The Trade Area Mix and Retailing Mix: A Retail Strategy · PDF fileBert Rosenbloom The Trade Area Mix and Retailing Mix: A Retail Strategy Matrix A useful analytical framework for

Bert Rosenbloom

The Trade Area Mix andRetailing Mix: A RetailStrategy Matrix

A useful analytical framework for selecting new store trade areas and foradjusting the merchandise of existing stores to their present trade areas.

THE selection of trade areas for new storesand the adaptation of existing stores to their

present trade areas present complex problems tothe ret£uler. A substantial body of literature hasemerged in which many concepts, methods, andtechniques are offered for dealing with some ofthese problems.' Much remains to be done, how-ever, particularly in developing frameworks fortrade area analysis that are understandable anduseful to retail management.

This article presents a three dimensional con-struct for describing a retail trade area. The con-struct, referred to as the trade area mix, is com-bined with the traditional retailing mix to form aRetail Strategy Matrix.

The Trade Area Mix:A Three Dimensional Construct

In Exhibit 1, the retail trade area is describedby three basic dimensions: (1) trade area geog-raphy (TAG); (2) trade area demand (TAD); (3)trade area heterogeneity (TAH). Each of these di-mensions is discussed below.

Trade Area Geography (TAG)

Trade area geography (TAG) refers to the geo-graphical extent of the trade area. This dimensionis usually emphasized in the definitions of the re-tail trade area found in the literature. Applebaum

• ABOUT THE AUTHORBert Rosenbloom is associate professor of marketingat Drexel University, Philadelphia.

and Cohen state, for example, that the retail tradearea is "the area from which a store gets its busi-ness within a given span of time."^ Huff defines itas:

A geographically delineated region, containingpotential customers for whom there exists aprobability greater than zero of their purchasinga given class of products or service offered forsale by a particular firm or by a particularagglomeration of firms.^

A number of ingenious methods and techniquesranging through mathematical models,*• observa-tions and surveys,^ mapping techniques,* drivingtime studies/ and other methods^ have been de-vised by researchers seeking to delineate the geo-graphical boundaries of retail trade areas.

The heavy emphasis on the geographical di-mension is not surprising because trade areageography is the seminal or prerequisite dimen-sion for describing a trade area. Managementmust know the geographical extent of its marketfor any given store before it can proceed with anyfurther appraisal of the store's market.

Trade Area Demand (TAD)

The second dimension, trade area demand(TAD) refers to the level of consumer demandwithin the geographically delineated trade area.TAD, in other words, describes the existing or po-tential purchasing power in a trade area for astore's offerings. The TAD dimension is examinedwhenever the retailer attempts to forecast salesfor a new store or existing one. A variety of tech-niques have been developed for forecasting retailsales particularly for new store trade areas.'

58

A Retail Strategy Matrix 59

Often the TAG and TAD dimensions are closelyand positively related as has been the case for themany suburban trade areas emerging since WorldWar II. Yet each of these two dimensions is notnecessarily a positive function of the other. Somegeographically large trade areas have a relativelylow level of demand while some geographicallycompact areas are relatively high in demand.

Further, for any given trade area, changes oc-curring over time in the two dimensions may notbe closely and positively correlated. This canhappen when infrastructual changes such as im-proved highways which may increase the extentof a trade area's boundaries are also accompaniedby an out-migration of population from thearea.'" Finally, it is possible for a trade area toexperience an increasing TAD even as its TAG be-comes smaller. if an in-migration of a moreaffluent population occurs. The central businessdistricts in a number of major U.S. cities repre-sent trade areas where this phenomenon may oc-cur." Thus explicitly distinguishing between the

TAG and TAD dimensions helps to foster a clearanalysis of trade areas.

Trade Area Heterogeneity (TAH)

Trade area heterogeneity (TAH), the third di-mension shown in Exhibit 1 refers to the mix ofconsumer market segments within the trade area.The diversity of consumer demand for productsand services is described by this dimension.

