You are on page 1of 10

GeoJournal 45: 255–264, 1998.

© 1998 Kluwer Academic Publishers. Printed in the Netherlands.


255

Classifications of retail stores and shopping centres: some methodological issues

Clifford M. Guy
Department of City and Regional Planning, Cardiff University of Wales, P.O. Box 906, Cardiff CF1 3YN, Wales, UK

Received 20 May 1998; accepted in revised form 14 October 1998

Key words: classifications, consumer behaviour, retail outlets, shopping centres

Abstract
This paper reviews current methods used in classifying retail outlets and areas devoted to retailing, in the geographical
and town planning literature. For retail outlets, classifications based upon types of goods sold, and types of shopping trip,
are discussed. This is followed by an analysis of modern large store development which reflects property developers’ and
retailers’ concerns. Areas devoted to retail uses are subdivided into unplanned ‘retail areas’ and planned ‘shopping centres’.
Traditional classifications based upon central place theory are reviewed for both of these types, and found wanting in the light
of recent changes in retail development practice and consumer behaviour. Classifications based upon physical development
characteristics and type of shopping trip are recommended. Finally, classifications of urban retail location are examined.

Introduction shopping developments in the various countries concerned


are classified in a consistent way (Guy, 1998a). Consistency
A familiar issue in geographical studies of retail devel- is also important when one is analysing phenomena such as
opment and change is the classification of retail outlets. retail internationalisation, or long-term changes in consumer
Classification is essential as a means of understanding and shopping behaviour within a region.
analysing relationships in the world of retailing. The inten- The paper proceeds as follows. The first major section
tion in this paper is however not to revive the rather arid discusses what is meant by a retail outlet or ‘store’, and
debates about ‘central place’ systems typical of retail geog- then explores some methods of classification. These draw
raphy up to the 1980s (e.g. Christaller, 1966; Berry, 1967; upon established notions of consumer attitudes and retailing
Beavon, 1977; Beaujeu-Garnier and Delobez, 1979), al- organisation and method. The following section discusses
though these matters will be referred to in passing. Instead, the more complex issue of classifying and defining ‘shop-
the treatment of classifications is founded largely in the more ping centres’. The question is addressed first of whether all
recent geographical literature relating to three aspects of re- retail outlets are eligible for inclusion in systems of shop-
tailing: consumer behaviour and choice; retailer strategy; ping centres, or whether the term implies certain minimum
and property development and town planning (e.g. Bromley standards of size, internal organisation, planning or owner-
and Thomas, 1993; Brown, 1992; Davies, 1976, 1984; Guy, ship. Various ways in which centres have been classified are
1980, 1994; Jones and Simmons, 1990; O’Brien and Harris, then discussed, moving from ‘classical’ central place sys-
1991; Wrigley and Lowe, 1996). tems to more recent approaches based upon aspects of retail
Why is it necessary to classify retail outlets? Firstly, a performance, consumer perceptions, development history
logical and appropriate classification should assist in sys- and location. Some classifications which attempt to combine
tematic and well-informed discussions of research in retail these dimensions are also examined. The final section em-
geography, whether of development trends, institutional in- phasises that there is no one system of classification which
fluences, retailer behaviour or consumer behaviour. This is universally applicable, even within one geographical area
suggests in turn that the means of classification involved and time period.
should be influenced in some way by the perspective of the
researcher and the objectives of the research. For example,
when assessing elements of consumer behaviour in a partic- Retail outlets – definitions and classifications
ular region, the classification of retail outlets should respect
both common practice in that region and the main criteria A retail outlet can be defined as a building from which retail-
which shoppers use when choosing where to shop. ing is carried out. In order to exclude buildings concerned
Secondly, a consistent and comprehensive classification solely with mail order sales, etc. a retail outlet should nor-
system allows researchers to compare and contrast empirical mally store retail goods which can be sold to members of the
findings across a variety of spaces, cultures and time periods. public from the premises, without prior appointment. While
It is easier, for example, to compare the impacts of differ- this general principle is widely accepted, variations can still
ent land use planning systems on retail change if modern occur in practice. For example, in the United States, motor
256
Table 1. UK government classification of retail goods Table 2. UK government classification of retail businesses

