Andy Wood, MD, GI Insight You Asked How does one approach the challenge of differing postcode geographies across Europe?

All direct marketing analysis is focused on postcodes. Geodemographic modelling could not exist without postcode geography. Geodemographics is the marketing science based on the “birds of a feather flock together” premise. In other words, people clustered in a small geographical area tend to be of broadly similar means (predicated on their house price/rent), have broadly similar family structures (predicated on housing type) and therefore behave in broadly similar ways, with fairly similar spending patterns. Census data is aggregated to provide an ‘average’ household description for each small postcode area. Of course, because census data is compulsory, it is also confidential, so only these average descriptors are released for commercial usage.

The key factor, for analytical and segmentation purposes, is the number of households represented by the smallest area that each system can express. To give an idea of the disparity of detail that emerges, the lowest postcode level in the UK represents, on average, around 15 households. In other countries, the lowest level can denote anything up to 1,000 households. Therefore any idea of pan-European campaign or CRM strategy integration has to be very carefully thought out.

Nevertheless, once this is understood, a level of cross-border integration can be achieved. The country whose smallest postcode unit has the highest volume of households is taken as the common denominator. Countries with finer postcode units have to be aggregated up so that volumes per unit are roughly equivalent. In this way apples are compared with apples. A number of segmentation companies have developed international geodemographic systems that allow this scale of postcode geography to be compared cross-border.

Analysis then takes place using this common postcode unit size. The findings and segments that are revealed by the analysis cannot, however, be immediately pressed into action. Where a country has postcode units smaller than the common standard described above, then one has to reapply the segments to this finer geography in order to see whether the patterns seen at the larger unit level remain consistent when broken down into smaller units.

If segments are seen to be consistent, then each country can then implement their campaigns. However, although they are using common targeting criteria, they each run their segmentation and selection at the finest (smallest) geographical level that they can, and taking notice of local marketing, data protection and suppression regulations. Response
modelling goes through exactly the same process.

GI Insight