The greater the diversity of demand, the higheris the degree of trade area heterogeneity. Hence,the more heterogeneous trade area is charac-terized by more market segments demandinggreater varieties of goods and services while theless heterogeneous (or more homogeneous) hasfewer market segments.'^ TAH is not necessarilyrelated positively to TAG. Trade areas large ingeographical size may be low in heterogeneityand vice versa. Further, although one would ex-pect the TAD and TAH dimensions to be relatedpositively, significant variations in heterogeneitycan exist among trade areas similar in levels ofdemand.'^

EXHIBIT 1. T H E TRADE AREA Mix CONSTRUCT

Trade AreaGeography

(TAG)

TradeAreaDemand

(TAD)

Trade AreaHeterogeneity

(TAH)

60 Journal of Marketing, October 1976

EXHIBIT 2SUB-MIXES AND VARIABLES OF THE RETAILING MIX

Goods and ServicesVariabies

CommunicationVariables

Physical DistributionVariables

Variety and AssortmentParkingSales ServiceCustomer ServiceCreditPrice LinesGuarantees and ExchangesAlterations and AdjustmentsDelivery

Personal SellingAdvertisingWindow DisplayInterior DisplayPublic RelationsStore LayoutCatalogsTelephone Sales

Store LocationDistribution CentersInventory ControlTransportationHandling Goods

Source: William Lazer and Eugene J. Kelley, "The Retailing Mix: Planning and Manage-ment ,'Vowrna/o/"/?e/ai7mg, Vol.37 (Spring 1961).

The Retailing Mix

The retailing mix developed by Lazer and Kel-ley is a conceptual model portraying the retailmanagement process.''* It consists of three relatedsub-mixes;

1. A goods and services mix2. A communication mix3. A physical distribution mix

Exhibit 2 lists some of the variables of each ofthe sub-mixes. The fundamental role of retailmanagement is one of blending the variables ofthe three sub-mixes so as to adjust the total offer-ing of the retail store to its market environment.Consumer satisfaction and profitable store opera-tions are achieved through optimal sub-mixblending.'^

The Trade Area Mix and the Retailing Mix:A Retail Strategy Matrix

Operationally, the market environment for a re-tail store is its trade area. Hence, the trade areamix construct provides a means for describing themarket environment (trade area) for any given re-tail store in terms of the spatial location of itsconsumers (TAG), consumer purchasing power(TAD), and the diversity of its consumer demand(TAH). By combining the trade area mix with theretailing mix in a matrix, the retailer is providedwith a framework for analyzing the trade areadimensions in relation to the retailing mix.Exhibit 3 shows the matrix. Each of the ninenumbered cells in the matrix helps to focus theretailer's attention on the possible strategic im-plications of the trade area dimensions on thevariables in a store's retailing mix.

Using the Retail Strategy Matrix

Goldstucker argues that too many retailers dis-play little strategic flexibility in their location de-cisions:

Retailers currently act within a frameworkwhich seems to place unnecessary constraintsupon their flexibility in selecting and adjustingto locations. One of these constraints is that aretailer usually views his type of operation asgiven and then seeks only sites which appear tobe compatible with it."*

Consequently many retailers fail to develop thestrategic combination of retailing mix variablesappropriate for particular trade areas beforelocating in them and/or lack a systematic proce-dure for adjusting to the areas over time. Thisleads to sub-par performance or ultimate storefailure.

The Retail Strategy Matrix offers the retailerassistance in dealing with this problem. Byanalyzing any given trade area using the RetailStrategy Matrix, the retailer gets a more coherentoverview of the three trade area dimensions asthey relate to the variables in a store's retailing mix.This helps the retailer to perceive more closelyhow his retailing mix for particular store unitscan be adjusted to fit more precisely the dimen-sions of various trade areas.

The next section presents examples of how par-ticular cells and combinations of cells within thematrix can be used by retailers to help them de-velop more flexible strategies for adjusting totrade areas.