Convenience goods Food retailers


Food Large grocery retailers
Alcoholic drink Other grocery retailers
Tobacco Dairymen
Other goods: Butchers, poulterers
newspapers and magazines Fishmongers
cleaning materials and matches Greengrocers, fruiterers
Bread and flour confectioners
Comparison goods
Clothing and Footwear Drink, confectionery and tobacco retailers
Do-it-yourself goods Retailers of confectionery, tobacco and newsagents
Household goods Off-licences
furniture, pictures, etc.
Clothing, footwear and leather goods retailers
carpets and other floor-coverings
Men’s and boys’ wear retailers
major appliances
Women’s, girls, children’s and infants’ wear retailers
textiles and soft furnishings
General clothing businesses
hardware
Footwear retailers
Recreational goods
Leather an travel goods retailers
radio, television and other durable goods
television and video hire Household goods retailers
sports goods, toys, games and camping equipment Household textiles retailers
other recreational goods Carpet retailers
books Furniture retailers
bicycles Electrical, gas and music goods retailers
Other goods Hardware, china and fancy goods retailers
pharmaceutical products and medical equipment Do-it-yourself retailers
toilet articles and perfumery Other non-food retailers
jewellery, silverware, watches and clocks Chemists
other goods Newsagents and stationers
Booksellers
Source: Unit for Retail Planning Information (1994).
Photographic goods retailers
Toys, hobby, cycle and sports goods retailers
Florists, nurserymen and seedsmen
vehicle sales are included in data relating to retail outlets,
Non-food retailers (not elsewhere specified)
whereas this is not the case in the UK. Some sources may
include within ‘retailing’ certain consumer services which Mixed Retail Businesses
do not always involve the physical transfer of goods, such as Large mixed businesses
hairdressing or travel agency. Other mixed businesses
In classifying retail outlets, several systems have been General mail order houses
used, each relating to some physical or economic character- Hire and repair businesses
istic. Four of the most common are discussed below. Television hire businesses
Other hire or repair
Classification by type of goods
Source: Business Statistics Office (1992, Table 2).
A classification in everyday use relates to the types of goods
sold by retail outlets. Many widely recognised types of shops
sell coherently organised ‘bundles’ of goods, such as furni- sources of hybrid categories such as ‘variety store’ or ‘large
ture or shoes. A typical official classification of such bundles mixed retailer’. The classification of ‘retail businesses’ in
of goods is shown in Table 1. use until recently in the UK is based mainly upon types of
This approach causes problems however when shops sell goods sold, but also includes some hybrid categories (Ta-
a variety of goods which apparently have little in common. ble 2). A further problem with this type of listing is that
For example the modern food superstore (in the UK) or hy- some goods are sold by several types of retailer: for ex-
permarket (France) might sell cooking utensils, children’s ample, newspapers are typically sold in Britain by ‘large
clothing, and compact discs as well as the food and groceries grocery retailers’, ‘retailers of confectionery, tobacco and
which are its main type of merchandise. The store might also newsagents’, ‘newsagents and stationers’, and ‘booksellers’,
offer services such as a pharmacy, post office, cafe, banking using categories listed in Table 2.
or dry cleaning. Other shops may deliberately sell as wide a Another general problem with goods-based classifica-
variety of goods as can be achieved in the space available, tions is that any one category of outlets may include dis-
ranging from the humble ‘corner shop’ to the mighty depart- parate types of shop which are regarded as dissimilar by the
ment store such as Harrods. This leads to the use in official general public and are not really in competition with one
257

another; for example, Fortnum and Mason (luxury foods), Arguably one could dispense altogether with the idea of
a Sainsbury store (mainstream groceries), a Netto store the ‘comparison outlet’ and instead use a distinction based
(limited range discount) and an independent greengrocery upon motivation for purchases. For example one can eas-
(specialist fresh foods) may all be included in a category of ily distinguish ‘household goods’, which are needed for
‘food stores’. home improvement and maintenance, from ‘personal goods’
which are bought for purposes of self-improvement or grat-
Classification by shopping trip purpose ification. Examples of the former are home improvement
goods and furniture; of the latter, clothing, footwear, books
For many years a distinction has been made in the mar- and CDs. Household goods are nowadays often bought in
keting literature between shopping trips which are intended large free-standing stores, known as retail warehouses in
for purchasing routine necessities and those which are more the UK (Hillier Parker, 1994) and fachmarkts in Germany
occasional in nature and directed more towards personal (Kulke, 1996), where a very wide range of brands and styles
gratification. The well-known classification of goods as con- are available. The shopping process might still be described
venience or comparison underpins Table 1. This distinction as ‘comparison’, but the methods of display and in-store
has often been extended to shops (Brown, 1992, pp. 21–22). service which have proved so successful in the convenience
Convenience goods have been defined as ‘goods which sector are applied. Personal goods, on the other hand, still
the consumer usually purchases frequently, immediately and tend to be sold in smaller stores, often grouped together to
with a minimum of effort’ (AMA, 1948). A convenience aid the process of comparison. The terms household shop-
outlet1 should therefore attract shopping trips which are ping and personal/fashion shopping are thus used in this
short in length and frequently made. The shopper uses the paper.
store on a routine basis and may often know exactly what
he/she intends to purchase there. The supermarket is the Classification by size and type of store
most obvious example, but other types of store which attract
routine trips from a mainly local catchment area, such as One of the most discussed features of retailing in the last 40
fresh food shops, newsagents, tobacconists, chemists and years or so has been the development of large stores, usually
post offices are often classed as convenience outlets. selling either convenience or household goods. These stores,
Comparison shopping has traditionally involved an el- which are generally on one level and provided with ample
ement of personal gratification: comparing goods between car parking space, have grown up in the suburban areas of
different shops is part of the process of choice, which is many north American and European cities: in convenience
made on the bases of ‘suitability, quality, price and style’ retailing, the supermarket, superstore and hypermarket3 ;
(AMA, 1948)2 . Comparison outlets are thus the destination in household goods retailing, the retail warehouse or fach-
for occasional shopping trips, often carried out in family markt, also known as ‘non-food superstore’ or ‘big box
or other groups, which are held to be more enjoyable than store’. These terms all denote a certain simplicity in store
routine convenience shopping. Traditionally the compari- construction and design: building and maintenance costs
son outlet has been small in size and emphasised quality of need to be kept low, and the interior functions rather like a
goods, awareness of fashions, and service to the shopper, warehouse with priority given to clear signposting and ease
rather than a wide range of goods or low prices. Clothing, of restocking the display shelves.
footwear and personal accessories such as jewellery are the A problem with much of the recent literature in retail
most commonly cited forms of comparison outlet. Compar- geography is that when discussing types of store it reviews a
ison shopping has also for many years been associated with limited number of large purpose-built outlets while ignoring
the department store, which began to appear well over 100 the more traditional small retail outlet. These are much too
years ago in major European and north American cities. varied in nature to fall into one category of ‘small shop’. If
The notion that comparison shopping is somehow dis- we attempt to relate shop type and size to the typology of
tinct from convenience shopping and that different types shopping trips discussed above, then the role of small shops
of retail outlet are involved has nowadays become rather becomes clearer. An attempt to classify retail outlets along
problematic. Firstly it is not unreasonable to state that super- these lines is shown in Table 3.
market shopping for weekly family food supplies involves Any classification by size, such as that employed in Ta-
comparisons on the bases of ‘suitability, quality, price and ble 3, can however be problematic. For example, should
style’. Secondly, it is often the case nowadays that compar- one use sales area (also known in the UK as net floor area),
ison shopping takes place within one large specialist outlet or total floor area (gross area or gross lettable area)? Food
rather than between several small ones, as for example in retailers tend to provide information based upon sales area
furniture or carpet retailing. whereas shopping centre developers and owners usually pro-
1 The term ‘convenience outlet’ is used here rather than ‘convenience store’,
vide gross area data. Ways in which ‘net’ and ‘gross’ area
are themselves defined can vary. Generally, sales area is
since the latter term has a more precise meaning in the English-language
literature: a small store (sales area typically 100–200 m2 ) selling a variety more indicative of the volume of merchandise in a store, but
of convenience goods and services to a local catchment population. 3 In most European countries the term ‘hypermarket’ is used instead of su-
2 The American Marketing Association actually used the term ‘shopping
goods’ in this connection, but this term has (ironically) gone out of fashion. perstore, and statistical sources generally use the term for any food store
of over 2500 m2 sales area.
258
Table 3. Examples of shop types classified by trip purpose and size

