Directory UMM :Data Elmu:jurnal:J-a:Journal Of Business Research:Vol50.Issue1.2000:

Fostering Client–Agency Relationships:
A Business Buying Behavior Perspective
J. David Lichtenthal
ZICKLIN SCHOOL OF BUSINESS–BARUCH COLLEGE
THE CITY UNIVERSITY OF NEW YORK

David Shani
KEAN UNIVERSITY

Account acquisition and retention is an ongoing problem facing advertising
agencies. Literature in this area has focused on the criteria used in agency
selection, the factors fostering continuity, and the forces prompting the
break-up of client–agency relationships. However, this classic industrial
service relationship has not been examined from a business-to-business
buying behavior perspective. A study was conducted with top agency
account acquisition personnel. This study found strong support for the
notion that business buying behavior models can be applied to client–
agency relationships. Furthermore, they may be applied to business-tobusiness service transactions as well. Many forces considered unique to
business buying behavior were prevalent for the selection of agency services
according to sales personnel involved in cultivating new business. The
findings suggest that agencies need to emphasize nonspecific campaign

forces effecting agency selection. Moreover, the study also points to the
importance of identifying the effect of internal organizational forces and
the roles buying center members play, side by side with campaign-specific
factors. Directions for future research are noted and managerial implications for business-to-business new account acquisition and selling are also
provided. J BUSN RES 2000. 49.213–228.  2000 Elsevier Science Inc.
All rights reserved.

A

cquiring and maintaining accounts for advertising agencies has always been important (Aaker, Batra, and
Myers, 1996; Russell and Lane, 1996; Wackman,
Salmon, and Salmon, 1987) and crucial to the survival of
agencies (Beard, 1996; Michell and Sanders, 1995; Michell,
Cataquet, and Hague, 1992). The wave of mergers and acquisitions has placed the structure of the advertising industry in
a flux and the competitive intensity faced by agencies has not

Address correspondence to Dr. J. David Lichtenthal, Associate Professor of
Marketing, Zicklin School of Business, Baruch College, The City University of
New York, Department of Marketing, 17 Lexington Avenue, New York, NY
10010. Tel: (212) 802-6516.

Journal of Business Research 49, 213–228 (2000)
 2000 Elsevier Science Inc. All rights reserved.
655 Avenue of the Americas, New York, NY 10010

subsided (Ducoffe and Smith, 1994). Creative consultancies
have developed, complementing conventional agencies, while
the emergence of media independents reflects the breakdown
of the agency-commission system (Michell, 1984). Therefore,
the agency’s role as a business service provider demands they
must optimally ease the communication between the advertiser and agency personnel (Cook, 1989; Tauber, 1986). In
addition, the selling side of agency–client relations has witnessed the creation of positions within agencies that are solely
responsible for soliciting new accounts (New York Times, 1990).
Understanding the forces influencing buying behavior for
advertising agency services is even more important for fostering and stabilizing the traditional client–agency relationship.
Historically, the study of this buyer–seller relationship had
focused predominantly on the target markets of the programs
themselves and had not been viewed in a business-to-business
context. Agencies could incorporate dominant current practices to reflect the fact that clients are likely subject to a myriad
of forces affecting organizational buying behavior (Johnston
and Lewin, 1994, 1993; Hutt and Speh, 1998).

While researchers have focused on issues relating to agency
selection and loyalty in client–agency relationships, they have
primarily done so by studying specific organizational or individual level factors. However, many factors likely influence
the client–agency relationship. For example, client industry
norms and practices and the various roles within client firms
are instances of factors that are neither organizational nor
individual. What we need is a more comprehensive set of
factors that also includes a broader range and variety of environmental and social forces affecting client–agency relations.
Closer inspection of client–agency relationships shows that
these are a subset of the organizational buying context. The
factors affecting client–agency relations resemble those forces
affecting organizational buying behavior (OBB). In the organizational buying behavior literature there is a dominant and
pervasive framework to study the dynamics propelling OBB.
ISSN 0148-2963/00/$–see front matter
PII S0148-2963(99)00014-4

214

J Busn Res
2000:49:213–228

In this study, we attempt to provide a richer representation
of the various forces affecting client–agency relations by
applying a broader and more diverse range of factors likely
affecting the marketing effort for this business service.
The organization of the article is as follows. First, a review
of the literature on client–agency relations is summarized and
the links to forces affecting business buying behavior is noted.
Subsequently, the Webster and Wind (1972a,b) (hereinafter
WW) organizational buying framework was used to identify
a comprehensive list of forces that might affect client–agency
relationships. An exploratory study was conducted that examined the relevance of this list. Implications for future research
and managing client–agency relations are presented.

Literature on Advertising
Client–Agency Relationships
Research in the area of client–agency relations can be classified
into three categories: (1) the criteria used by clients in agency
selection; (2) the factors fostering continuity of client–agency
relations and; (3) the forces prompting the break-up of client–

agency relationships. In brief, much of the research has focused on how agencies and clients come together, stay together
or break up.