Strategic Flexibility ThroughMatrix Cell Analysis

Ceil 1—TAG and the goods and services mixconverge in this cell. The retailer's analysis here

A Retail Strategy Matrix 61

•'""-•s,, ^ Trade Area^ ^ ^ ^ Mix

Retailing """'-v..,, ^Mix ""- -s,,, ^

Goods and ServicesVariables

CommunicationVariables

Physical DistributionVariables

Trade AreaGeography

(TAG)

1

4

7

Trade AreaDemand(TAD)

2

5

8

Trade AreaHeterogeneity

(TAH)

3

6

9

EXHIBIT 3. RETAIL STRATEGY MATRIX

should focus on possible relationships betweenone or more of these variables (see Exhibit 2) anda store's TAG. For example, one relationship sug-gested is that between variety of merchandise of-fered and the distance of consumer travel. It hasbeen argued that an increase in the vai'iety ofmerchandise variable can increase the distanceconsumers are willing to travel to gain access tothe store offering i t . " Exhibit 4 illustrates thisrelationship.

In Exhibit 4, the level of merchandise variety at

V| is associated with a willingness of an averageconsumer to travel the distance D,. By increasingthe variety of merchandise offered to Vo, the dis-tance consumers are willing to travel increases toDa, which extends the TAG. Eventually a point ofmaximum distance is reached (E^ in Exhibit 4)beyond which increases in variety will not attractmore distant consumers. The actual levels of mer-:chandise variety and distances involved beforethis point is reached will vary widely for differentkinds of retailers in different trade areas.'*

DistanceofConsumerTravel

D,

Variety olMerchandise

EXHIBIT 4. RELATIONSHIP BETWEEN VARIETYAND DISTANCE OF CONSUMER TRAVEL

62 Journal of Marketing, October 1976

Barney's, one of the world's largest men's storesis a good example of a retailer whose strategyemphasizes this relationship between variety anddistance. Barney's one store, located at SeventhAvenue and 17th Street in New York City, hastraditionally emphasized a vast variety of bettermen's suits and other apparel as an alternative tobranch stores in prime locations. The customerCcin find virtually any style of clothing from con-servative to ultra high fashion in sizes rangingfrom 34 extra short to 54 portly. Barney's hasbeen correct in its belief that many suburban con-sumers would be willing to overlook the conveni-ence of their suburban shopping centers to gainthe advantage of a greater variety of men's cloth-ing. Consequently, the store has enjoyed a sub-stantial patronage (sales volume is in excess of$33 million) from a TAG that encompasses mostof the New York Metropolitan Area.'^

Different types of retailers may increase a givenstore's TAG by emphasizing other "goods andservices mix" variables. For example, many smallindependent jewelry stores located in the so-called "Jeweler's Row" in downtown Philadelphiaattract a large portion of their consumers fromthe distant suburbs by emphasizing price.

An independent audio store located in a suburbof the same city increased its trade area by em-phasizing the "personal selling" and "customerservice" variables. Some consumers were willingto travel extra miles to gain the advantages of theexpert advice available from this store's sales per-sonnel and the complete on-premise repair serviceoffered. 2°

These examples suggest that retailers can reapbenefits in the form of an expanded trade area byanalyzing the effects that a store's goods and ser-vices variables have on a store's TAG.

Cells 2 and 3—These two cells are concernedwith the "fit" of a store's goods and services mixvariables with the level and diversity of consumerdemand in any given trade area. In general, thetwo cells should be evaluated simultaneously be-cause they are closely related. Changes in thelevel of demand (TAD) will often be accompaniedby changing market segments (TAH), and viceversa.

As an example of how a retailer can use thesecells to sharpen his analysis, and subsequentstrategy, consider the Abraham and Straus de-partment store located in downtown Brooklyn,New York. This huge store (1.4 million squarefeet) has been highly successful for decades and astandout performer in its parent organization.