Sales area (m2 ) Convenience Household Personal/fashion


shopping shopping shopping

Under 250 Convenience store Fashion boutique


Butcher Shoe shop
Pharmacy

250–1000 Small supermarket Hardware store Bookshop


Video hire Sports goods shop

1000–2500 Large supermarket Retail warehouse


Fachmarkt

Over 2500 Hypermarket Retail warehouse Department store

is often difficult to measure. Gross area can be measured classification, for example the food superstore or hypermar-
(approximately) from large-scale maps. ket, which can be defined by its range of merchandise, own-
Classification by size also implies the use of arbitrary ership type, size and internal design, or some combination
boundaries. For example, hypermarkets are generally recog- of these attributes.
nised in Europe as having a lower limit of 2500 m2 sales For most purposes, particularly examinations of retail
area, but in Germany a lower limit of 1500 m2 is some- structure at national or local level, a classification based
times used. In the UK, a lower limit for food superstores upon types of goods sold seems most appropriate. This can
(the equivalent of hypermarkets) of 25 000 ft2 (2323 m2 ) is with advantage be grouped according to shopping trip pur-
often used. pose, as suggested in Table 3. In many investigations, a
second dimension, probably ownership type or size, needs
Classification by Store Ownership to be used.

Another common method of classifying retail outlets is ac-


cording to the means of ownership. The simplest distinction Shopping centres: definitions and classifications
is between independent and multiple retailers, the former
owning one shop (or a small number of shops), and the latter Classifying shopping centres is a much more complex and
owning several or many shops. The dividing line between uncertain business. Indeed, much of the debate in retail ge-
the two categories often varies according to the country ography up to about the end of the 1970s was concerned
concerned or even the statistical series used. Intermediate with definitions and classifications of ‘shopping centres’,
categories involving affiliation to a particular supplier, or ‘business centres’, ‘retail areas’, etc. An initial problem in
franchising, complicate the situation. The status of the co- empirical work is often to define the physical limits of a par-
operative movement is not clear – it may sometimes have its ticular centre, or indeed to decide whether a group of shops
own category in retail classifications. is a ‘centre’ at all. One school of thought is that the term
Nevertheless, store ownership type is a vital dimension in shopping centre should be applied to any group of shops,
retail geography. In European statistical series the following whether old, new, planned, unplanned, purpose-built or con-
classification is used (Euromonitor, 1996): verted. The opposite view is that the term should be confined
• Co-operatives. to planned, purpose-built retail development (as for example
• Department and Variety Stores, and Mail Order. in texts such as Dawson, 1983; Dawson and Lord, 1985). In
• Multiples. this paper this view is accepted, and unplanned clusters of
• Voluntary Associations and Buying Groups. retail outlets are referred to as retail areas.
• Non-affiliated Independents.
In some cases, ‘multiples’ may further subdivided, as in Classifications of retail areas: the central place hierarchy
the UK where a distinction is made between retail firms with
10 or more outlets (‘large multiple retailers’), and those with Unplanned retail areas have typically grown in a haphazard
2-10 outlets (‘small multiple retailers’) (Business Statistics manner, often through gradual conversion from other land
Office, 1992, Table 2). uses. An unplanned area is likely to have a multiplicity of
property owners, and to consist of many separate buildings
Classifications of retail outlets – conclusions which are varied in physical appearance. Unplanned retail
areas have generally developed in two ways. In the historic
It is now clear that several means have been used by geogra- centres of towns, they have grown from origins as periodic
phers and planners to classify retail outlets. Some systems markets. In suburban areas, they have grown through con-
are fully inclusive while others are not. Some generally versions of residential or other property, in order to serve
recognised types of outlet imply a number of dimensions of
259
Table 4. Classification of central areas in UK, 1989
Such analyses must however be criticised on several
Grade Multiple branch Number of Cumulative grounds. Firstly, designation of a hierarchy suggests that
score centres total there are clear distinguishing characteristics for each level,
when in fact empirical analysis suggests that a continuum
I National 186 1 1 exists between small and large retail areas within most cities
II Metropolitan 93–132 6 7 (e.g. Davies, 1974; Potter, 1981). Secondly, retail and ser-
IIIA Major regional 35–74 99 106
vice ‘ribbons’, particularly in European towns, have usually
IIIB Minor regional 25–34 60 166
grown to serve the needs of adjacent communities rather than
IVA Major district 6–24 290 456
those of highway users, as Berry appeared to suggest.
IVB Minor district 2–5 370 826
A third criticism, posed by Davies (1984), Dawson and
