Вы находитесь на странице: 1из 16

Behavior- and Outcome-Based Sales Control Systems: Evidence and Consequences of Pure-

Form and Hybrid Governance


Author(s): Richard L. Oliver and Erin Anderson
Source: The Journal of Personal Selling and Sales Management, Vol. 15, No. 4 (Fall, 1995),
pp. 1-15
Published by: Taylor & Francis, Ltd.
Stable URL: http://www.jstor.org/stable/40471597
Accessed: 28-07-2016 02:10 UTC

REFERENCES
Linked references are available on JSTOR for this article:
http://www.jstor.org/stable/40471597?seq=1&cid=pdf-reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted
digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about
JSTOR, please contact support@jstor.org.

Taylor & Francis, Ltd. is collaborating with JSTOR to digitize, preserve and extend access to The Journal
of Personal Selling and Sales Management

This content downloaded from 132.174.255.223 on Thu, 28 Jul 2016 02:10:11 UTC
All use subject to http://about.jstor.org/terms
Behavior- and Outcome-Based Sales Control
Systems: Evidence and Consequences of
Pure-Form and Hybrid Governance
Richard L. Oliver and Erin Anderson

Anderson and Oliver (1987) presented a behavior vs. outcome sales control
taxonomy based on methods of monitoring salespersons' effort. Under an
outcome system, sales revenue is thought to be a sufficient criterion for sales
proficiency. Under a behavior system, the sales process itself is critical and
"input" monitoring activities are maximized. Oliver and Anderson (1994) tested
these differences on salesperson perceptions and found effects on affective and
cognitive consequences. This paper extends these findings and shows that, while
conventional behavior and outcome strategies appear to be operating in accord
with theory, evidence is found for an integrated or "hybrid" philosophy exhibiting
degrees of both. The incidence and nature of such strategies are discussed in
terms of their consequences and implications.

Anderson and Oliver (1987), hereafter A&O, pro-


pose two "alternative" managerial strategies which
Richard L. Oliver (Ph.D., University of Wisconsin-Madison) is
the Valere Blair Potter Professor of Management, Owen Gradu- they label as "outcome control" and "behavior con-
ate School of Management, Vanderbilt University. He researches trol" and position as end points on a control system
consumer and sales psychology with a special focus on customer continuum. In their formulation, an outcome-based
satisfaction, service quality, and sales behavior, and holds the
position of Fellow of the American Psychological Association for control system involves relatively little monitoring
his extensive writings on the psychology of the satisfaction of salespeople, relatively little managerial direc-
response. He is co-editor of Service Quality: New Directions in
tion or effort to direct salespeople, and straightfor-
Theory and Practice and has published articles in the Journal of
Marketing, Journal of Marketing Research, Journal of Personal ward, objective measures of results (e.g., sales). In
Selling & Sales Management, Journal of Consumer Research, contrast, a behavior-based control system is typi-
Psychology & Marketing, Behavioral Science, Journal of Eco-
nomic Psychology, Journal of Applied Psychology, Applied Psy- fied by considerable monitoring of activities as well
chological Measurement, Psychometrika, Organizational Behav- as results, high levels of managerial direction and
ior and Human Decision Processes, and others.
intervention in activities, and subjective and more
Erin Anderson (Ph.D., UCLA) is a Professor of Marketing at complex methods of evaluating performance, typi-
INSEAD (The European Institute of Business Administration).
As Europe's first business school, INSEAD is located in cally centered around the salesperson's job inputs.
Fontainebleau, France. Anderson specializes in issues of vertical When compared to outcomes, inputs include per-
integration and organizational control in sales and distribution. sonal qualities (e.g., aptitude), activities such as
She has published in a variety of journals, including Journal of
Marketing Research, Journal of Marketing, Management Sci- new account generation, and sales strategies. They
ence, Marketing Science, and Journal of Law, Economics, and are not themselves indicators of sales results (out-
Organization.
comes) but are expected to generate results at a
The authors gratefully acknowledge the support of the Electron-
later point in time. Performance evaluation on in-
ics Representatives Association (ERA) and the Manufacturers'
Representatives Educational Research Foundation (MRERF), puts is necessarily more subjective because the
and thank the editor and two anonymous reviewers for helpful manager must decide if and to what degree inputs
suggestions. The first author thanks the Dean's Fund for Faculty
Research of the Owen Graduate School of Management for generate outcomes. Moreover, inputs do not lend
providing partial support for this project. themselves to objective measurement, as they are
not always readily countable. This emphasis on in-

Journal of Personal Selling & Sales Management,


Volume XV, Number 4 (Fall 1995).