Federated Department Stores. During the lastseveral years, however, the Brooklyn store has ex-perienced a decline in sales growth and seriouserosion in its profit margins. Some trade sourcesbelieved this was an inevitable consequence of thegeneral decline i n demand in the downtownBrooklyn trade area.-'

The new top management which took over inearly 1973 did not support this view. After con-ducting a series of market surveys of the Brooklyntrade area, management found that the area wasstill characterized by a high level of consumerdemand. But it also found that the market seg-ments making up the demand had changeddramatically over the last decade. The mix of in-come, age, occupations, and ethnic groups wasdifferent. Moreover, attitudes and lifestyles hadchanged. Management believed that thesechanges combined to foster a larger group ofmore fashion-conscious customer segments andfewer customer segments demanding a utilitarianprice emphasis. In-store surveys of A.<£S. custom-ers, however, revealed that only a small numberof the more fashion conscious customers were ac-tually shopping at A.&S.^^

In terms of the TAD and TAH dimensions,A.tSS.'s management thus believed that the store'sunfavorable performance was not a function of adeclining TAD, but an indication that the store'sexisting goods and services mix was not ade-quately meeting the demands of the changingmarket segments (TAH).

The new top management team is now at-tempting to adjust, or as they put it, "fine tune,"the store's goods and services mix to better meetthe new TAH dimension. In hne with its attemptto attract more of the fashion conscious marketsegments into the store, a heavier emphasis hasbeen placed on fashionable women's accessories,cosmetics, and men's wear. At the same time suchstaples as notions, drugs, stationery, food, andlower-priced lingerie, have been de-emphasized.The results so far appear to be very promising.^^

In summary, A. di S. management was carefulto distinguish between the TAD and TAH dimen-sions in its analysis of the Brooklyn store's tradearea. If management had interpreted the unfavor-able sales and profit picture to be a reflection of ageneral decline in TAD, it might very well haveopted to drastically reduce the size of the store'soperation or perhaps, even to close it down. In-stead management discerned a changing TAHdimension and chose to adjust to it. In so doingA.&S.'s Brooklyn store has remained viable andprofitable.

A Retail Strategy Matrix 63

Thus retailers who are faced with changingtrade areas should be careful to distinguish be-tween a real decline in the purchasing power ofthe trade area (TAD) and a change in the mix ofcustomer segments (TAH). If the latter is the case,the store might well be able to remain profitableby adjusting to meet the demands of the changingmarket segments (TAH).

Cell 4 — The adjustment of one of these vari-ables—advertising—to a store's TAG has tra-ditionally been a problem for retailers. Becausethe pattems of media coverage for newspapers,radio, and T.V. often extend well beyond a givenstore's trade area boundaries, substantial "wastecireulation" is common in retail advertising.^*While the problem is most pronounced for thesmaller single-unit retailer who typically has avery limited TAG, this media fit problem is alsoshared by large multi-unit retailers. For example,the Supemiarkets General Corporation, a large EastCoast supermarket chain, freely admits that thereis substantial "waste circulation" in some of itsstore trade areas through media advertising, andthe same holds true for other supermarketchains.^' A large department store chain and drugchain also report the same problem. *" Unfortu-nately, there is little that these retailers can doabout this problem because the coverage pattemsof available newspaper, radio, and T.V. media arenot geographically flexible.

Something can be done about the media fitproblem, however, when selecting trade areas forproposed stores. By giving more weight to mediacoverage patterns relative to a prospective store'sTAG, those trade areas that rank especially lowon media fit can be rejected. The above men-tioned retail chains have, in fact, recently put in-creased weight on the adequacy of media fit whenevaluating trade areas for new store units. While-\\\ of these retailers report that they would notreject an otherwise acceptable trade area solelyon the basis of poor media fit, it has played a partin their rejection of marginal trade areas. Buteven more importantly, this increased emphasison media fit has forced retailers to give more at-tention to adjusting advertising media usage toreduce "waste circulation." One result has beenan increased emphasis on the use of more geo-graphically selective media such as direct mailand circulars. For example, a Supermarkets Gen-eral Corporation store located in Patterson, NewJersey used direct mail and handout circulars toefficiently zero in on the large Spanish populationin the store's trade area. Not only were these di-

rect mail and handouts specifically geared to thetype of products demanded by this Spanish mar-ket segment, but they are also printed inSpanish.^'

Cells 5 and 6—These two cells deal with theappropriateness of a store's communication mixfor the TAD and TAH dimensions of its trade area.Given the systems nature of the three sub-mixesof the retailing mix, communication strategy de-veloped by the retailer must be related to andsupportive of the goods and services strategy.^'Thus, the strategy for Cells 5 and 6 should parallelthe strategy for Cells 2 and 3.