Source: Reynolds and Schiller (1992), Table 3. Sparks (1986), and other writers, is that large modern retail
stores which are specialised in purpose, such as hypermar-
kets and retail warehouses, serve fairly extensive catchment
neighbouring areas with convenience or occasionally com- areas and are difficult to fit into conventional hierarchies of
parison goods (see Guy, 1994a, Chapter 6 for a more detailed retail areas. This is because, compared with traditional retail
discussion). areas of comparable size, they attract shoppers from a wider
Much effort has been expended by geographers in clas- catchment area but take a smaller market share. This reflects
sifying such retail areas. Typically such classifications have their overwhelming reliance on private car travel, and the
attempted to relate retail and business areas to their catch- notion that car users are less likely to discriminate between
ment population. Broadly speaking, two geographical scales shopping opportunities solely on the basis of travel distance.
are used in this type of analysis: firstly, national or regional; Many surveys have shown that consumer choice of shop-
secondly, urban. ping destination reflects several qualities, such as variety of
National or regional classifications are made in an at- goods, price of goods, cleanliness, spaciousness and security
tempt to determine the relative status of central shopping of the centre, and quality and quantity of car parking provi-
areas (town or city centres). Each area is classified according sion, for example. Central place theory, which relies mainly
to some index of retail ‘size’ or ‘importance’. Volume of upon distance as a choice criterion, becomes inadequate in
non-food sales would be a typical indicator. If such data this situation.
are not available for recent years, as in the UK, ‘points’
systems have to be used. An example is the system used Classifications of shopping centres by size and function
by the consultants Hillier Parker in the UK which awards
‘points’ according to the number of major non-food mul- As mentioned above, the central place model has also been
tiples’ outlets located within the central area (Schiller and used to classify purpose-built shopping centres. This is more
Jarrett, 1985; Reynolds and Schiller, 1992). It is then neces- typical in north America, where planned centres rather than
sary to allocate central areas to ‘levels’ in the classification. retail areas are the norm in most urban areas. Neighbourhood
The classification and associated points score used in these and community centres serve local residential populations
papers is shown in Table 4. and predominantly feature convenience shops, whereas re-
While there are always some problems in this kind of gional centres provide a much wider range of goods and
analysis, particularly relating to definitions of ‘levels’ or services and are located close to major road intersections in
‘grades’ in the classification, the use of central place analysis order to serve a larger catchment population.
at national level is relatively straightforward. Just as the central place model of retail areas is now seen
In contrast, use of central place analysis within urban as inadequate, so is the equivalent north American model
areas has been rather problematic. It was pioneered in the of suburban shopping centres. Although some types of large
late 1950s by a group of geographers working in the United stores such as supermarkets, discount stores (e.g. WalMart),
States (e.g. Berry, 1963). A sequence of local centre, neigh- and ‘junior department stores’ are probably located within
bourhood centre, district centre and town or city centre was planned centres to a greater extent than is found in Europe,
commonly recognised4 (see Table 5). In Britain and else- others such as specialist non-food stores and club discount
where, town planners used such classifications in respect stores have tended to locate along major highways to form
of both existing retail areas and future planned centres (e.g. sprawling ‘retail strips’ (Lord and Guy, 1991). More re-
Burns, 1959). cently, an important trend in north American retailing has
Berry’s (1963) analysis of Chicago was probably the been for these ‘big box’ stores to cluster into large-scale
most detailed and sophisticated of its type during this pe- power centres (Toderian, 1996), similar in nature to the
riod. As well as recognising a hierarchy of unplanned retail British retail parks which are discussed below. These cen-
and business areas, it also suggested hierarchies of (a) busi- tres are possibly of regional scale in terms of population
ness ribbons located largely on major highways, and (b) catchment, but have a much more limited range of goods
planned centres. Berry also designated separate hierarchies compared with regional shopping centres.
in ‘high-income’ and ‘low-income’ areas of the city. A second problem occurs, similar to that referred to
above in connection with hierarchies of urban retail areas.
4 At that time, the term ‘centre’ was commonly used for what this paper
Planned centres are typically evaluated by consumers on
terms ‘retail areas’. bases such as price, ‘atmosphere’, quality of goods sold, or
260
Table 5. A typology of unplanned retail areas in British towns

