Geodemographics is the “analysis of people by where they live” (Sleight, 1997, p. 16). It is the suggestion that WHERE you are, says something about WHO you are; that knowing where someone lives provides useful information about how that person lives. To quote some product advertising, it is the possibility that “we know who you are, because we know where you live.” The figure illustrates this link between people and places. It is a simple idea – one that has shown itself to be of commercial value and the catalyst of a rapidly growing and globalizing industry.
THE USE OF GEODEMOGRAPHICS
We are neither the first to take an interest in geodemographics; nor, we hope, will we to be the last! Over a decade ago Brown (1991, p. 221) commented that,
“[g]eodemographics has come into use as a shorthand label for both the development and the application of are typologies [neighbourhood classifications] that have proved to be powerful discriminators of consumer behaviours and aids to “market analysis”.
The “proof” is found in the increased value of the geodemographic market. In Britain this was estimated at a value of £25 million in 1992 (Sleight, 1997, p. 15, citing Mitchell, 1992). By 1995 the same market was valued at £54 million. In 1998, Directions Magazine (www.directionsmag.com) reported its “conservative estimate” of 20,000 companies in the USA and Canada using commercial neighbourhood classifications as part of their marketing information. Weiss (200) reports that US marketers spend an estimated $300 million annually on nation’s 100 million households: “cluster-based marketing has gone mainstream and is now used by corporate, non-profit, and political groups alike to target their audiences” (p. 4). The market has continued to evolve, the most recent stimulants being the release of twenty-first-century census data and the emergence of extensive “data warehouses” associated with a growing trade in consumer-oriented data.
It is also over 10 years ago that Leventhal (1993, p. 223) recognized the potential of geodemographics to inform strategic marketing planning and communications, presenting examples of its application to the Market Research Society under three main headings:
- Survey design: samples may be stratified or selected using geodemographics, and many large-scale surveys take advantage of this facility.
- Retail planning: knowledge of the types of people living in catchment area can be a key ingredient in understanding store performance and the same information can help in deciding a store location.
- Direct marketing: the selection of prospects (prospective customers) can be improved by using geodemographics, whether for direct mail, “door-to-door” distribution or sales calls.
USING GEODEMOGRAPHICS FOR RETAIL TARGETING
A retail company can take its client list and sort it into different types of consumer, making its judgement by where the client lives. The sorting first begins by linking the address of each client to a predetermined classification of the type of area that address is found in. As a consequence, the clients are segmented into groups not actually on the basis of their own, individual characteristics but according to some sort of social average for the area in which they live – by the type of area in which they reside (this distinction is important and one we return to). The area type is defined by the classification used to sort the consumers into groups. Such a classification would normally be purchased from a third party data vendor. A “look-up” file allows the retail company to determine in which type of neighbourhood each of its customers lives.
HOW IT WORKS: A SHORT THEORY OF GEODEMOGRAPHICS
In seeking to define the nature of geodemographics, one of us (Harris, 2003, p. 225) has suggested it is “the analysis of socio-economic and behavioural data about people, to investigate the geographical patterns that structure and are structured by the forms and functions of settlements”. Taking time to consider what that long-winded statement actually means sheds light on why neighbourhood classifications may be useful, predictive tools, for analysis and decision taking.
The relationship between places and people is neither one way nor nolely the consequence of external factors. When people speak of “their neighbourhood” or “their community” they do so in a way that suggests an attachment to place. Harris (2003) implies that there is an interrelationship between people and places – the link is illustrated by the figure. Therefore, the physical, social and economic properties of settlements in some way reflects the character, choices, preferences, ideals, affluence, consumer lifestyles (and so forth) of past and present populations living in those settlements but also are a consequence of governmental policies, for example in respect of planning controls and social housing initiatives. Because a place usually pre-dates the residents so the relationship is two way: the style and character of the settlement “draws in” certain population groups, perhaps by choice, perhaps by necessity; those residents then shape further evolution of the area. Longley and Batty (1996, p. 76) write that:
“[t]he behaviour of individuals in [geographic] space together contribute to the development of places over time and these place effects in turn condition subsequent spatial [geographical] behaviour.”