In the case of the A. & S. department store dis-cussed above, several parallel changes were madein the communication mix variables to comple-ment the goods and services changes. In particu-lar, newspaper ads were changed to reflect thegreater emphasis on fashion. The window dis-play, interior display and store layout variableswere also altered to reduce the previous clutteredlook of the store, and to create what A. & S. man-agement refers to as "a new fashion ambiance" tobetter complement the merchandise.^'

These complementary changes in the com-munication mix have resulted in a more com-prehensive adjustment to the trade area.

Cell 7—In this cell the key decision variable formany retailers is store location relative to thegeography of a given trade area. As mentionedearher, much existing trade literature emphasizesthis issue.^° The assumption underlying virtuallyall of this work is that consumers engage inbalancing the desirability of near and distant re-tailers against the cost, time, and energy whichmust be spent in overcoming travel distance fac-tors. If these factors are too high for patronizing adistant store, the customer will seek out one thatis closer." Thus finding a highly convenient loca-tion in terms of actual travel distance, driving orwalking time, or in some cases, proximity to masstransit points, has been a sine qua non of a goodretail location. This is particularly the case forstores selling convenience goods where intensecompetition is prevalent and a differential advan-tage based on goods and services and communi-cation variables is difficult to achieve. Drug-storeretailers, for example, argue that locational con-venience is more important than ever, especiallyin light of the effect of the energy crunch on thedistance which consumers are willing to travel.'^Recent research by the Supermarkets General Cor-poration shows that locational convenience con-

64 Journal of Marketing, October 1976

tinues to be a factor of overwhelming importancein supermarket patronage.'^

The range of trade area choices available to re-tailers handling specialty goods, however, isgreatly reduced by this emphasis on excellent lo-cation. The perspective offered by the RetailStrategy Matrix can increase the range of tradearea choices available to shopping and specialtygoods retailers. By looking at Gill 7 in conjunc-tion with other cells in the matrix, the retailer candiscern strategic trade-offs between location andother variables for attracting consumer patronagewhich may yield a higher return on investmentthan a "100% location." For example, considerthe trade-off previously discussed for Barney'smen's clothing store. This retailer opted for a lessdesirable location, but was able to more thanoffset this apparent disadvantage with a largevariety and assortment of merchandise to attractpatronage. The jewelry retailers and the audiostore previously discussed made trade-offs of priceand customer service in lieu of prime locations. Interms of the matrix, all of these trade-offs werebetween Cell 7 and Cell 1.

Another trade-off is possible between Cell 7and Cells 3 and 6 (TAH). For example, Pier I Im-ports, a large national retail chain specializing inimported furniture and gift items, generallyavoids prime mall and shopping center locations.Instead the firm selects somewhat "out of the waylocations" often on main highways near prime lo-cations. Most of these stores are free standing, andemphasize a large display of an unusual assort-ment of merchandise aimed primarily at youngermarket segments (see Cells 3 and 6). The savingsin rented cost from using these less prime loca-tions (often as much as one third to one half lessthan the cost per square foot of prime shoppingcenter or mall locations) more than offsets thehigher advertising expense needed to supportthese locations.'*

Thus, for specialty and shopping goods retail-ers, the matrix can suggest strategic trade-offswhich can increase the range of location choicesavailable to them.

