Segmentation of Consumers according to Type and Level of Engagement with Electronic Communications and Technologies, 2004

<p>Abstract copyright UK Data Service and data collection copyright owner.</p>This project was designed to identify the principal general categories of consumers in Great Britain, on the basis of their different levels of engagement in and use of electronic information technologies. A key purpose of the classification was to allow surveys and other data files to be coded with a consumer segmentation relevant to this particular area of behaviour. Around 40,000 electors in Great Britain were thereby allocated to one of 23 distinct clusters in the data. The classification system was built using a series of publicly-available data items known about virtually all British electors. These have been combined in such a way as to optimise discrimination across a series of 80 different measures of consumer behaviour sources, taken from a lifestyle questionnaire database. Thus, the classification was compiled using results of some 80 response codes on over 500,000 lifestyle questionnaires collected over the three years prior to the project. <br> <br> The classification system is in the form of a look-up table relating the cluster codes to a set of values on two multivariate classification systems, one at the person level (<i>Pixel</i>), the other operating at the postcode level (<i>Mosaic</i>). The information used to compile the classification included: length of residence and household composition obtained from the electoral registers of Great Britain local authorities; estimates of age based partly on attributes from said electoral registers; textual parsing of information from the Post Office <i>Postal Address File</i> (PAF); and names and addresses of company directors supplied by Companies House and from a number of shareholder registers. Because of their proprietary nature, these source files are not included in the dataset.<br> <br> An arrangement has been negotiated between the Economic and Social Research Council (ESRC) and Experian Ltd., whereby Experian will append their proprietary <i>Mosaic</i> and <i>Pixel</i> codes to survey respondent and/or customer/client records. The look-up table contained in this dataset thus allows the cluster code to be derived from these two data items. As a result of the annual updating of Experian's electoral register, there should be no obstacle to the coding of files containing new electors or those who have recently moved house.<br><B>Main Topics</B>:<br>The data include four variables, comprising <i>Mosaic</i> and <i>Pixel</i> classification codes, and codes denoting likely access to and ability to use electronic communications and technologies. See documentation for further details.<br> <br> Standard Measures:<br> <i>Mosaic Classification,</i> (1991), produced by Experian Ltd