Modelling and Evaluating Posters in a Multi-media Environment

Peter Masson
Bucknull and Masson, United Kingdom


Economy of personal interviewing normally dictates that random samples are drawn in stages involving geographic clustering of respondents. Sampled informants within such clusters will not be representative, in terms of their movements past the (un-clustered) 'population' of all poster sites – even though they may be a good demographic representation of the general population. Poster reach and frequency estimates based on such clustered samples may severely under-represent the real position.

Consequently measures of poster site exposure (in reach and frequency terms) have largely been site centric models of aggregate behaviour. The downside to this is that there is only a very limited opportunity to evaluate poster exposure in combination with other media types or to evaluate poster performance within detailed target groups. In today's environment, requiring a more holistic (360° planning) approach to campaign planning and evaluation, this site centric approach places considerable constraints on the way that the poster medium can market itself.