Cells 8 and 9—Strategy for these two cellsshould parallel that developed for Cells 2 and 3,and Cells 5 and 6. A retailer's adjustment of thephysical distribution variables relative eo TADand TAH should, therefore, be congruent with thegoods and services and communication strategyadjustments. In the above discussion of the A. &S. department store, several adjustments werealso made in physical distribution mix variables

to complement the goods and services and com-munication changes. Specifically, a $5 millionremodeling of the store building is underway tofacilitate better display, inventory control, andmerchandise handling. Traffic flow pattemswithin the store are being studied to promote.more customer exposure to the new merchandiseassortments and to enable customers to movemore easily through the store. For example, sixdepartments have been removed from the firstfloor leaving only men's wear, women's acces-sories, and cosmetics. The increased space enablesA. & S. to better display the more fashionablemerchandise in these departments and facilitatescustomer access to it.^''

Large multi-unit retailers attempting to adjusttheir store product mixes to more precisely meetthe demands of different trade areas may alsohave to alter some of their existing procedures inthe use of distribution centers, inventory control,transportation, and handling. For example.Supermarkets Genera! Corporation found it cum-bersome and inefficient to supply specializedethnic foods from its main distribution center inWoodbridge, New Jersey to a number of its stores.Instead, it has often used local vendors locatednear the particular store trade areas to supply thestores wtih these products. Supennarkets Generaladmits that this has created some problems, par-ticularly in inventory control. Its management be-lieves, however, that the extra effort needed tosolve them is worthwhile because it enables thestores to have product mixes that are more intune with their particular trade areas.'*

Sorting and Weighting Trade Area Data

Trade aica data of all sorts are continuallyflowing into the retail firm. Besides the retailer'sown research efforts, unsolicited information re-ported by local media, feedback from consumersand vendors, reports by local government plan-ning agencies and research by universities locatedin or near the trade area are additional sourcesfor the retailer.

Two related problems faced by the retailer are:

• How to sort this mass of trade area data intomeaningful and easily accessible categories.

• How to determine which data are most im-portant.

For the first problem, the Retail Strategy Ma-trix is highly useful because it offers a conceptualpigeon-holing device for sorting trade area data.By setting up files labeled according lo each of thematrix cells, ehe retailer has a set of nine catego-

A Retail Strategy Matrix 65

ries into which he can place trade area informa-tion for later analysis and use. Hence, when heuses the matrix to evaluate a trade area to deter-mine if the location is suitable or if the existingretailing mix needs strategic adjustments, thedata stored in the files will exist in a form that isreadily accessible.

The Retail Strategy Matrix is also very usefulfor dealing with the second problem of determin-ing the relative importance or weight that shouldbe given to various trade area data. The perspec-tive offered by the matrix helps the retailer tofocus more clearly on the strategy questions thatare relevant to his particular situation. This, inturn, helps him to identify what trade area dataare most important to his specific needs. Forexample, a retailer (such as A. di S. departmentstores), focusing on whether changes should bemade in certain goods and service variables toadjust to new market segments (Cell 3), shouldgive great weight to data that would be useful inidentifying these market segments, especiallychanges in the demographic and psychographicprofiles of the trade area's population. A retailer(such as Barney's men's clothing store), consider-ing the likehood of attracting customers fromgreater distances by adjusting ehe variety of goodsvariable (Cell 1), would place a high level ofwei^t on data relevant to consumer shoppingpatterns, such as actual driving time, and con-sumer attitudes toward increasing their drivingtimes. A retailer (such as Supennarkets General

Corporation), focusing on the question of adjustingadvertising coverage to the TAG (Cell 4), wouldattach special importance to data on media avail-ability and its coverage.

Thus, by helping the retailer to focus moreexplicitly on the strategy issues relevant to him,the Retail Strategy Matrix helps to provide abasis for evaluating the importance of trade areadata.

Summary and Conclusion

A three dimensional construct called the TradeArea Mix was presented. When combined withehe traditional Retailing Mix in a matrix format,an analytical framework, referred to as the RetailStrategy Matrix, was developed.