Retail area Catchment area Shop numbers and types

Central area Whole town, plus surrounding Over 100; mainly personal and
suburbs and rural areas fashion goods
District centre Inner urban or suburban area of 50–100; convenience, house-
20 000–50 000 population hold and personal goods
Neighbourhood Residential area of about 20–40; convenience and some
centre 10 000 population household goods
Local centre Immediately surrounding area 1–10; convenience goods
of 500–5000 population

Source: adapted from Guy (1994a) Table 6.1.

adequacy of car parking. Many research studies have linked not be roofed over. Its lower size limit is often taken to
particular centres to particular social groups in the catchment be 10 000 m2 of gross retail area. It includes one or more
area (e.g. McGoldrick and Thompson, 1992). large ‘anchor stores’ (department or variety stores), sev-
eral (often over 50) smaller retail units, and (in larger
Classification of shopping centres by physical form and more recent centres) a food court, and leisure uses
such as cinema or ice rink. Such centres attempt to
A different approach to classifying shopping centres lies replicate the amount and variety of shopping space in
in their physical characteristics such as appearance, tenant long-established central shopping areas. Small compar-
mix, and size. The advantage of this descriptive approach ison goods shops selling clothing, footwear, leisure and
is that it employs some of the criteria which shopping cen- luxury items are particularly important, and help dis-
tre developers use in describing and evaluating shopping tinguish the centre from retail parks and other types of
centre proposals. It also matches criteria which consumers centre.
often use in identifying shopping destinations and choosing (iv) The Regional Shopping Centre is a large shopping mall,
between them. built in a free-standing position rather than as part of
It may be useful first to provide simple descriptions of an existing central area. The lower size limit to this
the main forms of modern shopping centre development in type of centre is often stated as 500 000 ft2 (as in Guy,
north America and western Europe: 1994b), but limits as low as 30 000 m2 have been used
(i) The Focused Centre. These have typically been built (e.g. in Reynolds, 1993). Very large centres of over
to serve surrounding residential areas with convenience about 800 000 ft2 (known in north America as super-
shopping needs, and consist of a supermarket, hyper- regional centres) are usually built on two levels, in order
market or (in north America) discount department store to reduce the total land take of the centre and to shorten
plus some small shop units. They are often known as lo- the amount of walking which shoppers have to carry out
cal centres or neighbourhood centres in north America; within the centre itself. In such a centre, the major stores
as district centres in the UK and centres intercommu- will probably trade from both floors.
naux in France (Reynolds, 1993). (v) The Factory Outlet Centre is similar physically to a rel-
(ii) The Retail Park. These schemes are usually built in off- atively small shopping mall but has no anchor store.
centre locations. They consist of several large stores The outlets sell price discounted goods and are usually
selling a wide range of bulky ‘household goods’ such managed by manufacturing or wholesale organisations.
as furniture, washing machines, DIY materials, etc. The purpose is to sell surplus goods such as ends of
More recently, retail parks in the UK have begun to sell ranges, outdated fashions, slightly imperfect specimens,
clothing, shoes and other personal/fashion goods (Guy, etc. (Fernie, 1995; Fernie and Fernie, 1997).
1998b). (vi) The Speciality Centre may resemble a modern shop-
The retail park has been defined in physical terms as: ping mall or may be converted from one or more old
‘. . . at least 50 000 ft2 gross lettable retail area, and built buildings. There is no anchor store, and the retail out-
and let as a retail entity. It should be sited outside the lets (mainly owned by independents) tend to specialise
town centre and contain at least three retail warehouses, either in one type of goods or in goods designed for a
defined as single storey retail units of at least 10 000 ft2 . visitor/tourist market. The term festival marketplace has
It should also include some purpose built pedestrian area been used for some examples, mainly in north America,
or joint car parking facilities’ (Hillier Parker, 1994). which have particularly strong tourist appeal.
The stores within a retail park may be physically sepa- A classification which includes these types of centre is
rate, or joined together, or a mixture of the two styles. shown in Figure 1. The two main criteria used are physical
Some retail parks include leisure uses such as cinemas (one building or several), and the composition of the centre,
or indoor bowling arenas. in terms of the number of buildings or the number of oc-
(iii) The Shopping Mall is in effect contained in one very cupiers. This classification is closely related to developers’
large building, although some of the internal spaces may perceptions in the following ways.
261

Figure 1. Retail development: a typology based on physical characteristics. Source: adapted from Guy (1994, Table 2.4).