Criteria Used in Agency Selection
Cagley (1986) found that perceptions of advertisers and agencies were equivalent on 14 of 25 selection criteria studied.
Based on attribute mean importance, both groups agreed that
agency personnel should have account responsibility. In addition, both sides see agency business and management skills
as important attributes. Agency personnel gave more importance to “relationship” related attributes than did advertisers.
Advertisers, on the other hand, gave less importance to marketing related services supplied by agencies. Self-selection
exists, on both sides of this dyad, pointing to the need for
mutual learning and understanding the other’s viewpoint.
Cagley and Roberts (1984) derived 25 attitudinal statements from discussions with agency personnel and published
literature. Four factors emerged: market analysis (i.e., research, creativity, media selection); operational scale (i.e., size
of accounts, ability to buy media); interpersonal relations (i.e.,
compatibility of personnel); veracity (i.e., strength of recommendations and objection to override client ideas). Industrial
advertisers feel a stronger need for sales promotion ideas and
capability.
Perceptions of creativity are a source conflict in client–
agency relations, among the top managements of the 50 largest
advertising agencies and advertisers (Michell, 1984). Both

sides agreed on the primary importance of the client–agency
relationship. However, the differences in the remaining four
categories studied was pronounced. The clients saw “creativity” as an interorganizational and a process phenomenon while
agencies emphasize creative environments and the personali-

Lichtenthal and Shani

ties of creative teams. Clients view creativity as a more structured process compared with agencies which stressed spontaneity. Successful relationships treat differences in creativity as
a matter for positive action prompting the emergence of an
account planning function.

Loyalty in Client–Agency Relationships
Michell and Sanders (1995) tested a 7-factor 57-item model
for predicting account loyalty among advertisers. Regarding
overall reasons for agency loyalty, clients ranked actual account characteristics well ahead of the other six dimensions
for remaining loyal. However, items with the highest mean
scores are associated with the general process involving suppliers and interpersonal characteristics such as mutual trust, high
caliber personnel, and mutual professional competence. They
reaffirm the findings of Michell, Cataquet and Hague (1992)
that relate loyalty to campaigns that generate sales, compatible

with client objectives and the agency’s closeness to their business. The cause of breakups is the absence of these three
broad aspects. Furthermore, Michell, Cataquet, and Hague
(1992) concluded that a prevailing set of variables exists which
are responsible for the breakup in agency relationships. These
factors appear consistently over time and between the United
States and United Kingdom. Termination is a process rather
than a single decision and that the formal break is related to
specific incidents.
Agency change could be predicted from a comparison of
switchers and non-switchers (Henke, 1995). Compared with
non-switchers, primary decision makers who would change
agencies expressed less satisfaction with agency media skills
and the size of their account (i.e., energy and attention given)
compared with other accounts. Less importance was placed
on creative skills and the agencies’ ability to win awards.
Creative aspects are important to winning the business and
diminish over time. It appears that agencies overestimate the
importance of their creative ability and achievements as the
relationship progresses. Thus, there is a need to focus more
on understanding the unique needs of the client as an “individual.” Buchanan and Michell (1991) used linear logistic regression model to measure the association between observable

structural factors (i.e., account size-age, past switching behavior, agency and client size, product class) in a client–agency
relationship and the risk of failure (i.e., log-odds of a breakup).
Based on more than 1,000 relationships in the United Kingdom, they found that larger accounts are more stable even if
due to shared association. In addition, new accounts may be
less prone to failure while a higher incidence of prior switching
behavior is positively related to subsequent switching behavior. Organizational factors are the more important structural
factors for determining relationship stability.
Clients and agencies do try to match up with similar partners as to client, size, agency size, and account size (Michell,
1988). Major clients remain loyal to larger agencies, in part
because, if necessary, they can replace creative teams to rein-

Client–Agency Relationships

vigorate the account. Major accounts are comparatively more
loyal despite account size preferring continuity. However,
there has been a polarization in client choice of agency type
(Michell, 1984). The largest agencies maintain their net position through adaptation of services provided. Movement has
been away from agencies with higher marketing reputations
toward agencies with higher creative reputations. A clear trend
exists toward the use of media consultants. Three declining

sectors are medium and small size agencies and in-house
direct accounts. Small agencies have declined at a faster rate.
Doyle, Corstjens, and Michell (1980) study the matched
views of companies that switched with those of their former
agency. They identify different perceptions of the reasons for
the breakdown. The primary reason for the account move,
from the agencies’ perspective, is changes in client policy
while from the client’s perspective, dissatisfaction with agency
performance is foremost. Furthermore, the breakdown is a
process of creeping disengagement preceded by clear signs of
vulnerability.

