TGI and its use in geodemography

Better discrimination and finer targeting are now possible

Richard Silman and Andrew Brown

Readers of Admap will be familiar with the Target Group Index (TGI). The TGI sample is derived from a sample stratified on a rolling ACORN basis: in other words, geodemographics are used to guide interviewers as to which streets to conduct the face-to-face omnibus interviews, the vehicle by which the TGI questionnaire is placed. This article explains how, via the postcode link, TGI respondents can be matched with their relevant geographical cluster group, whilst retaining their anonymity. BMRB's new costing system and approach to the geodemographic market is described. The new arrangements are intended to help clients to use geodemographic data in conjunction with TGI to achieve discrimination and fine targeting.

FOR MANY YEARS NOW it has been possible to analyse the Target Group Index Survey (TGI) on a geodemographic basis. The advent of PC-based survey analysis systems, coupled with major advances in Geographic Information Systems (GIS) has resulted in a marked increase in the level of usage of TGI linked to geodemography. The purpose of this paper is to review such current uses on TGI, and to discuss developments in the market and how they might affect the marketing and media industries.

As a result of TGI respondents giving up their names and addresses, it is possible via the postcode link to match the TGI respondent with the relevant geodemographical cluster group whilst retaining his or her anonymity.

Exhibit 1 shows an example of the discrimination offered by the CDMS Superprofile solution. We have analysed three different target audiences:

  • Owners of hover lawnmowers
  • Heavy users of malt whisky
  • Heavy users of cigarettes
Taking them in order, if we look first at the geodemographic profile of hover lawnmower ownership, we can see significant biases towards their living in groups such as the 'Nest Builders' (123), 'Settled Suburbia' (117) and 'Thriving Greys' (112), while the groups least likely to contain hover lawnmower owners were the 'Have Nots '(68), 'The Urban Venturers' (64) and 'The Senior Citizens' (77). Intuitively, one would not expect the poorer groups to have hover mowers as they may either not be able to afford to buy one, or indeed have no need as they do not have a garden. The most interesting figure pertaining to hover mower ownership is the index of 88 for the 'Country Life' group. It is possible that in these areas where gardens (grounds ?) are likely to be significantly larger than average, the owner might either use a gardening company to mow his lawn or have a ride-on mower.

EXHIBIT 1. 3 DIFFERENT TARGETS ANALYSED BY CDMS SUPERPROFILE GROUP

OWN A HOVER LAWNMOWERHEAVY USERS OF MALT WHISKYHEAVY USERS OF CIGARETTES
TARGETINDICIES ON ALL ADULTS
Affluent Archivers10112346
Thriving Greys11211765
Settled Suburbia11710581
Nest Builders12310394
Urban Venturers64119108
Country Life8817470
Senior Citizens7791101
Traditional Industrial Heartlands10782119
Struggling Council Tenants9186141
Have Nots6851179
Base: All adults
Source: TGI 1994

In the case of heavy drinkers of malt whisky we see a slightly different profile, key groups being 'Affluent Achievers' (123), 'Country Life' (174) and 'Urban Venturers' (119). Malt whisky drinkers are not found in groups such as the 'Have Nots' (51) and 'Traditional Industrial Heartlands' (82).

Finally, looking at heavy smokers, the analysis shows a rather depressing picture of unhealthy lifestyles in the poorest groups, 'Have Nots' (179), 'Struggling Council Tenants' (141) and 'Traditional Industrial Heartlands' (119).

Information such as this can be used in many ways.

One of the most common uses would be to relate the biases in Geodemographic Profile to the Postal Address File (PAF). For example, one may be wishing to carry out a direct mail shot, informing potential purchasers of a new product launch for a hover mower. The marketer can then upweight postcodes which have biases towards 'Nest Builders' and 'Settled Suburbia' as they are more likely to reap better returns than postcodes that are found in the 'Urban Venturers' and 'Have Nots' groups.

Another potential use for the data in Exhibit 1 might be to use the target of heavy smokers for targeting sites within a poster campaign.

In Exhibit 2 (the Radio Clyde example) we show how Radio Clyde, although not measured as a named station on the questionnaire, might use TGI geodemographics to provide an alternative way of using TGI to sell the medium.