The physical composition of a centre reflects the cen- ilar leases as none of them is powerful enough to be able to
tre’s main type of appeal for shoppers. A centre which attract preferential terms.
consists of separate buildings, or a simple row of stores This suggests a classification which is shown in Figure 2.
connected by an open walkway, is relatively cheap to con-
struct but lacks visual appeal and character. This type of Classification by retail offer or trip purpose
centre is often intended to emphasise low prices and con-
venience for the shopper. Typical centres of this type are The simple classifications discussed so far do not provide a
neighbourhood and community centres, which are gen- full explanation. For example, retail parks and factory outlet
erally destinations for convenience shopping; and retail centres are similar in terms of size and ownership, but in
parks and factory outlet centres, which are destinations for fact offer different types of merchandise and serve different
household and comparison shopping. On the other hand, consumer markets.
shopping malls emphasise quality, luxury and atmosphere, This indicates that a classification of shopping centres
especially if climate-controlled. They are destinations for based upon types of shopping trip, as discussed earlier in this
personal/fashion shopping trips which are made for purposes paper, may be worth exploring. Thus, convenience shopping
of enjoyment rather than for purchasing essential items. trips are related largely to centres anchored by supermarkets
The epitome in this respect is the north American super- or other large food stores. Household shopping trips are re-
regional centre, or the tourist-oriented multi-level shopping lated most strongly to retail warehouse parks although they
‘complex’ or ‘emporium’ found in large cities in south-east may also involve use of more traditional types of urban sub-
Asia. centre. Personal/fashion shopping involves a greater amount
of window-shopping and comparison behaviour, hence is
Classification by centre ownership and tenancy typically associated with town centres or shopping malls. In
arrangements north America and to some extent the UK, personal shopping
may also take place in off-centre department stores and retail
Another important issue to the developer is the nature of the parks, where these sell personal goods (Guy, 1998b).
retail occupiers and the types of tenancy agreement with the Two further categories of shopping behaviour, not pre-
centre’s owners. A free-standing store, such as a hypermar- viously considered, need to be added at this point. The
ket, is often owned by the retailer concerned. At the other shopping centre industry is increasingly aware of the recre-
extreme, the super-regional centre is likely to be owned by ational aspects of consumer shopping trips. Similarly, the
a major national or international property company, and the tourist industry is increasingly using retail sales to boost its
individual shops are leased to retailers. The major anchors profitability. Specialist types of shopping centre which delib-
for the centre (stores which are important enough to gen- erately encourage ‘leisure shopping’ include factory outlet
erate large numbers of trips in their own right, and which centres and festival marketplaces.
thus generate trade for other retailers) are likely to benefit Secondly, a type of shopping behaviour characteristic of
from long leasehold ownership and will pay far lower rents commuters and travellers needs to be considered. This is the
per ft2 or m2 than the smaller units (Guy, 1994a, pp. 164– purchase of either convenience items needed at home, or
166). Such stores include department stores, or, in smaller items needed on a journey, such as newspapers and books.
centres, variety stores or hypermarkets. Between the owner- This ‘incidental’ shopping is associated with ancillary shops
ship arrangements of the free-standing hypermarket and the which range from small kiosks at railway stations to almost
super-regional centre lie various stages. The focused centre full-scale shopping malls built in some international airport
is often owned by the hypermarket operator concerned. The terminals.
retail park or factory outlet centre is typically owned by a
property company, and all the retail occupiers will hold sim-
262

Figure 2. Retail development: a typology based on ownership characteristics.

Table 6. Types of shopping centre related to location


Classifications of retail locations
Type of centre TC ETC ORA OU NRA ET
The final question which concerns retail geographers and
analysts is the location of a store or centre. The location Free-standing store X X X X
of a specific retail development implies a unique relation- Focused centre X X X
ship with the local residential population, and a competitive Retail park X X
Shopping mall X X X
stance relative to other retailing in the area. More generally,
Regional centre X X
however, it is useful to distinguish certain broad types of lo-
Speciality centre X X X
cation. This helps both to categorise retail areas or shopping
Factory outlet centre X X
centres, and to analyse locational decisions of developers
and retailers. Key: TC, town centre; ETC, edge of town centre; ORA, other retail area
(unplanned); OU, other urban (e.g. industrial, waterfront); NRA, new
residential area; ET, edge-of-town; X, signifies that this type of centre is
Intra-urban retail locations likely to be found in this type of location.
Source: Adapted from Guy (1994a), Table 2.9.
The locational context for retail development has particu-
lar significance for land use planning, where the supposed
problems of ‘out-of-town’ or ‘off-centre’ development are • Other urban: any urban area with some form of ex-
constantly debated (Guy, 1998a). In this context, the UK isting commercial development, usually industrial or
government has adopted the following classification of retail waterfront but in some cases retail.
locations within urban areas: • New residential area: usually edge-of-town.
• Edge-of-centre: ‘For shopping purposes, a location • Edge-of-town: a site originally rural in nature or bounded
within easy walking distance (ie. 200–300 m) of the by rural land.
primary shopping area . . . ’ In Table 6 a simple analysis is made of relationships
• Out-of-centre: ‘A location that is clearly separate from a between these locational types and certain types of mod-
town centre, but not necessarily outside the urban area’ ern retail development. The table shows which types have
• Out-of-town: ‘An out-of-centre development on a green- been built in which locations within the last 20 years or so
field site, or on land not clearly within the current ur- in western European countries. Current western European
ban boundary’. (Department of the Environment, 1996: legal restrictions and land use planning policies would now
Annex A.) preclude some of these options, such as regional shopping
For geographical as well as planning purposes, it is im- centres in edge-of-town locations (Guy, 1998a).
portant to be aware of the ways in which retail development
is related to the use made of land in its vicinity. New retailing ‘Non-hierarchical’ classifications of locations
may replace other land uses (such as industrial), and may
also serve an adjoining residential area. In this respect the An alternative approach to classifying retail locations lies in
following classification builds on the above to express the an analysis of their form and function. The starting point
various possibilities for location within typical urban areas: is Berry’s (1963) classification, already discussed, which
• Town centre: the historic central retail core of a town, combined planned shopping centres and unplanned retail
possibly including modern shopping malls. areas. This has however been criticised by Brown (1991,
• Edge of town centre: an area characterised mainly by p. 374) on two grounds: firstly, because of its reliance on
non-retail land uses but within easy walking distance of the now outdated notion of the central place hierarchy; and
the retail core. secondly, because of confusion between physical and func-
• Other retail area (unplanned): this refers to traditional tional characteristics, as for example in the ‘automobile
suburban retail areas and ribbon developments. row’, which is both shopping ribbon and specialised area,
although classified as the latter.
263
Table 7. Non-hierarchical classification of retail locations