Ihis Retail Strategy Matrix offers the retailer auseful framework for analyzing trade areas. Byperiodic and regular evaluation of the cells in thematrix, as they apply to trade areas for both exist-ing and proposed stores, the retailer can track andforecast changing trade area dimensions in amore systematic fashion. This provides a morecoherent framework for planning retailing mixstrategies in response to changes.

Given the increasingly complex environmentwithin which the retailer must operate, analyticalframeworks such as the Retail Strategy Matrixcan help to bring order to the complexity of themarketplace, a task sorely needed by retailers inevaluating their trade area data.

End notes

1. See for example: Jac L. Goldstucker. "Trading Areas"in Science in Marketing, George Schwartz, ed. (New York:John Wiley and Sons, 1965). pp. 281-300; Donald L,Thompson, "Future Directions in Retail Area Research,"Economic Geography, Vol. 42 (January 1966). pp. 1-18; LouisP. Bucklin, "Trade Area Boundaries: Some Issues in "Theoryand Methodology," Journal of Marketing Research, Vol. 8(February 1971). pp. 30-37; William Applebaum, "MethodsFor Determining Store Trade Areas, Market Penetration andPotential Sales," Journal of Marketing Research, Vol. 3 (May1966), pp. 127-141; Peter L. Simons. "The Shape of Subur-ban Retail Market Areas: Implications From a LiteratureReview," Joumal of Retailing, Vol. 49, (Winter 1973-74),pp. 65-78.

2- William Applebaum and Saul B. Cohen, "Store Trad-ing Areas in a Changing Mar\i£t," Journal at Retailing, Vol. 37Area." Journal of Marketing. Vol. 28 (July 1964), p. 38.

3. David L. Huff, "Defining and Estimating a TradingArea.' Joumal of Marketing, Vol 28 (July 1964), p. 38.

4. William J. Reilly, Methods for the Study of Retail Rela-tionships, Bureau of Business Research, Research Mono-graph No. 4 (Austin: The University of Texas, 1929); Paul D,Converse, "New Laws of Retail Gravitation," Journal ofMarketing. Vol. 14 (October 1949), pp. 379-84, See also:Huff, same as reference 3.

5. For a review of some of these methods see: Ronald R.

Gist, Retailing: Concepts and Decisions. (New York: JohnWiley and Sons, 1968). pp. 161-164.

6. Edward M. Mazze, "Determining Shopper MovementPatterns by Cognitive Maps," Joumal of Retailing, Vol. 50(Fall 1974), pp. 43-48.

7. James A. Brunner and John L. Mason, "The Influenceof Driving Time on Shopping Center Performance," ^oNma/of Marketing. Vol. 32 (April 1968), pp. 57-61.

8. See Goldstucker, same as reference I, pp. 292-299.9. See for example: Richard L. Nelson, The Selection of

Retail Locations, (New York: F. W. Dodge Corporation, 1958),pp. 140-164. For two examples of more sophisticated ap-proaches see: LA. White and J. B. Ellis, "A Systems Con-struct for Evaluating Retail Market Locations," yonma/o/^Marketing Research, Vol. 8 (February 1971), pp. 43-46, andDavid B. Mackay, "A Microanalytic Approach to Store Loca-tion Analysis," Joumal of Marketing Research. Vol. 9 (May1972), pp. 134-140.

to. For a discussion of some of the factors that can affectboth the geography and level of demand in trade areas see;Donald L. Thompson, "Consumer Convenience and RetailArea Structure," Joumal of Marketing Research, Vol. 4 (Feb-ruary 1967), pp. 37-44.

11. See for example: E. B. Weiss, "New Store Locations inthe Core City." Stores, (May 1972), pp- 43-44,

12. Wroe Alderson, Dynamic Marketing Behavior,

66 Journal of Marketing, October 1976

(Homewood. PL Richard D. Irwin, Inc., 1965). pp. 23-40.13. Joseph B. Mason and Charles T. More, "An Empirical

Reappraisal of Behavioristic Assumptions in Trading AreaStudies," Joumal of Retailing. Vol. 46 (Winter 1970-71), pp.36-37.

