0019-8501/01/$see front matterPII S0019-8501(99)00103-0
Industrial Marketing Management
, 473486 (2001) 2001 Elsevier Science Inc. All rights reserved.655 Avenue of the Americas, New York, NY 10010
Business to Business Market Segmentation
Per Vagn FreytagAnn Hjbjerg Clarke
This article discusses the characteristics of industrial mar-kets in relation to some of the major industrial market segmen-tation models. To understand the different market situations,we describe a scale with simple market transactions at one endand complex relationship management at the other, suggestingthat the segmentation approach must be different for each endof the spectrum. The article presents a general industrial seg-mentation model directed towards situations characterized byrelationships and networks. The model stresses the importanceof having a deep understanding of the customers characteris-tics, needs, future directions, as well as identification of whatkind of overall relationship is required by the customer. Thismodel involves identifying, selecting, and monitoring ofsegments. 2001 Elsevier Science Inc. All rights reserved.
Segmentation is a crucial activity in marketing, but theway it is viewed has changed over time. Some authors
view segmentation as being closely related to another ofmarketings major thoughts, the marketing concept .The essence of the marketing concept is that the best wayto address the customer is by satisfying their needs andwants. These needs and wants thus need to be fully under-stood, and several ways exist to collect and analyze thenecessary information. Which ways are used depends onthe guiding methodology and techniques applied, but sta-tistical analysis approaches are the most common.
Other authors do not see segmentation as a statisticalanalysis technique, but as a tool for resource allocation.They regard segmentation as an overall way of identify-ing different target groups (for the purpose of makingsome general strategic decisions, such as which busi-nesses the company should be in and how resourcesshould be allocated) . A major question is thereforewhether or not both points should be addressed at thesame time. In principle, segmentation is about identifyingand targeting customer groups through their needs andwants, as well as determining which customers and needswill be addressed and with what manner and intensity .
The industrial market often is characterized by cooper-ation between customer and supplier; with the supplier
Address correspondence to Dr. A. H. Clarke, University of SouthernDenmark, Department of Marketing, Grundtvigs Alle 150, 6400 Sanderborg,Denmark.
thus having in-depth knowledge of the customers needsand wants. When it is known beforehand that customersare interested in such collaborative relationships, this de-sire can be used as a segmentation base. In connectionwith Wind and Cardozos micro and macro segmentationmodel , they talk about the influence different combi-nations of decision units (buying center) can have on themarket segmentation. They, however, do not directlymention the relevance of the buyers attitude to collabo-ration as a segmentation base.
Another approach is represented by Bonoma and Sha-piro  in their nested approach. They do not explicitlytalk about using the customers attitude to collaborativerelationships as segmentation base either. However,within the framework of the nested approach, it is possi-ble to use questions about the customers interest in col-laborative work, because either the purchasing strategyor the buying center can be used as a segmentation base.
This article develops an analytical framework that ex-plicitly presents the opportunity to use the customers in-tentions towards collaboration with the supplier as a seg-mentation base.
INDUSTRIAL MARKETING THEORIES
Segmentations main uses are therefore of both opera-tional and strategic nature. It should result in guidelines forthe operational level that have the purpose of gaining acontinual competitive advantage for the company. Indus-trial marketing literatures various approaches all appear to
be pointing in roughly the same direction. Jackson  dis-tinguishes between lost-for-good and always-a-sharesituations, with the difference being how closely the sup-plier and customer cooperate. Lost-for-good situations arecharacterized by strong bonds between the supplier andcustomer that make switching suppliers expensive for thecustomer. Always-a-share situations are mainly character-ized by weak bonds and suppliers competing on price.
Jacksons  approach is closely connected to the stra-tegic management approaches called lean managementor lean thinking . In lean management, customershave close relationships with some of their suppliers,based on things such as mutual product or process devel-opment, or using only one or a limited amount of suppli-ers in certain areas. The opposite is arms length man-agement, which keeps more distance to the suppliers whoas a result mainly compete on price.
