258 The Marketing Book
use products with specific benefits in specific
situations. Table 10.4 provides an example with
respect to the suntan lotion market with four
situations and four target groups. In principle,
four times four products may be designed.
Taking skin colour and skin factors into
account, even more different product formulas
may be marketed.
So far, all of the above segmentation bases
have been derived from anonymized market
research. Targeting would often be determined
by estimating where and how potential seg-
ments might be found and reached. A company
might have all of these details of its customers
in its transactional database, but the discussion
so far has tackled these characteristics from the
more anonymized perspective. Attention is
now turned, however, to data that are more
closely linked with specific individuals. It is
from this point that the new data-driven
marketing, based on emerging marketing met-
rics and the pressures for marketing to be
more accountable, has led to quite different
approaches to segmentation and targeting.
Data-driven segmentation
In order to build up a picture of how marketing
has increasingly moved to technological target-
ing, it is necessary to go back a couple of
decades, to 1981 to be precise. Indeed, earlier it
was implied that 1981 was a watershed year for
segmentation and targeting. This was the year
that the National Census in the UK was used
for the first time in a major and influential
manner in marketing.
Geodemographics
It is a proposition of this chapter that the use of
the UK Census in 1981 was one of the more
significant events in moving from generalized
customer profiles to more individualized
approaches. From that Census, some 40 census
variables were cluster analysed and the emerg-
ing clusters of households led to the creation
of 39 neighbourhood types in the first geode-
mographic system in the UK (ACORN – A
Classification Of Residential Neighbourhoods,
developed by CACI – Consolidated Analysis
Centres Incorporated). Table 10.5 summarizes
the ACORN neighbourhood categories that
could constitute potential market segments for
organizations. ‘Me-too’ versions are now
offered by different companies. For example,
what originally was the credit referencing arm
of Great Universal Stores (CCN – Consumer
Credit Nottingham) produced MOSAIC (which
is now owned by Experian). This (and now
other geodemographic systems) overlays cen-
sus data with financial details.
Table 10.6 provides an example of the
depth of profiling that is possible for each sub-
segment.
Figure 10.1 demonstrates one use of geode-
mographics. It is possible to profile a catchment
area (for example) for the potential citing of a
retail outlet. The MOSAIC (in this case) cate-
gory overlays of the local map shows where
different segments live. If, for example, the
retailer is mainly targeting the ‘stylish single’
segment, the map shows the area of greatest
concentration of this segment. Indeed, names
and addresses of those in this segment can be
purchased in order to target these potential
customers personally.
The basic rationale behind geodemograph-
ics is that ‘birds of a feather flock together’,
making neighbourhoods relatively homoge-
neous. An easy criticism in repost is that ‘I am
not like my neighbour’. Another major limita-
tion of census data relates to the difficulties
associated with updating information, partic-
ularly because in the UK the Census is only
carried out every 10 years. Experian has reallo-
cated approximately 7 per cent of postcodes
and have six name changes in the MOSAIC
typology, both as an update and to improve
clarity of meaning. There are suggestions that
annual updates might be based on survey
research, especially of the ‘lifestyle’ type, which
is discussed in the next section.