Forces Prompting Break-up
of Client–Agency Relationships
Murphy and Maynard (1996) explored cognitive conflict between agencies and clients comparing decision profiles regarding hypothetical campaigns in five areas. While substantially
in overall agreement with their clients, agency rankings on
the five decision areas did show some discrepancies. Agency
professionals agreed with clients on the importance of message/creativity and then budget. Media planning was third,
while strikingly they gave less weight to market research and
the client/agency relationship. These latter discrepancies were

not surprising, as it is expected that clients will worry about
product development and agencies worry about relationships.
While using pooled data from both groups could obscure
individual differences, cognitive disagreement appears to be
ruled out as a source of conflict. The results, according to
the authors, point to interpersonal factors and organizational
deficiencies that might cause both groups to perceive poor
communication and misunderstanding. Agencies and clients
think alike but often believe they do not.
Effects of client representative role ambiguity could be
found along several dimensions. For task-interactive services
like agency use, a positive relationship exists between experienced role ambiguity and the client’s job tension-anxiety and
perceptions of conflict in agency relations (Beard, 1996). An
inverse relationship exists between role ambiguity and the
client representatives’ tenure/experience, time working with
an agency, and client satisfaction with the agency. Client role
ambiguity has significant impact on four relationship consequences: conflict, tension-anxiety, satisfaction and performance/relationship. Fostering overall client satisfaction suggests that agencies need sales personnel they can train to be
sensitive to a client’s experience of their roles.
Hotz, Ryans, and Shanklin (1982) conducted a study with

J Busn Res
2000:49:213–228

215

the top 100 advertisers and the account executives who manage the accounts. The research goal was to detect sources of
dysfunctional behavior and learn the extent to which each
side is in concord with the other’s shortcomings. The main
factors derived from specific contributory areas include: personnel turnover at the agency, assistance given to the agency
by the advertiser, client organization effectiveness with its
advertising activities, and degree of agreement (on both sides)
about the agency’s role. An inverse relationship exists between
high turnover at the agency and their effectiveness in handling
the account. The client must monitor the quantity and quality
of assistance it is offering its agency and its own effectiveness in
advertising decision making. Agency and advertiser personnel
agree on the role the agency is to play in the client’s marketing
plan. Both sides must present what they expect to put into
the business relationship to reduce conflict.
In summary, we can observe that past research on client–
agency relationships suggests factors that resemble those forces
affecting OBB. Specifically, some criteria used in agency selection can be classified as mainly individual level factors resembling the more micro-forces affecting OBB. Factors such as
mutual learning and perceptual incongruence are used and
are present in WW. Among the factors affecting loyalty in
client–agency relations are interpersonal characteristics such
as mutual trust and professional competence (Michel and
Sanders, 1995). As well, organizational characteristics such
as size noted by Buchanan and Michel (1991) and Michel
(1988) plus changes in client policy (Doyle, Corstijens, Michell, 1980). Some of the forces prompting the break-up of
client–agency relationships are similar to those forces affecting
OBB. In particular, Murphy and Maynard (1996) and Hotz
et al. (1982) emphasized the effects of organizational factors
with the interaction of individual factors. This interaction is
suggested by WW. Beard (1996) called attention to the influence of group forces in the form of role ambiguity and its
impact on sustaining client satisfaction. In the background,
these research streams makes the tacit assumption of an OBB
context.

The Business Buying
Behavior Perspective
Prior literature on agency–client relations has not looked at
the broad forces affecting organizational buying dynamics of
clients and the implications for agency marketing effort. Zeithaml, Varadarajan, and Zeithaml (1988) suggest research on
the contextual relevance of buyer behavior variables holds the
potential for major contribution to the execution of marketing
strategy.
Researchers generally agree (Johnston and Lewin, 1994,
1993; Johnston and Spekman, 1987, 1982; Lilien, Kotler, and
Moorthy, 1994; Hutt and Speh, 1998) that four frameworks
of OBB are the most accepted: those of Robinson, Faris, and
Wind (1967), Sheth (1973), Choffray and Lilien (1977) and

216

J Busn Res
2000:49:213–228

Webster and Wind (1972 a,b). These frameworks have shaped
our understanding of organizational buying behavior since
the 1960s (Hutt and Speh, 1998; Wilson, Lichtenthal, and
Rethans, 1986; Johnston, 1981). A brief overview of each is
provided.
Robinson, Faris, and Wind (1967) developed a process
model known as the BUYGRID framework from descriptions
of three purchase situations. The BUY CLASSES element (i.e.,
straight rebuy, modified rebuy, new task) has guided research
that helped empirically derive the classification of business
buying situations altering the duration of the organizational
buying process. The BUY PHASES element has guided researchers who have empirically derived variations on the total
number of stages and their order in the buying process (Lichtenthal, 1988; Johnston, 1981).
Sheth (1973) developed an organizational level framework
on the process of organizational buying based on case studies
and literature reviews. Three elements emerged: a psychological world of the decision maker, the conditions that precipitate
joint decision making and the process of conflict resolution.
The variables and their linkages are suggested including situational factors. Choffray and Lilien (1977), drawing on the
conceptual work of Sheth (1973) and Webster and Wind
(1972a), develop an empirical model. They focus on the links
between characteristics of the organization’s buying center
and three major stages in the organizational buying process.
They developed an operational model of OBB that explicitly
addresses the output of group decision making by giving a
conceptual framework, measures and methodology for predictive purposes.
The most comprehensive and influential framework is that
proposed by Webster and Wind (1972 a,b). Using case studies
and literature reviews, WW heuristically derived a structure
of the set of four forces influencing OBB (Appendix A).
Their original conceptualization of the forces shaping OBB
is described as a general theory of OBB. Four classes of forces
are seen to act in OBB: environment (E), organization (O),
group (G), and individual (I). A limitation of their approach,
as with most complex models, is that relationships among
these forces and operationalization procedures were not well
specified. WW proposed that environmental forces external
to the firm provide a subtle and pervasive context in which
the other forces operate. Organizational and group forces act
together to determine a frame of reference that guides individuals in their interpretation of other organizational actors.
The ubiquitous nature of WW’s contribution has been
noted by many researchers. Johnston and Lewin (1994, 1993)
used the major models of OBB to organize more than 150
articles in the field. Using the WW framework, they classified
98 studies and found the distribution of the research within
four broad classes was: environmental (8%), organizational
(22%), group (55%), and individual (15%). Among the studies
they reviewed, all four broad classes were used as both dependent and independent measures. While the authors of most