In the Swedish network theory [8, 9], the industrialmarkets are characterized by cooperation and mutual ad-aptation between the supplier and customer. This theorymakes a distinction between the overall situation and sin-gle episodes, making it possible to see different imagesof the market, depending on the companies involved, theindustry, the contents of the exchange process, etc.
These points of view see segmentation as more than justa technique for analyzing the environments and for allocat-ing marketing resources. They engage segmentation in theinternal workings of the company, and the company, as apart of the market, shares in shaping the environment.
A scale showing the facets of the relationship betweensupplier and customer can illustrate these three ap-proaches. One end of the scale is characterized with no co-operation, no bonds, no switching cost, and an nonadaptedproduct. Close cooperation, strong bonds, high switchingcost, and a highly adapted product mark the other end. Thedifferent factors often will be interrelated. The two ex-tremes are labeled simple market transaction and com-plex relationship management. At a certain point on the
The two extremes are labeled simple market transaction and complex
PER VAGN FREYTAG is Associate Professor of Marketing at the University of Southern Denmark.
ANN HJBJERG CLARKE is a doctoral candidate in Marketing
at the University of Southern Denmark.
scale, the term market is replaced by relationship. Thismarks the change in emphasis from satisfying the cus-tomer in the short term to satisfying the customer in thelong term; however, the exact point is difficult to identify.
The best way to segment a market in a given situationmay depend on what part of the scale the companys focuslies. This distinction has not been made in segmentationtheory before. In Wind and Cardozos  two-step, macro-micro model and in Shapiro and Bonomas  multiplestep, nested approach, the premise is that segments can gofrom being on a very general level to an individual level bybreaking down the segmentation bases until segmentsare acceptable in the terms of accessibility, measurability,and substantiality. Arndt , Hutt and Speeh , andKotler  claim that it may be more suitable to start outfrom the premise that potential buyers are different, andthat the task is to find similarities, to go from the individ-ual level to a general levelin other words to build up.
In an ex post situation, the buyers needs and wants havebeen made manifest. A company that has adapted to theseneeds and wants over time comes close to complex relation-ship management, whereas a company that has not adaptedto the buyers needs and wants while serving the market iscomparable to the simple market transaction situation.
In simple market transaction situations, few, if any, ad-aptations are made to cater for the buyers needs and wants.Who supplies the product is often of little importance to thebuyer, providing major criteria are met, that is, stable deliv-eries, uniform quality, and low price. Suppliers regard sim-ple market transactions as matters of volume and the bestpossible price. In complex relationship management situa-tions, other criteria carry the importance, that is, the capa-bility of adapting to individual customers, the capabilityand willingness to enter development projects, etc.
Segmentation should be based on the current marketsituation; it must identify the main reasons the buyersbuy. One approach is to look at the actual product pur-pose and the needs and wants that arise from particular
usage, because these are the most central in buying orga-nizational behavior .
This understanding of segmentation comes very close toPlank  in his critical review of industrial market segmen-tation. One of Planks major conclusions is that User re-quirements clearly have to be defined in a more systematicmanner. A preliminary conceptual model needs to be de-veloped around user requirements (p. 90) . It is possiblethat neglecting to use the customer as the foundation for de-veloping segmentation models is a fundamental reason forthe problems associated with implementing them.
In an empirical study, Abratt  found that Themost difficult stage appears to be the transaction of re-sults into an effective implementation strategy (p. 79).Abratt found that the three most common variables usedto segment industrial markets are: geographic (87.5% ofthe companies), demographic (62.5%), and how often theproduct is used (62%). Because user requirements are notcommonly used, it comes as no surprise that implementa-tion frequently gives difficulty.
The following material purposes to conceptualize thefoundation and major procedures for an implementableindustrial marketing segmentation model suitable forcomplex relationship management situations. The reasonfor concentrating on complex relationship situations, isthat the industrial marketing literature recommends them,whereas to date most segmentation models [4, 10, 14] arejust based on simple market transaction situations.