This content downloaded from 132.174.255.223 on Thu, 28 Jul 2016 02:10:11 UTC
All use subject to http://about.jstor.org/terms
2 Journal of Personal Selling & Sales Management

Table 1
Characteristics of Behavior- and Outcome-Based Sales Control Systems

Behavior Control

Structured Low to little structure

Fixed compensation (salary) Variable compensation


(commission, bonus)

Management monitors behavior Management monitors results


more than results more than behavior

Much supervision and many Little supervision and few managers


managers (low span of control) (high span of control)

Much contact with management Little contact with management

Much reporting Little reporting

Performance evaluated subjectively Performance evaluated on a few


on many criteria observable results

Firm bears risk Salesperson bears risk

puts is fundamental to the strategy A&O refer to as consequences of management's strategy should vary
behavior control. Differences between a control sys- proportionately as the sales organization's philoso-
tem which monitors behaviors as inputs versus one phy trends either toward behavior or outcome con-
monitoring outcomes only are highlighted in Table trol. Thus, one would expect that the transition
1. from one form of control to the other would result in
As discussed in A&O, a number of cognitive, af- fairly smooth changes in the consequences. For ex-
fective, and behavioral differences are predicted to ample, A&O suggest that behavior control should
result from the two control philosophies. Generally, foster less extrinsic and more intrinsic motivation.
behavior control governance (as opposed to outcome If this is so, then salespeople under more outcome
control) should foster greater attention to company control would be expected to report more extrinsic
and product knowledge, greater affect (e.g., satis- and less intrinsic motivation as management's con-
faction, commitment) toward the firm and accep- trol philosophy becomes more outcome control ori-
tance of company procedures, greater levels of in- ented.
trinsic vs. extrinsic motivation, greater focus on Oliver and Anderson (1994), hereafter O&A, have
planning and prospecting behaviors as opposed to recently tested their propositions using a diverse
selling, and lower (but higher quality) sales results. group of sales representatives from a large number
The descriptions above illustrate the differences of sales organizations (rep firms) in the electronics
one might expect to observe at the extreme. These components industry. Using their philosophical cri-
extremes might suggest that the two control sys- teria for the control system differences and empiri-
tem philosophies are mutually exclusive in that the cal findings from scale validation, they constructed
respective elements cannot coexist within the same an index of behavior to outcome control. This scale
sales organization. However, implicit in the A&O was then used to judge individual salespeople as to
propositions is the possibility that firms alterna- the degree of perceived behavior-outcome control
tively may position themselves along the continuum used by management, in the perspective of the re-
and selectively use elements of both systems. In sponding salesperson. The index score was then
such a case, the cognitive, affective, and behavioral correlated with the predicted cognitive, affective,

This content downloaded from 132.174.255.223 on Thu, 28 Jul 2016 02:10:11 UTC
All use subject to http://about.jstor.org/terms
Fall 1995 3