Lichtenthal and Shani

of these articles did not directly cite WW, it apparently acts
as a “background gestalt” for selecting and developing measures for research on OBB. Furthermore, the incidence of
citation, for the WW framework in the Social Science Citation
Index (SSCI) for the first 20 years after its publication (i.e.,
a census from 1972–1991), reveals that the Webster and Wind
(1972b) article was cited 37 times. Researchers relied on WW
in selecting variables in empirical work eight times (22%). In
the remainder, WW was used to discuss buying center roles
11 times (30%), to refer to comprehensive models seven times
(19%), to refer to the process of OBB four times (11%), to
note the complexity of OBB twice (5%), to discuss the need
for certainty and risk reduction twice (5%), and for other
purposes (8%). The WW framework is widely recognized as
a set the forces active on the buyer’s side that could be used
for developing a characterization of business buying behavior.
In the study described below, the WW framework with its
distinct subdivision of four sets of forces into E,O,G,I, is
applied to the client–agency context. It will become apparent
that this framework enriches our understanding of this buyer–
seller relationship as well.

Methodology
Developing Measures of the Forces
To identify a set of forces likely to be active in affecting
buying behavior for services, appropriate measures must be
developed. Hambrick (1984) suggested the use of an existing
framework to guide in the derivation of measures for collecting
data. For example, Bunn (1993a) developed a taxonomy of
buying patterns and situations consisting of six prototypical
buying decision approaches based on numerically derived
patterns similar to previous classification schemes with some
modifications.
In this study, the chapters on environmental, organizational, group, and individual forces (Webster and Wind,
1972b, pp. 40–117) were studied to derive the content of the
items for the instrument. A list of statements was developed
to capture as many distinct aspects of each force as possible.
Following Bailey (1994) and Hunt (1983), the statements
were made to be mutually exclusive and collectively exhaustive. These statements are listed in Table 1.
Of the 72 items uncovered, 15 were about the environment
(E), 23 involved the organization (O), 13 concerned groups
(G), and 21 referred to individual (I) forces. These items were
used to construct a questionnaire that asked sales executives
to rate them as to the frequency with which they encounter
them during the course of converting prospective clients to
accounts. The content of the items were modified to reflect
the context of OBB in buying advertising services. A frequency
scale was chosen over an importance scale because a frequency
scale does not require evaluation (Moncrief, 1986; Walker,
Churchill, and Ford, 1979). For example, as measures for the
set of forces, some items might be active frequently but might

Client–Agency Relationships

J Busn Res
2000:49:213–228

217

Table 1. Rank and Mean Frequency of Each Item by Major Class
Class
Individual (I)
I14:
Client experience with an agency influences its chances to be selected in future
I20:
The expertise of agency personnel affects agency selection
I19:
The trustworthiness of agency personnel affects agency selection
I11:
A favorable predisposition makes a client more receptive to marketing
communication from agency
I06:
Agency selection is based on the agency’s ability to develop programs that
meet the client’s needs
I16:
By obtaining more information about the agency, client personnel reduce risk
in agency selection
I13:
A favorable predisposition helps a client to remember more marketing
communication from the agency
I01:
Client personnel involved have a constant need to be reassured
I15:
An ongoing client-agency relationship leads to habitual reselection of that
agency
I10:
Risk reduction motivates clients more so than other factors to prefer one
agency over another
I03:
Client personnel are confident individuals
I12:
Client personnel “hear what they want to hear” regardless of marketing
communication from the agency
I05:
In selecting an agency, client personnel, regardless of what they believe, act
based on company expectations
I18:
Client personnel change campaign objectives in order to reduce their
individual professional risk
I09:
Personal recognition motivates clients more so than other factors to prefer one
agency over another
I08:
Personal advancement motivates clients more so than other factors to prefer
one agency over another
I07:
In selecting an agency, client personnel regardless of company expectations act
based on their own beliefs
I21:
In order to reduce risk in agency selection, client personnel are loyal to an
agency
I04:
Personnel involved in agency selection formulate their preference independent
of each other
I17:
In order to minimize their risk in agency selection client personnel reduce
their personal involvement
I02:
Client personnel involved are able to tolerate ambiguity
Group (G)
G03:
There is a person who controls the flow of agency related information to other
client personnel
G08:
The opinion of the main negotiator for the client tends to align the
expectations of others involved
G04:
The person who initiated the process for agency selection can be easily
identified
G13:
Among client personnel involved one individual has ultimate decision
authority to select an agency
G02:
There is a person who sets the requirements of the campaign
G01:
There is an individual who negotiates the requirements of the campaign
G07:
Internal bargaining is used by client personnel to resolve conflicts in selecting
agencies
G11:
There is an individual who negotiates but does not set the financial terms for
the client
G06:
Once the agency is selected, there is one person in the client firm who acts as
liaison
G09:
All those involved at the client firm, simultaneously listen to an agency
presentation
G12:
There is an individual who sets but does not negotiate the financial terms for
the client
G10:
There is an individual who develops a list of qualified agencies but does not
participate in negotiation
G05:
People involved in agency selection have roughly equal status in the formal
organizational hierarchy