Function

Form General Specialist Ancillary

Cluster (unplanned) Town centres and ‘Cafe quarter’ Sandwich bars in


suburban retail financial district
areas
Cluster (planned) Shopping malls Factory outlet centre Shops in airport
Retail parks Festival marketplace departure lounge

Linear Traditional ‘Ethnic’ shopping Shops in pedestrian


shopping ribbons street subway

Isolated Corner shop Retail warehouse Newspaper stall


Hypermarket

Source: Adapted from Brown (1991, Figure 2).

Table 8. Some dimensions used in retail classifications


From these arguments Brown (1991) develops a ‘non-
hierarchical’ classification of retail locations. Table 7 depicts Dimension Typical categories
this classification and includes many of the types of retail
area and shopping centre referred to in this paper. The two Goods sold Food store, ladies’ wear shop
dimensions of this classification are physical form (clus- Trip purpose Convenience, household, personal/fashion
tered, linear or isolated); and function (general, specialist Size of store Supermarket, hypermarket
Store ownership Independent, multiple, franchisee
or ancillary). This classification has some appeal in that the
Catchment area Neighbourhood centre, district centre
physical characteristics are easily identifiable from large-
Physical form Cluster, linear
scale maps or a geographical information system, and the
Development history Planned, unplanned
functional characteristics are identifiable from superficial
Development type Retail park, regional shopping centre
observation. However, it is less useful than that shown in Function General, specialist, ancillary
Table 6 in analysing development trends, as it does not Location Town centre, edge of town
fully correspond with the criteria used by the development
industry or by town planners.

Conclusions been accessible within urban areas. It has been necessary


thus to devise classification systems which incorporate (at
This paper has reviewed many of the methods used by retail least by proxy) some if not all of these variables.
geographers and other analysts in classifying retail outlets, The most successful way of doing this, which also re-
retail areas and shopping centres. The detailed discussion flects the decision criteria used by retail property developers,
has been largely of systems of classification used in the UK is to involve either a more sophisticated classification of
and north America, but the broad approaches are commonly shopping trips, or some type of simple description of the
used throughout developed countries. built form of the shopping destination concerned. These no-
It is clear that until recently classification relied largely tions underlie most of the classifications discussed in this
upon two rather simplistic notions about consumer shopping paper. These are summarised in Table 8.
behaviour. The first of these was the distinction between Although classifications have become better related to
convenience and comparison trips: this distinction was held the main determinants of retailer and consumer behaviour,
to be self-evident and clearly related to types of shops. it would incorrect to insist that one system of classification
Recent trends, including the growth of large mixed retail should be used in preference to another. There are two im-
businesses, and of recreational or ‘leisure’ shopping, have portant issues here. Firstly, a compromise always has to be
rendered the convenience/comparison distinction less valid found between comprehensiveness and simplicity. Too many
as a means of classifying retail goods and outlets. categories make analysis and interpretation difficult; too few
The second notion was that consumers would tend to buy can obscure essential insights. Secondly, the means of clas-
the goods required (convenience or comparison) at the near- sification should partly at least relate to the purpose of the
est available retail area or shopping centre. This assumption, research exercise. For example, a researcher analysing shop-
which underlies central place theory, is clearly incorrect for ping destination choices from the results of a travel survey
many actual shopping trips since other determinants of shop- should be aware that an ‘evoked set’ of outlets or shopping
ping choice are also important. The position since the 1960s centres from which shoppers make their choices should be
has been that many different shopping opportunities, distin- defined with respect to a particular combination of location
guished by environmental quality, parking facilities, price and trip purpose. This suggests use of trip categories (as in
structures and other criteria as well as size and location have Table 3 of this paper) to classify outlets or shopping centres.
264