EXHIBIT 2. TARGET: HEAVY USERS OF POTATO CRISPS

Index on all adults
All adults
Heavy commercial radio118
Medium commercial radio86
Light commercial radio94
Glasgow Herald71
Any Regional Daily Morning94
Any Regional Daily Evening91
Base: Radio Clyde (TSA)
Source: TGI 1994

In Exhibit 3, we have broken out the Piccadilly Radio Broadcast Area (TSA) in terms of Mosaic category.

EXHIBIT 3. TARGET: PICCADILLY RADIO TSA

MOSAIC GROUPINDEX ON ALL HOUSEWIVES
High Income Families96
Suburban Semis103
Blue Collar Owners168
Low Rise Council130
Council Flats90
Victorian Low Status122
Town Houses and Flats70
Stylish Singles19
Independent elders69
Mortgaged Families41
Country Dwellers7
Institutional Areas0
Base: All Housewives
Source: TGI 1994

There are three key groups who emerge with disproportionate incidence compared with the national profile: 'Blue Collar Owners' (168), 'Low Rise Council' (130) and 'Victorian Low Status' (122). Having determined the key groups it is then possible to analyse these groups in relation to their propensity to consume various products.

In Exhibit 4 we have created a file of all the food categories on TGI and extracted the element defined as Heavy Usage. (Heavy users would typically account for about 80 per cent of volume consumption in any given market The table is ranked on an index demonstrating the propensity for 'Blue Collar Owners' to be heavy users of food types, top categories in this case being ready to eat desserts, instant snack meals and tinned milk puddings. In summary, the geodemographic system provides a linking mechanism between the medium and product and brand usage, where single source data have not been collected.

EXHIBIT 4. TARGET: BLUE COLLAR WORKERS

HEAVY USERSINDEX ON ALL HOUSEWIVES
Ready-to-eat Deserts135
Instant Snack Meals124
Tinned Milk Puddings122
Tinned Ham/Luncheon Meat116
Cakes and Gateaux116
Low Fat Dairy Spreads116
Milk Drinks115
Yoghurts115
Baked Beans112
Stock Cubes110
Base: All Housewives
Source: TGI 1994

Despite BMRB's initial involvement with the geodemographics industry, in the early days all subsequent work was done on an arms' length arrangement. BMRB were responsible for coding the relevant segmentation to each and every respondent on the TGI, and this information was then available on data tape. With the development of PC systems it thus became increasingly easy for clients not only to run cross tabulations, but also run cross tabulations by geodemographic coding. The ease of use provided by BMRB has resulted in the TGI being the source of countless geodemographic cross-tabs since 1979. These cross tabs are used to derive profiles and penetration of product, brand or media usage within each neighbourhood type.

The arrangement that BMRB made with the geodemographic companies was to treat them as 'media owners'. Thus, in return for the payment of an upfront fee, their 'title', or in this case the geodemographic coding, was included on the TGI. Once included, it was then possible for the geodemographic company to run data via the BMRB PC system CHOICES, at their leisure. As the use of these data became more and more extensive, BMRB became more remote from the actual usage of the data. This, combined with the high upfront fee, resulted in no additional codings - from either new players such as Infolink (Define) or Euro Direct (Neighbours and Prospects) or specific codings (such as Scottish Acorn) being added to the TGI.

It was this situation that lead BMRB to a fundamental rethink of its strategy for dealing with this area and during the course of 1994 we have been developing a new approach to the geodemographic marketplace. This has led to some significant changes both in our relationship with the geodemographic companies and with regard to the segmentations now available on TGI.

The major change implemented by BMRB has been to halve the cost of coding the initial segmentation and to halve the cost again of additional segmentations. However, in return, BMRB now levy a royalty on all cross-tabulations done that utilise TGI data. These royalties, complete with identification of the end users, are passed back to BMRB by the geodemographic agency on a quarterly basis. This will now ensure that BMRB has a much closer relationship with the geodemographic agency and is reasonably aware of which clients are accessing TGI information, cross-tabulated by the geodemographic segmentation. An added benefit of this development is that, in return, BMRB is now able to identify which TGI clients are using which geodemographic segmentations.

With a batch of new, competitive area classification systems, the TGI can be usefully employed as a 'vehicle' for comparing the classifiers. This will allow comparison of the degree of discrimination between systems.