and behavioral consequences, also as reported by cognitive, affective, and behavioral profiles im-
the same salesperson. Generally, the predicted ef- plied by the A&O propositions. No form of con-
fects of control philosophy on selected affective and trol exists in isolation. Rather, controls (may)
motivational consequences were supported. How- combine synergistically to influence the attain-
ment of a given objective... .
ever, counter to the hypothesized effects, the na-
ture of the control system did not appear to impact Additionally, Ouchi (1979; Ouchi and Maguire
sales strategies or aspects of performance. 1975) proposes that, under conditions of high and
Additionally, O&A contrasted their results with low combinations of: (a) ability to measure outcomes
those of Cravens et al. (1993), who also examined and (b) knowledge of the transformation process by
elements of the A&O propositional inventory. Us- which inputs are converted to outcomes, four gover-
ing different philosophical and methodological as- nance possibilities emerge. If measurement ability
sumptions, the Cravens et al. findings were gener- is high and transformation knowledge is imperfect,
ally in accord with those of O&A. Specifically, both outcome control is recommended for the reason that,
studies found associations between a behavior-based if the optimal behavior patterns are unknown, be-
supervisory style and affect, as well as acceptance havior monitoring is wasteful. Alternatively, if
of company procedures. Further, Cravens et al. fo- transformation knowledge is accurate and measure-
cused greater detail on performance outcomes and ment ability is low, then behavior control is recom-
found support for a greater range of performance mended using the same frame of logic. Moreover,
effects, particularly with respect to how effectively under ideal conditions of accurate knowledge and
the sales force uses inputs. Although they found no high ability to measure outcomes, either form of
direct effect of supervisory style on the more out- control and conceivably both are feasible and nei-
come-oriented performance measures, they did de- ther is a priori preferable. Under the fourth alter-
tect indirect effects. native of low ability and uncertain knowledge, a
Restricting discussion to the O&A findings, O&A "clan" structure is recommended, but is not consid-
allow that their results are of a "modest order of ered further here as it is believed to be a relatively
magnitude" and speculate that a potentially un- rare occurrence (Ouchi 1979).
known dimension or dimensions may be missing Although data on the frequencies of use of these
from the index. Alternatively, they propose as a strategies are not abundant, the potential "inde-
"fascinating issue" the possibility that salespeople pendence" of outcome and behavior control as sub-
are not facing control system elements which A&O stitutable or complementary approaches has been
argue to be normatively correct. Indeed, while O&A tested in Ouchi and Maguire (1975). These authors
argue that behavior and outcome control are oppos- found that the correlation among perceptions of
behavior and outcome control received from man-
ing philosophies and suggest that firms may locate
in intermediate positions on the behavior-outcome agement was 0.18, a significant but modest asso-
continuum (e.g., moderate supervision and balanced ciation. This suggests that the two control mecha-
salary/commission plans), they do not entertain the nisms may be used concurrently, at least in the
possibility of combinations of opposing philosophies Ouchi and Maguire context (retail department
(e.g. straight salary but little supervision). In fact, stores).
they argue that such opposing strategies are "inter- Thus, the supportive but modest pattern of the
nally contradictory," and go on to say that "few, if O&A results suggests that the transition from be-
any, managers will create 'inconsistent1 combina- havior to outcome control (or from outcome to be-
tions" (p. 55). havior control), as tapped by their index, may not
Other researchers, however, have proposed hy- cause the hypothesized consequences to respond in
brid strategies. Jaworski (1988), for example, cites linear fashion. This may be so because managers
evidence (e.g., Kim 1984; Porter, Lawler, and Hack- may be violating the A&O normative prescriptions
man 1975; Terborg and Miller 1978) that behav- in that intermediate positions on the O&A index
ioral (process) and outcome control mechanisms are may actually reflect the use of opposing strategies
effective when used in combination. He states (p. as contrasted with intermediate strategies.
31): To investigate this possibility, further examina-
tion of the O&A data using categorical (cluster)
A reasonable set of questions, then, is to what
extent: (a) do sales organizations exhibit these
analysis is undertaken here. Categorical analysis
divergent control philosophies, and (b) do sales- is useful when it cannot be expected that a combi-
people within these organizations display the natorial index is linear in all of its component vari-