CONCEPTUALIZING AN INDUSTRIAL SEGMENTATION MODEL
Before the conceptual frame is outlined, the industrialmarkets characteristics and trends are described. Shethand Sharma  point out the increasing turbulence inindustrial markets and regard relationship marketing asan appropriate strategic response. With increasing tur-bulence in the market place, it is clear that firms have to
Use the customers intentions towards collaboration with the supplier as a
move away from transaction oriented marketing strate-gies and move towards relationship oriented marketingstrategies for enhanced performance (p. 91) .
Tendencies likely to be identified during research aresummarized as follows:
Supplier as a customer Service procurement Crossnational rules Crossnational values Global sourcing Bonding with suppliers Hub and spoke organizations Supply experience curves Partnering Crossfunctional supplier teams
The conceptual segmentation framework must be ableto deal with these tendencies, and relationship marketingappears to be appropriate as relationships help build pre-dictability into the companys environment. Relationshipmarketing requires that any tendencies be dealt with in astrategic manner, as the organizations of both the sup-plier and customer are affected in several ways and at allorganizational levels.
A relationship approach involves trying to understandthe individual needs and wants of the customer. The overallperspective on the customer therefore should be that thecustomer is unique, equal to the main idea of the build-upapproach within segmentation theory . It is important tounderstand all the consequences of using a relationship ap-proach, because managing supplier/buyer relationships isnot an easy task. It normally would be impossible to have a
Segmentations main uses are of both
operational and strategic nature.
FIGURE 1. Relationship marketingconsequences on strategic issuesand segmentation.
close relationship with every customer, so strategic deci-sions need to be made concerning with whom, how close,when, and for what purpose the relationships will be. Inparticular, it is a question of what core competencies shouldbe offered to the customer and what governance conceptsand actual structures to choose  (Figure 1).
When starting the segmentation process, instead ofseeing customers as identical, the build-up approach be-gins by viewing customers as different and then proceedsto identify possible similarities between them. In a turbu-lence market, using a build-up approach is more suitablethan a breakdown approach. The selling company also isable to play a part in shaping its own environment by thechoices it makes. Partnering and bonding are examples ofhow companies can influence the turbulence of their en-vironments. The segmentation model thus should be ableto identify the factors affecting the markets turbulenceand be able to adjust to these by being dynamic.
A segment often is described as something that has to beidentified, chosen, and thereafter targeted by an adjustedmarketing mix . This can be seen as a static (nondynamic)way of dealing with segmentation in markets where custom-ers develop their needs and wants in interaction with theirsuppliers. Segments are developed as a result of the interac-tion processes between the buying and the selling company.Who plays the most active part in initiating the contact andhow the needs and wants are developed may differ consider-ably from situation to situation.
Sometimes the buyer will be the most active part bytaking initiative, because the concept of reverse market-ing is proposing . Conversely, the seller at times maybe the most active part. Therefore, it is not an unreal hy-pothesis that both buyer and seller will take initiatives,developing the buyers needs and wants by cooperating,and simultaneously direct the sellers capabilities and(strategic) development towards these needs and wants.
Derived from this, a more suitable basis for a segmen-tation model is that segments are developed in the inter-action between two or more parties.
THE DIS (DYNAMIC INTERACTION SEGMENTATION) MODEL
The following section gives further detail on how tocreate segments in the more complex, relationship fo-cused part of the market (Figure 2).
The buyers perception of their own needs and wantsstill are regarded as important variable of segment identi-fication. Needs and wants are developed through interac-tion between buyer and seller but also are influenced bythe activities of the competitors and general changes inthe environment (see the DIS model). As an example,mutual product development projects often lead to thedetermining for the buyers needs and wants. Competi-tors will try to influence the buyers perception of hisproblem and the solution . Changes in the environmentalso will a...