Rank
Within

Overall
Rank

Mean
(n 5 97)

1
2
3
4

1
2
3
4

6.40
6.30
6.13
6.05

5

5

5.97

6

6

5.96

7

7

5.89

8
9

8
27

5.56
4.76

10

37

4.50

11
12

39
42

4.45
4.32

13

49

4.20

14

51

4.07

15

53

4.04

16

56

3.89

17

57

3.85

18

58

3.69

19

60

3.56

20

64

3.34

21

70

2.42

1

10

5.54

2

14

5.22

3

18

5.05

4

22

4.97

5
6
7

26
31
36

4.76
4.68
4.53

8

40

4.40

9

46

4.28

10

48

4.24

11

52

4.06

12

67

3.07

13

69

2.70

(continued)

218

J Busn Res
2000:49:213–228

Lichtenthal and Shani

Table 1. continued
Class
Organizational (O)
O16:
Personnel involved from client departments are concerned with different aspects
of the agency’s offering
O01:
The client’s corporate marketing goals affect agency selection
O15:
Those involved in agency selection are more loyal to their own department’s
preferences
O14:
The selection of an agency is centralized rather than decentralized in the client
firm
O02:
The client’s campaign objectives influence agency selection
O20:
Client personnel involved engage in extensive information search about
agencies
O22:
The search for agencies is based on past experience with the process
O07:
Formal authority is used by client personnel to resolve internal conflicts
concerning agency selection
O08:
The formal authority of those involved varies over the process of agency
selection
O23:
More than one agency is used in order to reduce uncertainty in agency
selection
O09:
Informal authority of those involved varies over the process of agency selection
O06:
Coordination among client personnel involved is achieved through effective
communication
O17:
More than one person at the client firm has formal authority to decide on
agency selection
O19:
More than one agency is used in order to reduce the client’s risk
O14:
Most client personnel involved are well informed about the issues concerning
agency selection
O12:
The interaction pattern among agency selection personnel is influenced by the
procedures and paperwork followed
O03:
Client personnel use persuasion rather than authority to advance their
preference for an agency
O11:
Financial rewards based on formal evaluation of client personnel affects agency
preference
O10:
Non-financial rewards based on formal evaluation of client personnel affects
agency preference
O05:
Client personnel involved have latitude to deviate from company rules
concerning agency selection
O21:
Client personnel change campaign objectives during the agency selection
process
O18:
A client’s conflicting goals are resolved in the process of agency selection
O13:
Manipulating client organizational factors is more effective than direct
persuasion of their personnel
Environmental (E)
E10:
Competitive intensity in client industry affects the extent of agency services
used
E08:
General availability of capital affects the extent of agency services used
E15:
Client industry business practices and norms affect the extent of agency services
used
E04:
Changes in client industry technology affect the extent of agency services used
E09:
Level of optimism or pessimism affects the extent of agency services used
E11:
Expectations of price changes for agency services affects the extent of agency
services used
E05:
Changes in client industry technology affect the way clients select an agency
E01:
A client’s preference for an agency is influenced by its proximity
E12:
Govt economic activities affecting the client industry influence the extent of
agency services used
E13:
Govt regulatory activities affecting the client industry influence the extent of
agency services used
E06:
U.S. economic growth rate affects the extent of agency services used
E14:
Client industry trade associations affect the extent of agency services used
E07:
U.S. unemployment rate affects the extent of agency services used
E03:
Client firm’s ecological concerns affect agency selection
E02:
Seasonal weather changes have an impact on agency selection

Rank
Within

Overall
Rank

Mean
(n 5 97)

1

9

5.54

2
3

11
15

5.40
5.21

4

16

5.15

5
6

17
19

5.11
5.04

7
8

20
23

5.01
4.97

9

24

4.87

10

28

4.76

11
12

29
30

4.71
4.70

13

34

4.60

14
15

35
38

4.56
4.46

16

45

4.28

17

50

4.19

18

55

3.96

19

59

3.63

20

61

3.49

21

63

3.37

22
23

65
66

3.24
3.24

1

12

5.34

2
3

13
21

5.27
5.00

4
5
6

25
32
33

4.79
4.65
4.62

7
8
9

41
43
44

4.32
4.31
4.31

10

47

4.24

11
12
13
13
14

54
62
68
71
72

4.00
3.46
2.91
2.42
1.46

Client–Agency Relationships

also be deemed not as important by some respondents. To
minimize the halo effect of continuously evaluating items from
one major OBB force and to guide respondents to give careful
and equal consideration to all the forces two steps were taken:
(1) the poles of the scales were alternated throughout the
instrument; and (2) the items were systematically rotated.
These two steps help ensure that sales executives would give
their assessments on a set of measures that is a proxy for the
entire WW framework without signaling its broad content or
structure.
Senior sales personnel from the top 50 agencies in the
Eastern United States were selected from Adweek Agency Directory. Thirty-nine of the fifty firms initially contacted by phone
agreed to participate. Eleven refused citing reasons of confidentiality. Each agency that agreed accepted 10 questionnaires
with postage-paid return envelopes and circulated these to
their sales executives responsible for new account development. To foster participation, guaranteed anonymity was
promised and respondents were not required to identify themselves or their firm. A total of 101 usable surveys were returned, yielding a response rate of 25.8% based on the number
of surveys distributed.