This content downloaded from 132.174.255.223 on Thu, 28 Jul 2016 02:10:11 UTC
All use subject to http://about.jstor.org/terms
4 Journal of Personal Selling & Salée Management

ables and the variable weightings or transforma- The salespeople who participated in this study
tions are not known a priori. It allows for non- were employed by sales agencies which represent a
unidimensionality in what would otherwise be a cross section of the membership of a major elec-
single bipolar composite. tronic components trade association (Electronic Rep-
Two reasons for examining the nature of hybrid resentatives Association, or ERA). ERA members
governance are noteworthy. First, it would be in- are independent companies (manufacturers' rep-
structive to see what form or "mix" of behavior and resentatives) which sell electronic components
outcome strategies are used by hybrid firms. Under (broadly construed, as some are actually subsystems
the assumption that firms using elements of seem- that can become components of larger systems) to
ingly incompatible strategies have reached a work- business customers. The products may then be
able equilibrium between emphasis on behaviors resold or incorporated into another product. Both
and outcomes, it is not unreasonable to ask what the product lines and the customer base are broad
form such an equilibrium would take. and diverse. The long gamut of products that con-
Second, the consequences of hybrid governance tain some electronics (i.e., electronic components)
can be examined and compared to those of behavior goes into many different markets (e.g., communica-
and outcome-control firms. Does the dual use of tions, instrumentation, consumer products) and to
elements of both strategies combine to help or hinder many types of buyers. Customers include original
the firm's goals and in what ways? Answers to these equipment manufacturers, assembly houses, dis-
questions were pursued as described in the follow- tributors, construction firms, and retailers.
ing discussion. Because the products and buying tasks vary
greatly, so does the nature of the purchasing unit
and purchasing process. While most sales tasks are
Method
straight rebuys involving individual salespeople, a
manufacturers' representative in the electronic com-
Sample
ponents industry can encounter any type of selling
Data were obtained from sales representatives situation. For example, Céspedes (1992) addresses
employed by a diverse set of independently owned the difficulty inherent in coordinating sales efforts
and operated sales agencies in the electronic com- across multiple locations of a key account purchas-
ponents industry. A sales agency is an independent ing many product lines. These difficulties can arise
company, analogous to an advertising agency or a not only for large manufacturers but for manufac-
consulting firm. The Census Bureau labels such turers' representatives as well.
companies "manufacturers' representatives" or While some writers continue to view manufactur-
"manufacturers' agents" (see Anderson 1986). A ers' representatives as small companies that run
manufacturers' representative provides selling ser- their sales forces in typical outcome-control fash-
vices to manufacturers, and is compensated by com- ion, paying their employees straight commission,
mission on sales, which becomes the first line of the offering minimal supervision, and incurring low
company's income statement (i.e., the bulk of its sales overhead (e.g., operating from the proverbial
revenue). From its income, the manufacturers' rep- office in the owner's garage), this characterization
resentative (the company) must cover all expenses may not be true today. In fact, there is no empirical
relating primarily to personnel who, in turn, are evidence to suggest that manufacturers' represen-
primarily salespeople. The company's owners (typi- tatives as a group are so homogeneous, small, and
cally a handful of founders and top employee sales- simple, and there is no theoretical reason why they
people, perhaps including a layer of professional should be. These companies are sales forces, like
managers) direct the firm's affairs, making deci- any other, and face all the decisions (organization,
sions such as which manufacturers to represent, compensation, supervision, evaluation, hiring, fir-
which customers to pursue, and how to manage all ing, training, and so forth) that other sales forces
employees, including salespeople. One important encounter. The manufacturers' representative can
decision is how to pay salespeople in terms of level be thought of as a free-standing company which
of payment, form of payment, level of benefits, and concentrates on its core competence - selling - and
so forth. In particular, a manufacturer's represen- does not produce what it sells. The company can be
tative as an entity is paid on commission, but the quite large and can involve a considerable infra-
organizational entity need not pay its own employ- structure in terms of management and support sys-
ees in the same fashion. tems (e.g., information technology, clerical support,

