Market segmentation 263
Because geodemographics are based on
census data, there is the corresponding criticism
of those systems. In addition, as sophisticated as
geodemographics are – certainly compared with
the simplicity of age, gender and occupation
(the main variables of the demographic alter-
native) – the approach is essentially the same, i.e.
it ‘profiles’ people. The method does not in itself
explain why people behave as they do and
neither does it provide individualized informa-
tion on what people buy. These issues are, to
some extent, addressed by other approaches,
outlined in the sections below on ‘new’ lifestyle
segmentation, transactional data and biographic
segmentation.
Names and addresses cannot be revealed
from the census, but the statistics for enumera-
tion districts can. Census data can be linked
with the postcode database (there is one post-
code for approximately 15 households) and
with the electoral roll (another database); it is
possible to identify individual households and
their characteristics. One of the current debates
(at the time of writing) is between the Informa-
tion Commissioner’s position that the electoral
roll was not collected for marketing purposes
and should not therefore be used in this way,
and on the other hand the marketing industry,
which argues for freedom in its use. Although
the electoral roll is rarely used as a list in itself,
it is used as the base for virtually every
targeting tool and geodemographics started
this process.
One concern of the Information Commis-
sioner is that some people may disenfranchise
themselves by not registering for fear of being
over-targeted by marketers. A legal case has
already brought this to a head. In November
2001, a member of the public won his case
against Wakefield Council after that Council
had not been able to confirm that his electoral
roll data (name and address) would not be
supplied to third parties, such as marketers,
without his consent (Acland, 2001).
The full implications of this case were not
clear at the time of writing, but if the principle
is generally upheld, this could be a significant
strait-jacket for data-driven segmentation and
targeting. Clearly, without the electoral roll it
would be difficult to identify individual house-
holds from census data. One study suggests
that a ban on using electoral roll data could cost
advertisers £55 million per year because ‘the
cost of not having access to the electoral roll
would be five pence per (mail) pack’ (Denny,
1999).
Having said this, it is likely that an opt-out
option will be added to the electoral roll and
this should help to alleviate privacy concerns at
the same time as shifting segmentation and
targeting in yet another (related) direction –
namely based on customer ‘permission’ to use
their personal details. Perhaps the latter will
itself become more sophisticated and customers
will give permission to specific organizations to
use their details for specified purposes.
There have been other developments of the
geodemographic principle, as hinted at above,
and these are discussed in the following cov-
erage of ‘new’ lifestyle segmentation, the use of
transactional data and the fusion of data to
form the new biographics.‘New’ lifestyle segmentation
A more recent development in lifestyle research
and segmentation is the ‘lifestyle survey’,
which has a somewhat different basis. These
surveys essentially ask respondents to ‘tick’
those responses that apply – Table 10.7 demon-
strates some typical questions, some of which
will be sponsored by specific companies.
This table reflects just a portion of typical
current lifestyle surveys. Many more questions
are included, covering claimed buying behav-
iour across many different product and service
categories. Some questions will be sponsored by
specific companies – for example, a car insur-
ance company might sponsor a question asking
for the month in which the car insurance is
renewed. Because these surveys are not anony-
mized, the data will be filed in a database by
name and address of respondent; it is likely that
the month prior to that respondent’s renewal