A further issue is to be able to translate the target market, derived from TGI, into a mailing target. This may be done by relating TGI to customer databases or lifestyle databases. A further opportunity may well be the use of product/market data from TGI to develop 'market specific' geodemographic segmentations. We believe this area is the biggest single area of opportunity in the area of financial research, and could be customised to individual clients or syndicated within industry sectors.

Exhibit 5 shows an example of the discrimination that can be achieved by using the Financial Acorn alongside TGI. Financial Acorn is a geodemographic cluster solution that incorporates data from NOP's Financial Research Survey, as well as census data. There are few surprises in this table which looks at Gold Credit card holders against the Financial Acorn solution, with key groups being 'Wealthy Equity Holders' (300), 'Younger Spenders' (275), 'Affluent Mortgage Holders' (150) and 'Better off Borrowers' 150).

EXHIBIT 5. TARGET: GOLD CREDIT CARD HOLDERS

FINANCIAL ACORN GROUPINDEX ON ALL ADULTS
Wealthy equity holders300
Affluent Mortgage Holders150
Comfortable Investors83
Better-off Borrowers150
Younger Spenders275
Settled Pensioners92
Working Families25
Thrifty Singles33
Middle Aged Assured42
Older Cash Users25
Low income Unemployed0
Base: All Adults
Source: TGI 1994

Computer technology has had a key role to play in the further development of the geodemographic market. The first companies to start mapping TGI data through GIS were CCN, CACI and CDMS. Having coded their geodemographic solutions on to the TGI survey, they would then manually import the TGI cross-tabulated output into the GIS. The computer would then integrate the data in respect of the biases, and highlight via a colour mapping scheme, where there were high incidences of the cluster groups showing the relative biases against product and brand users. The user can generate a map showing potential sales/usage within any given geography. This in turn can be related to the client's own location data, drive time constraints and sales territories to provide a comprehensive picture of the market geographically. Such systems have proved invaluable to manufacturers, retailers and the media companies alike. In response to this demand for TGI in mapping software, we have invested in software to provide a seamless link from our own TGI analysis software, CHOICES, to GIS mapping programs such as CCN's Micromap, and are currently discussing similar such links with a number of other geodemographic companies.

In addition to the growth of companies specialising in the basic analysis and mapping there have been a significant number of new companies which could be defined as Value Added Resellers (VAR). These VARs differ substantially in their nature, ranging from companies offering bureaux analysis of data (Census, TGI and other) although not possessing their own geodemographic classification, through to companies providing bespoke cluster solutions for clients and gravity modelling. These VARs will be able to access TGI, via the CHOICES software, in the future to further develop analysis techniques.

For some time the TGI has been used to assess store development potential, branch performance analysis and, more recently, gravity modelling. The latter is sometimes known as 'Impact Analysis': and the technique can be used to assess the 'gravitational pull' of some stores or centres, against others. It deals with 'flows' of shoppers from residential areas into these competing stores or centres, and estimates the relative pull of each centre. The technique will generate estimates of turnover for particular centres or outlets, but its main application, as its name suggests, is for modelling impact of proposed changes in shopping geography.

In conclusion, we believe that the geodemographics market will grow strongly, spurred on by the release of the new census data, and the market entry of new census agencies and data suppliers to join the existing protagonists. The number of classification systems will therefore increase substantially.

Additionally, there will be more 'customised classifications' for markets, client companies and products. This will be facilitated by the dramatic development in hardware and software which have reduced the costs and complexity of developing such products.

We believe that market sectors such as fmcg will embrace the use of geodemographic data more wholeheartedly in the future. The availability and application of TGI in helping clients address this area will have been made substantially easier via the new arrangements set in place.

NOTES & EXHIBITS


Richard Silman

Richard Silman

Richard Silman was at AGB for eight years, working on all aspects of TV audience research, in time becoming director of TV research. He is now managing director of BMRB International. In addition to his responsibility for the full suite of TGI products he is also in charge of BMRB's omnibus and media ad hoc surveys, and the development of PC systems for TGI, NRS and TV.
Andrew Brown

Andrew Brown

Andrew Brown is currently the Director of Client Service and Sales for TGI, managing a team of executives promoting the TGI portfolio of services. He also coordinates the cinema industry research currency, CAVIAR. Prior to joining BMRB, he worked in the marketing and research department of Anglia Television and spent a year in AGB's client service operation.