Selecting Key Informants
To ascertain the active forces affecting OBB for business services, it is important to use respondents who may have knowledge about business buying behavior in a broad range of
business markets (Silk and Kalwani, 1982; Bunn, 1993; Kohli,
1989). According to Bunn (1993a) early research relied on the
use of key informants from the buying side (i.e., purchasing
managers’ reports of their own behavior and the behavior of
others involved), and it evolved to include informants from
the selling side (i.e., sales representatives and sales managers).
Earlier, Moss (1979) had reported three case studies wherein
industrial sales personnel were effectively used as a source of
in-depth marketing intelligence about buyers.
Researchers have not used reports on industrial buying
behavior from sales executives frequently, and these reports
should be considered valid sources of information about OBB.
As an observer external to the buying firm, sales personnel
should be able to see the broader set of forces affecting the
OBB process (Anderson, Chu, and Weitz, 1987). Furthermore,
organizational buyers have been shown to hold perceptual
views skewed to favor their own importance and functional
positions in the organization (Silk and Kalwani, 1982). In
effect, organizational actors are unable to “step out of” their
context and location within the firm to obtain a view including
more macro forces. Furthermore, “snowballing techniques”
(i.e., respondent reporting the involvement of another actor
until exhaustion) can result in the overall sample becoming
too small (Wind and Thomas, 1980).
Sales executives in large advertising agencies are plausible
key informants. Typically, the clientele of large agencies includes many organizations from a variety of industries. More-

J Busn Res
2000:49:213–228

219

over, choosing a new advertising agency is a high involvement
(new task) decision for the organizational buyer (Webster,
1990). The process of choosing an agency involves many
individuals over a substantial period of time increasing the
likelihood of the maximum number of forces being active and
experienced by sales representatives. Therefore, the senior
sales personnel of such agencies likely accumulate knowledge
and experience on the buying behavior of various organizations. Campbell (1955) proved that knowledgeable people,
when answering well-designed questionnaires within their
area of expertise, provide quality data. More recently, Weitz,
Sujan, and Sujan (1986) stress salesperson knowledge with
an ability to process buyer information as a determinant of
selling success.
Senior sales personnel, like account supervisors, are typically responsible for convincing the potential client to select
one agency over another, while account executives are responsible for servicing and satisfying the client once on board. It
is the account supervisor’s job to know and understand the
needs and requirements that shaped the clients’ selection of the
agency. They must understand their clients’ buying behavior.
When respondents are asked about specific purchases they
have participated in, it may restrict the ability to generalize
the results (Kohli, 1989; Silk and Kalwani, 1982 in Bunn
1993a). Therefore, sales executives were asked to recall their
most recent selling experiences rather than a decision instance.
Data was collected from sales supervisory personnel believed
to possess experience and knowledge about a variety of organizations using agencies’ services.

Data Analysis
The data was analyzed in several steps. First, the sample was
examined to assess respondent profile and experience. Second,
all 72 items were examined to establish the extent of occurrence of the OBB forces. The pattern of groupings within each
of the four main classes was explored as well. Finally, the
derived framework was examined using all 72 items and compared with the proposed structure of the WW framework.
The pattern of groupings among the four main classes was
also examined.

Respondent Profile and Experience
The confidence that can be placed in the results is in part
contingent upon the characteristics of the respondents as key
informants. An examination of the respondent’s job title, sales
experience, and their industry exposure as shown by their
account activity was conducted.
A majority of the respondents in the sample hold senior
account and upper-level management positions and had sales
responsibilities: vice president, 21.0%; account executive,
19.0%; account supervisor, 17.0%; senior vice president,
16.0%; senior account executive, 6.0%; executive vice president, 6.0%; miscellaneous, 15.0% (n 5 100).

220

J Busn Res
2000:49:213–228

As a group they have considerable experience. On average,
the respondents have been in the advertising business 11.15
years (n 5 96), with their current agency 5.21 years (n 5
98), and in their current positions 3.56 years (n 5 97).
The external validity of the framework that emerges is
partly contingent upon the respondent’s breadth of experience
in converting prospective consumers of agency services to
clients. In an open-ended format, the respondents reported
the three industries or firms with which they most frequently
did business. While information concerning each agency’s
clients is publicly available, the experience of individual sales
executives could vary. The responses showed that 65 industries were represented at the level of detailed industry (4-digit
SIC code), 53 at the industry group level (3-digit SIC code),
and 30 at the major group level (2-digit SIC code). The distribution was flat, with no one industry predominant. Overall,
the respondent group is well qualified as key informants, given
their organizational positions, years of experience and breadth
of industry exposure.

