows retention and delayed action responses to be monitored.