This content downloaded from 132.174.255.223 on Thu, 28 Jul 2016 02:10:11 UTC
All use subject to http://about.jstor.org/terms
Fall 1995 5

a corporate jet, a dedicated commercial office build- were measured with two and three scales, respec-
ing). Hence, considerable variation in control sys- tively, scored in the direction corresponding to be-
tems used by manufacturers' representatives is to havior control. Element (3) was the self-reported
be expected, an observation supported by the present percent of salary in the salesperson's compensation
data. plan.
This study used the O&A sample of 347 sales- The two supervisory style scales measuring clas-
people with complete data on the control system sification element (1) contained items tapping ex-
classification criteria. The average respondent was tent of supervision and an absence of a bottom line
male (92%) and 39 years of age, was a college gradu- orientation. The three evaluation scales represent-
ate (64%), had been in sales for 12 years and the ing element (2) contained items measuring infre-
present job for 5.5 years, and received $49,500 in quent outcome monitoring, process monitoring via
total compensation in the past year. Fourteen per- paperwork, and process monitoring via effort and
cent of the sample were on full salary; 23% were on attitude evaluation. Further details are provided in
full commission; all others were on combination O&A.
plans. To form a behavior control and outcome control
Data on the sales agencies represented by the (BCOC) index (see Bollen and Lennox 1991;
final sample of respondents showed that they were MacCallum and Browne 1993), the two supervision
diverse in a number of key firm variables. The manu- and three evaluation subscales were converted to z-
facturers' representative may cover its market area scores and additively combined with percentage of
(say, a three-state region) with several geographi- salary in the pay plan (also a z-score) to form an
cally distinct sales offices. It is the salespeople in index of behavior control (versus outcome control -
these offices (i.e., the employees of the manufactur- represented by lower scores).
ers' representative) whom we studied. The average Hypothesized consequences. All measures of the
number of sales offices per firm was three and hypothesized consequences of a behavior vs. out-
ranged from one to 19. The number of outside sales- come control philosophy are discussed in O&A. They
people employed by each firm averaged five in num- are used here without modification.
ber and ranged from one to 53, while the number of
office personnel, including managers and staff,
ranged to 96. Each rep firm served an average of Analysis
fifteen principals, ranging from one to 40. Finally,
the average number of accounts handled by each A fe-means clustering approach was chosen be-
salesperson was 63, ranging from five to 300. This cause of the large sample size and because cluster
level of heterogeneity in both salesperson charac- memberships were not known in advance. Ap-
teristics and firm data allows for a greater diversity proaches based on fe-means are iterative in that
of governance strategies across salespeople than many cluster center estimations are calculated, all
would be evident for more structured industry based on a pre-specified number (fe) of clusters, un-
samples. til no cluster centroid can be found which improves
cluster separations. Using the z-scores from the
Measures behavior control classification criteria as clustering
variables, k was varied from a low of two until
Control system classification criteria. A&O iden- solutions containing less than 10% of the sample
tified three characteristics of behavior vs. outcome size emerged. This sample size deterioration was
control sales philosophies. Specifically, more behav- approached at a four cluster solution and occurred
ior-oriented control systems have: (1) a higher ex- rapidly at five clusters. Based also on interpreta-
tent of supervision and contact between salespeople tion and goodness of fit, the three cluster solution
and their managers, and a supervisory style em- was selected for further analysis. As noted, higher
phasizing process over results ("bottom-line") be- order cluster solutions produced small sized addi-
haviors, (2) more subjective evaluation methods tional clusters and only marginal reductions in the
based on an emphasis of process behaviors over error sums of squares function. All clustering vari-
outcome results, and as an implementation proce- ables contributed significantly to the cluster sepa-
dure (3) a greater proportion of salary (commission rations at the .01 probability level by one-way
and bonus) in the pay package. Elements (1) and (2) ANOVA.

This content downloaded from 132.174.255.223 on Thu, 28 Jul 2016 02:10:11 UTC
All use subject to http://about.jstor.org/terms
6 Journal of Personal Selling & Sales Management