Main Classes of Forces Affecting OBB
The respondents were required to assess the extent to which
the specified OBB items from the WW framework are affecting
their clients’ buying behavior. Table 1 contains the summary
of the means of the 72 items. Each item was scored on a scale
of 1 (never) to 7 (always). Items were considered significantly
active for characterizing buying behavior, if they ranked above
a cut-off level across all respondents of above 3.5. Of the 72
items derived, 60 (83.3%) have a mean extent of occurrence
greater than 3.5. Respondents found these items to be more
active than not in their clients’ buying behavior.
In their chapters devoted to the four main forces, Webster
and Wind, (1972b, p. 40–107) emphasized the interaction
within each class. Factor analysis is preferred to cluster analysis
in this situation because, from a theoretical standpoint, WW
already contains loosely specified classes and plausible groupings (Stewart, 1981). Because cluster analysis is a theoretically
unbounded numerically driven procedure, it is less appropriate. Exploratory factor analysis allows any patterns (i.e.,
on an intra-class and inter-class basis) to emerge within the
theoretical bounds of the WW framework.
Similar to Moncrief’s (1986) use of factor analysis to develop a sales position taxonomy, the following procedures
were used for all factor analyses performed in this study:
(1) consideration was given to any factor initially extracted
(principle components method) with an eigen-value greater
than 1.0; (2) all those factors were subject to varimax rotation;
(3) items included had a loading of 0.40 or greater. This cutoff level was chosen based on a visual examination of all
loadings that revealed a significant drop in the loadings below
0.40; (4) any item with cross-loading of 0.40 or higher on
more than one factor was deleted; (5) as mentioned previously,
items with a mean extent of occurrence of 3.5 or below were
omitted.

Lichtenthal and Shani

For all 60 items surviving there were at least 97 least usable
cases (except one item with 91) even after listwise deletion.
A 2.0 or higher on a semantic scale suggests that the force is
somewhat active. A 3.5 is theoretically the number at which
the respondent group finds this force more active than not.
The full range of the 1–7 scale was used for every item included
in the analysis. These procedures were used to lower the
likelihood of spurious factors or items remaining or being
developed.
Table 2 shows the results from the four separate factor
analyses conducted on each of the set of items from main
forces (i.e., E,O,G,I).
For each of the major four categories, the table shows (1)
the framework proposed by WW; (2) the derived structure
based on the factor analyses; (3) the cumulative variance for
an force; (4) the extent to which the numerically derived
structure corresponds to the WW framework; and (5) a summary of findings for each force. A discussion of the main
findings follows.
In the environmental class, EF1 and EF2 correspond to
the technological and economic subclasses as specified in
WW. EF3 is a hybrid of legal and political subclasses. EF4 is
a hybrid of cultural and economic (industry) conditions. The
physical environment does not appear to be active probably
because of the study’s urban location. In the organizational
class, the eight factors derived use all the items from the
organizational class as suggested by WW. In the group class,
six factors are derived. Four correspond to four of the five
original buying center roles. There was some role fracturing
(i.e., hybrid roles). Respondents indicated that the number of
people typically involved from the client firm during the buying process averages 5.45, with a standard deviation of 1.878
(n 5 95). Other studies also empirically establish that OBB
is a multi-person phenomenon (Silk and Kalwani, 1982; Moriarty and Bateson, 1982). One factor corresponds to a negotiation role and one to the manifestation of lateral relationships.
In the individual class all seven factors derived correspond
to the individual subclasses as specified by WW. Five were
hybrid.

Underlying Structure of the Forces Effecting OBB
The research purpose is to investigate the general structure
of the framework that emerges, and to explore the nature of
the interactions among the OBB classes. First, it was to be
determined if there is a distinction between the external and
internal classes. Webster and Wind (1972b, p. 40) noted that
“Environmental influences are subtle and pervasive. They are
hard to identify and describe, and they provide the context
within which organizational, group, and individual factors in
turn exert their influence.” Therefore, the items that represent
the environment, which is external to the buying organization,
should group together, while the organizational, group, and
individual items should exhibit a higher degree of interaction.
Second, within the set of forces internal to the buying firm,

Environmental (E)
Original WW
Framework

Organizational (O)

Cultural

Goals and Tasks

Economic

Structure (communication, status,
authority, rewards, workflow)
Technology
People

Legal
Physical
Political
Technology
Empirically Derived EF1 Technology effects the
Structure
extent and way agency
services are used
EF2 Economic conditions visa-vis monetary policy and
anticipated price changes
EF3 Legal and political forces
government - regulatory
ativity
EF4 Business practices and
competitive intensity

Group (G)
Gatekeeper, User, Buyer, Decider,
Influencer
Tactics of Lateral Relationship

Individual (I)
Personality
Role Set

Social negotiations
Performance of Buying Committee

Motivation
Learning Process
Interaction with Environment
Preference Structure
Decision Rules
OF1 Authority of those involved varies GF1 Negotiator sets financial terms and IF1 Process-results of learning about
and is used to resolve conflict
selects vendors
the agency
OF2 Marketing and campaign goals