For these reasons, no attempt is made in this paper to Davies R.L., 1977: Marketing Geography: With Special Reference to
recommend a ‘best’ system of classification. Table 8 sum- Retailing, Methuen, London.
Davies R.L., 1984: Retail and Commercial Planning, Croom Helm,
marises the main dimensions used in practice to construct Beckenham.
classifications. A categorisation appropriate to any partic- Dawson J.A., 1983: Shopping Centre Development, Longman, London.
ular study might be achieved through the use of two or Dawson J.A. & Lord J.D., 1985: Shopping Centre Development: Policies
three of these dimensions. Some such combinations have and Prospects, Croom Helm, London.
Dawson J.A. & Sparks L., 1986: Issues for the planning of retailing in
been discussed in this paper, but others are feasible. It is Scotland, Scottish Planning Law Practice 18: 38–40.
arguable also that systems may have to change in order to Department of the Environment, 1996: Planning Policy Guidance 6: Town
accommodate new forms of retailing as they emerge. Centres and Retail Developments, HMSO, London.
Most academic papers end with a call for further research Euromonitor, 1996: European Marketing Data and Statistics 1996, Eu-
romonitor, London.
to be made into some relevant topic. Mere classification of Fernie J., 1995: The coming of the fourth wave: new forms of retail out-of-
the spatial characteristics of retailing is not perhaps a topic town development, Int. J. Retail Distr. Manag. 23: 4–11.
for research in its own right, unless related to some more Fernie J. & Fernie S.I., 1997: The development of a US retail format in
substantive research issue. However, one approach which Europe: the case of factory outlet centres, Int. J. Retail Distr. Manag. 25:
342–350.
has been missing in this paper would be concerned with Guy C.M., 1980: Retail Location and Retail Planning in Britain, Gower,
consumers’ own images and perceptions of shopping oppor- Farnborough.
tunities: the classifications discussed have emanated from Guy C.M., 1984: The urban pattern of retailing: B. Within the UK’. In:
Davies R.L. & Rogers D.S. (eds), Store Location and Store Assessment
deterministic ‘rules’ of shopping behaviour, or from per- Research, Wiley, Chichester.
ceptions of property developers and town planners. There Guy C.M., 1994a: The Retail Development Process, Routledge, London.
is a need to consider the dimensions which shoppers them- Guy C.M., 1994b: Whatever happened to regional shopping centres?,
selves use to differentiate shopping opportunities. Are these Geography 79: 293–312.
Guy C.M., 1998a: Controlling new retail spaces – the impress of planning
related more to the shops contained within retail areas and policies in Western Europe. Urban Studies 35: 953–979
shopping centres, to physical descriptors of these areas and Guy C.M., 1998b: High street retailing in off-centre retail parks: a review
centres, or to the type of shopping trip involved and the of the effectiveness of land use planning policies. Town Planning Rev.
way in which this expresses other personal and household 69: 291–313
Hillier Parker, 1994: Retail Warehouse Park Development Master List 1994,
characteristics? Out of such research might emerge a more Hillier Parker, London.
simple and universal means of classifying retail stores and Jones K. & Simmons J., 1990: The Retail Environment, Routledge, London.
shopping centres. Kulke E., 1996: German fachmarkts – their effects on the structural and
locational change in German retailing, European Retail Digest, Spring
issue: 8–12.
Lord J.D. & Guy C.M. 1991. Comparative retail structure of British and
References American cities: Cardiff (UK) and Charlotte (USA). Int. Rev. Retail,
Distr. Consumer Res. 1: 391–436.
American Marketing Association, 1948: Report of the Definitions Commit- McGoldrick P.J. & Thompson M.G., 1992: Regional Shopping Centres:
tee, J. Marketing 13: 202–217. Out-of-town Versus In-town, Avebury, Aldershot.
Beaujeu-Garnier J. & Delobez A., 1979: Geography of Marketing, Long- O’Brien L.G. & Harris F., 1991: Retailing: Shopping, Society, Space, David
man, London. Fulton, London.
Beavon K.S.O., 1977: Central Place Theory: A Reinterpretation, Longman, Potter R.B., 1981: The multivariate functional structure of the urban
London. retailing system: A British case study, Trans. Inst. Brit. Geog. 6:
Berry B.J.L., 1963: Commercial Structure and Commercial Blight, Depart- 188–213.
ment of Geography, University of Chicago, Research Paper 85. Reynolds J., 1993: The proliferation of the planned shopping centre. In:
Berry B.J.L., 1967: Geography of Market Centres and Retail Distribution, Bromley R.D.F. & Thomas C.J. (eds), Retail Change: Contemporary
Englewood Cliffs, Prentice-Hall, New Jersey. Issues, UCL Press; London, pp. 70–87.
Bromley R.D.F. & Thomas C.J. (eds), 1993: Retail Change: Contemporary Reynolds J. & Schiller R., 1992: A new classification of shopping centres
Issues, UCL Press, London. in Great Britain using multiple branch numbers. J. Property Res. 9: 122–
Brown S., 1991: Retail location: the post hierarchical challenge, Int. Rev. 160.
Retail, Distr. Consumer Res. 1: 367–381. Schiller R. & Jarrett A., 1985: A ranking of shopping centres using multiple
Brown S., 1992: Retail Location: A Micro-Scale Perspective, Avebury, branch numbers. Land Devel. Studies 2: 53–100
Aldershot. Toderian B., 1996: Big-box retailing: how are municipalities reacting? Plan
Burns W., 1959: British Shopping Centres, Leonard Hill, London. Canada 36: 25–28.
Business Statistics Office, 1992: Retailing: Business Monitor SDA25, Unit for Retail Planning Information, 1994: UK Goods Based Retail Ex-
HMSO, London. penditure Estimates and Price Indices, URPI Information Brief 94/2,
Christaller W., 1966: Central Places in Southern Germany (trans. C.W. Reading.
Baskin), Prentice Hall, Englewood Cliffs. Wrigley N. & Lowe M. (eds.), 1996: Retailing, Consumption and Capital:
Davies R.L., 1974: Nucleated and ribbon components of the urban retail Towards the New Retail Geography, Longman, Harlow.
system in Britain, Town Planning Review 45: 91–111.

You might also like