Results evaluation mean scores tend to be higher than the


paperwork/call-related scores, which indicates that
To interpret the three cluster solution, the dis- all groups monitor attitude and effort to some ex-
play of cluster means is reported in Table 2. Shown tent. The percent of salary figures display wide
are the BCOC index means, and the index compo- variance, a statistic masked in the standard score
nent means (the two supervision scales, the three data. The percent salary reports range from 46%
evaluation scales, and salary). The top half of the (Cluster 3) to 62% (Cluster 1), a pattern consistent
Table reports the actual z-score means while the with the behavior vs. outcome control distinctions
bottom half shows the raw score scale means. These made earlier.
latter data are not mean-adjusted. Table 3 shows the cluster differences in the list of
Cluster 1 (26% of the sample) is shown to have all consequent variables in O&A. In some sense, it is
of the characteristics of a behavior control philoso- analogous to Table 5 in O&A, where behavior con-
phy with the exception that respondents report that trol and outcome control are considered as polar
their sales offices use less paperwork and call-re- opposites on the same unidimensional continuum.
lated inputs than most. This first group reported In many cases, the results in Table 3 are similar to
experiencing high levels of supervision, relatively those of O&A. For example, the product/company
low emphasis on quantifiable results (sales, profits) knowledge scale, which was not correlated with the
and a management mentality that stresses what behavior-outcome control index, also did not reveal
they do, not just the results they achieve. These significant differences across clusters. The findings
salespeople perceive that management pays close for the organizational commitment scale, however,
attention to whether they display favorable job at- were consistent with predictions in both analyses.
titudes and whether they seem to be trying hard. Table 3 shows that the average commitment score
Salespeople in this group, on average, have the high- is indeed higher for behavior control groups and
est percentage of salary-dependency among the highest of all for BC1, as the simple correlation of
three groups. This cluster will be referred to as 0.27 suggested in O&A.
behavior control (1) or BC1. Generally, the cluster means provide an explana-
Cluster 2 (39%) also shares the characteristics of tion for the more modest levels of associations found
a behavior control firm except that these salespeople in the O&A correlational results. It appears that
perceive that management uses an above average both the BC1 and BC2 groups respond similarly to
level of outcome evaluation methods. Supervision, many of the hypothesized consequences. Examples
high levels of input evaluation, and average use of are risk-seeking preference, extrinsic motivation,
salary in the plan mark this cluster. It appears that selling smarter, and sales expense control. In many
the sales offices represented by this group are heavy cases, however, the BC2 group scores exceed those
users of any and all evaluation methods, including of BC1. Examples are sales expertise, teamwork,
outputs, within a behavior control philosophy. This and job satisfaction. This latter pattern would tend
cluster will be referred to as BC2. BC2 appears to to reduce the zero-order correlational results be-
be a hybrid control system, somewhat closer to be- tween the behavior-outcome control index and vari-
havior than to outcome control and is distinguished ables more highly related to the BC2 group, as the
by its extensive use of multiple criteria, spanning BC1 group scored marginally higher on the BCOC
both inputs and outputs. index (Table 2).
Finally, Cluster 3 (35%) appears to be an outcome To test for the similarity of BC1 and BC2 versus
control cluster. The level of supervision is well be- OC (i.e., a stepped response function), F-tests of the
low average, a bottom-line orientation prevails, both overall group differences and a test of the contrast
measures of input-related evaluations are below between the BC groups and the OC group are also
average, and the percentage of salary in the plan is shown in Table 3. Generally, this contrast was sig-
lower than the other clusters. Unlike a "pure" out- nificant in all instances of a significant ANOVA
come control philosophy, however, output monitor- finding. Thus, it appears that variations in the two
ing appears less frequently than one would expect. behavior control philosophies as discovered here do
Nonetheless, this cluster will be designated as one not change levels of the consequent variables as
of outcome control (OC). much as the difference between behavior control
The raw scale scores mirror the pattern of ^-scores and outcome control.
except that actual levels of the representative vari- To further understand the manner in which the
ables are shown. For example, the attitude/effort consequent variables respond to the BCOC index, a

This content downloaded from 132.174.255.223 on Thu, 28 Jul 2016 02:10:11 UTC
All use subject to http://about.jstor.org/terms
Fall 1995 7

Table 2
Cluster Centroids for the Three Cluster Solution

(1) (2) (3)


Control factor

Mean cluster Ζ scores

BCOC Index .291 .212 -.460


Extent of supervision .382 .448 -.797
Not bottom-line oriented .800 .141 -.766
Infrequent outcome monitoring .695 -.627 .183
Evaluation subjectivity
(paper/call related) -.672 .762 -.353
Evaluation subjectivity
(attitude/effort related) .306 .515 -.816
Salary in plan .233 .031 -.212