GF2 Liaison-flow of information

IF2 Confidence in agency

OF3 Communication and information
search

GF3 Determines requirements

IF3 Self-aggrandizement

OF4 Risk-uncertainty reduction
through multiple sourcing
OF5 Different people decide about
different aspects of offering
OF6 Selection process is centralized
OF7 Participants are loyal to their
department’s preferences than those
of other departments
OF9 Financial rewards based on
personal evaluation effects selection
58.1%
OF1 corresponds to authority structure

GF4 Opinion leader-ultimate authority

IF4 Risk reduction

GF5 People involved are of equal
status and they bargain
GF6 Initator role

IF5 Independent thinking about agency
selection
IF6 Straight rebuy tendency
IF7 Insecurity and selective perception

60.0%
IF1 is hybrid from learning and
preference structure
IF2 is a hybrid of personality,
motivation and preference classes
IF3 is a hybrid of motivation and
perceived risk classes
IF4 is constituted from the motivation
class
IF5 is a hybrid of personality and role
set
IF6 is a hybrid of role expectations and
learning classes
IF7 is constituted from personality

221

(continued)

J Busn Res
2000:49:213–228

Cumulative variance 55.5%
56.2%
Degree of
EF1 and EF2 corresponds to
GF1 is the buyer’s role
correspondence
economic and technological
forces
EF4 is a hybrid of culture and OF2 & OF6 constitutes goals and
GF2 is the gatekeeper’s role
economic forces
tasks
The physical environment
OF3 corresponds to communication and GF3 is hybrid of de facto decider and
taxon is not active
workflow structure
buying center listening to presentation
OF4 corresponds to behavioral theory GF4 corresponds to social negotiation
of the firm
OF5 & OF7 constitutes the buying
GF5 corresponds to lateral relationships
center as a group
OF9 corresponds to reward structure
GF6 is a role splintered from the user
class

Client–Agency Relationships

Table 2. Factor Analyses on Main Forces

Five elements are hybrid indicating a
high degree of interaction among
these forces.

Three elements are pure buying center
roles.
Two elements are pure group level
classes.
One element is a hybrid of role and
group classes.
One element is a splinter from one role
class.

Two elements are pure.

Lichtenthal and Shani

Two elements are pure and The structure class is constituted from
one is a hybrid.
four elements.
One subclass of the
The task class is a pure element.
framework is not
The people subclass is constituted from
manifest.
two elements.
Summary

Environmental (E)

Organizational (O)

Group (G)

Individual (I)

J Busn Res
2000:49:213–228

Table 2. continued

222

organizational and individual classes should interact with each
other to a higher degree than they do with the group class:
“. . . individual and organizational goals combine in a unique
way to determine a frame of reference or a point of view that
guides each individual and determines their interpretation of
the behavior of other members of the buying center” (Webster
and Wind, 1972b, p. 80). Therefore, for pairwise interactions
of O,G,I class items the O-I combination should exhibit the
highest degree of interaction.
To explore the pattern of groupings among all the major classes,
an exploratory factor analysis on all 72 items simultaneously
was performed. A total of 65 items forming 25 factors survived
the procedure outlined earlier (Table 3).
There are three pure environmental factors. F1 represents
the macro and the industry environment. F7 captures the
effect of the technological environment, and F23 reflects the
geographic proximity of the agency. These environmental factors are meaningful and not contaminated with items from
the other three major classes. The remaining factors consist
of a mix of items from the other three classes. Therefore, the
WW hypothesized broad distinction between E and O,G,I
emerged from the data.
An Item Separation Index (ISI) was developed specifically
for detecting mutual exclusivity of classificational schemes. A
basic requirement for good classification schemes is mutual
exclusivity (Bailey, 1994; Hunt, 1983). The ISI is a relative
measure of the mutual exclusivity of the various classes (i.e.,
the four sets of OBB forces). The index ranges in value between
0 and 100. A high value of the ISI suggests that the items
belonging to a particular class load together and do not to
interact with other classes. A low value of the ISI suggests
that the items from one class interact with items from another;
therefore, the classes are not mutually exclusive. This index
is sensitive in that even one contaminating item in a factor
will show interaction. In addition, only nearly pure factors
will drive the index to high values.
The ISI was calculated as follows: (1) the “pure factors”
were identified (i.e., those consisting exclusively of E or O or
G or I items); (2) for each of the major classes, the number
of items in the pure factors were summed; (3) for each of the
four forces, the number of the items from step 2 were divided
by the total number of items that survived the screening for
that main class (i.e., E,O,G, or I, respectively); (4) that number
was multiplied by 100. The resulting percentage and summary
of the interaction among the items for the analysis of E,O,G
and I is given in Table 4.
The environment class has an ISI value of 80, while the
values for the organization, group and individual force factors
are 35, 44, and 62, respectively. The high ISI value for the
environmental force factors indicates that E items group together to create a relatively pure class. The lower ISI values
for the other three forces suggests that items from those classes
tend to interact. To explore the pattern among just the O,G,I
items, an additional factor analysis (not shown)was done exFACTOR ANALYSIS OF ALL ITEMS SIMULTANEOUSLY.

Client–Agency Relationships

J Busn Res
2000:49:213–228

223

Table 3. Factor Analysis for All For