The interrelationship suggests that measures of the physical social and economic properties of settlements can yield useful information about the characteristics, preferences and lifestyles choices of the populations resident within those settlements, because people land places are dependent on each other.
These theoretical ideas are summed up by the adage “birds of a feather flock together”. This, according to Flowerdew and Leventhal (1998) is the basic tenet of geodemographics. In fact, birds of a feather may not just flock together but also increasingly become alike. This is because very few of us (the birds) live in complete isolation from the rest of the society (even if there are times when we wish otherwise!). It is likely that many of our behaviours, choices, aspirations and ideals are influenced by those with whom we interact in our everyday lives (and vice versa) and to assume otherwise is known as the atomistic fallacy. Despite the emergence of cyberspace and the popularity of online chat rooms or other forms of communication, it remains reasonable to suppose that geographical distance and location impart constraints on who we meet and when. Weiss (2000, p. 25) argues that there is value in classifying populations at a neighbourhood level and it relates to a “core truth… you are like your neighbors.”
This “law” is an expression of what spatial statisticians refer to as spatial autocorrelation (Cliff and Ord, 1973). This type of autocorrelation is present in a dataset if it can pensioners, lone-parent households, eat-out regularly couples or sports-car-owning adults) display a non-uniform and non-random patterning but, instead, are clustered into particular localities.
Geodemographic methods assume (positive) spatial correlation when residents of the same neighbourhood are taken to share, in broad terms at least, some common socio-economic and/or behavioural person living in a certain neighbourhood – the assumption being that proximity is related to similarity. However, Tobler’s first law does need to be modified when looked at in a geodemographic context. The geodemographic methods that are the subject of the book assume not only that proximate populations are related but so also are populations living in the same “class” of neighbourhood. In other words, near and far things are related – by neighbourhood type.
Finally and most pragmatically, the majority of the book is about classifying small areas (“neighbourhood” in a strictly technical sense) and about using geography as the basis for modelling, even inferring people’s demographic, socio-economic and behavioural characteristics. We do not ignore individual or household classifications but the focus is more geographical.
Geodemographics is the analysis of people based on a statistical classification of the area in which they live. The classification aims to capture the important socio-economic “dimensions” of, and differences between, neighbourhoods. The geodemographic approach has been found by many to be useful aid for guiding decision making and the management of geographical information.
Neighbourhood classifications usually are produced by grouping together a large number of usually administrative units into a much smaller number of groups, clusters or neighbourhood types on a like-with-like basis. A common choice of data to define the similarity, or otherwise, of neighbourhoods are national census statistics. Such classifications have been the bedrock of a rapidly growing industry that has its origins in urban geography and sociology. Present applications include survey design, retail planning, planning and decision taking in both the public and private sectors. Neighbourhood classifications can be used to look for geographical patterns in various socio-economic, behavioural, attitudinal or consumer datasets.
The usefulness of neighbourhood classifications derives from the idea that knowing where someone lives provides useful information about how someone lives. A simple theory of geodemographics is there is an inter-relationship between people and places, and also between individuals and the people they regularly meet. The adage, “birds of a feather flock together” and Tobler’s “first law of geography” go some way to explaining why neighbourhood classifications can usefully be applied to extract information about people from information about places. When direct knowledge about potential customers, clients or consumers is not consistent, neighbourhood analysis provides an inferential tool linking what is known to what is not.
- Geodemographics has been described as the analysis of people by where they live.
- It has been widely used to inform strategic marketing and planning.
- Neighbourhood classifications usually are produced by grouping a large number of administrative units into a smaller number of clusters, on a like-with-like basis.
- An assumption is that “birds of a feather flock together” such that populations living in the same neighbourhood type share broad socio-economic and consumer characteristics.
- Multivariate classification techniques simplify a complex geographic reality to make the basis and process of decision making easier, faster and more intelligible to stakeholders.
- Geodemographics is better for exploratory analysis than for hypothesis testing.