14. William Lazer and Eugene J. Kelley, "The RetailingMix: Planning and Management," Joumal of Retailing, Vol.37 (Spring 1961), pp. 34-41.

15. Ibid., p. 34.16. Jac L. Goldstucker, "A Systems Framework for Retail

Locations," in Science, Technology and Marketing, RaymondA. Haas, ed. (Chicago: American Marketing Association,1966). p. 421.

17. Martin Zober, Marketing Management (New York:John Wiley and Sons, 1964), pp. 131-132. See also: David J.Rachman, Retail Strategy and Structure (Englewood Cliffs,New Jersey: Prentice-Hall. 1969), pp. 176-78.

18. See for example: Robert O. Hermann and Leland L.Beik, "Shoppers' Movements Outside Their Local RetailArea." Joumal of Marketing, Vol. 32 (October 1968), p. 51.

19. From an interview with Milton Guttenplan, Directorof Advertising and Marketing, Barney s Clothes Inc., June 4,1976- For additional background on Barney's see: "A $33-Million Men's Shop Tries Social Climbing," Business Week.(September 1, 1973). pp. 48-49.

20. Bert Rosenbloom, "Find the Best Niche,"Audio Video,(October 1975), pp. 39-41.

21. From a speech given by Alan Gilman, President,Abraham and Straus Department Stores, at the 9thAbraham and Straus Management Institute Conference,"Managing to Meet Change," Brooklyn, New York, Octoher13, 1975.

22. From an interview with Harvey M. Gutman, ResearchDirector, Abraham and Straus Department Stores, January15, 1976.

23. For a further discussion of some of these changes see:Isador Barmash, "Remodeled A. & S. Tells What's in Store:A New Stress on Fashion," The New York Times, October 18,1975, p,45.

24. Lawrence W. Jacobs, Advertising and Promotion forRetailing: Text and Cases, (Glenview, Illinois: Scott, Fores-man and Company. 1972), pp. 116-118,

25. From an interview with Charles I. Bernhaut. Directorof Market Research, Supermarkets General Corporation,December 24, 1975.

26. From an interview with David J. Rachman, Professorof Marketing, Baruch College. City University of New York,January 19, 1976.

27. From an interview with Arnold Cohen. AdvertisingDirector, Supermarkets General Corporation. June 8, 1976.

28. Lazer and Kelley, same as reference 14.29. Barmash. same as reference 23.30. For example, the material cited in reference 1.31. For additional discussion of this topic see: Douglas J,

Datrymple and Donald L. Thompson, Retailing: An Eco-nomic View. (New York: The Free Press, 1969), pp. 98-104,

32. Ronald Gargano, "The Changing Drug Store: Escap-ing from Giganticism," Chain Store Age Executive, (Sep-tember 1975), p, 39.

33. Bemhaut interview, same as reference 25.34. From an interview with Donn Fletcher, Vice President

and General Manager, Pier I Imports, June 14, 1976.35. Barmash. same as reference 23,36. Cohen interview, same as reference 27.

MARKETING MEMO

Bootstrapping: Better than Judgment . . .

Bootstrapping involves the substitution of a simple linear model of judgmentsin place of the judgments themselves. It has been found that in many decisionmaking contexts the bootstrapped decisions are better than the judgments fromwhich they were derived. It appears that the linear model is quite successful atcapturing the policy of the judge and then making decisions without humaninconsistency. Most of the work done on bootstrapping has been done in acontext—such as forecasts—where the criterion or accuracy is clearly defined.

Bootstrapping works because the linear model is able to make extremely goodapproximations of most decision processes. The model then makes these judg-ments without random error. Thus by bootstrapping one replaces the randomerror of the judge with the nonrandom error of the model.

—Joel Huber, "Bootstrapping of Dataand Decisions," y Jwrrtfl/ of ConsumerResearch. Vol. 2 (December 1975),pp. 229-234. at p. 229. Published bythe Journal of Consumer Research,Inc.