Raw scale scores*

Extent of supervision 5.402 5.482 3.994


Not bottom-line oriented 5.584 4.802 3.723
Infrequent outcome monitoring 3.733 2.436 3.231
Evaluation subjectivity
(paper/call related) 2.777 4.581 3.178
Evaluation subjectivity
(attitude/effort related) 5.956 6.157 4.878
Salary in plan (%)

'Based on 7- point scales (with the exception of salary)


Note: All cluster differences significant at ρ < .01 .

summary of the significant findings in Table 3 is not tenable. Only three variables (commitment, par-
reported in Table 4. The first column shows the ticipative decision making, and perceptions of pay
instances where the BC1 group scored higher than as controlling) met this criterion and, in all three
the BC2 group which, in turn, scored higher than cases, scores for both BC1 and BC2 were statisti-
the OC group as a linear assumption would sug- cally similar. Thus, although the BC1 scores were
gest. The second column shows the instances where higher than that of BC2, they were not sufficiently
the BC1 and BC2 groups were largely similar in so to rule out the step function assumptions.
score, but were statistically greater than the OC In six cases (extrinsic motivation, direction of
group. In effect, this column displays variables motivation, selling smarter, controlling expenses,
where behavior control results in higher variable satisfaction, and a supportive organizational cul-
levels than outcome control, as the A&O proposi- ture), the step function findings fit the data best. In
tions would suggest. Finally, the last column shows the remainder of cases (nine in all), the hybrid group
instances where the hybrid group scores higher than scored higher than the OC and the BC1 groups,
both BC1 and OC. In some cases, the hybrid group although in five of these, the BC1 group could not
could be considered unique in that its score domi- be considered statistically lower. Note that the three
nated the other two groups (e.g., intrinsic motiva- critical performance variables (relative performance,
tion), while in others, it generated scores higher performance rank, and the pursuit of performance
than the BC1 group, but not statistically so. goals) were best explained by the "hybrid high" ex-
The results show that the "linear" assumption is planation.

This content downloaded from 132.174.255.223 on Thu, 28 Jul 2016 02:10:11 UTC
All use subject to http://about.jstor.org/terms
8 Journal of Personal Selling & Sales Management

Table 3
Cluster Means on the Hypothesized Consequences of Control Philosophy

Contrast
Cluster Cluster One-way 0,1.-2)

Cognitions/capabilities
Product/company knowledge 1 . BC1 3.68 2.95 0.53
2. BC2 3.36
3. OC 3.58
Sales expertise/competence 1.BC1 5.31 8.19b 3.70*
2. BC2 5.44
3. OC 5.05

Affects/attitudes
Organizational commitment 1.BC1 5.91 16.54" 5.74b
2. BC2 5.77
3.OC 5.19
Accepting of authority/direction 1.BC1 5.37 9.65" 4.02b
2. BC2 5.52
3. OC 5.04
Accepting of cooperation as part
of a sales team 1.BC1 5.61 6.86b 2.79o
2. BC2 5.82
3. OC 5.49
Accepting of performance
reviews 1.BC1 4.37 4.75" 2.05*
2. BC2 4.69
3. OC 4.26
Risk seeking preference 1 . BC1 3.76 0.83 -1 .28
2. BC2 3.77
3. OC 3.92

Motivation
Nature of motivation (Intrinsic) 1.BC1 4.79 7.69" 2.01·
2. BC2 5.22
3. OC 4.78
Nature of motivation (Extrinsic) 1 . BC1 2.98 9.45b -4.30"
2. BC2 2.96
3. OC 3.67
Direction of motivation (Agency) 1.BC1 0.45 3.97· 2.50a
2. BC2 0.47
3. OC 0.42

Behavioral strategy
Planning 1.BC1 4.70 3.92· 1.08
2. BC2 4.96
3. OC 4.73
Call activity (% of total time) 1.BC1 0.22 0.55 -0.61
2. BC2 0.24
3. OC 0.24
continued . . .

This content downloaded from 132.174.255.223 on Thu, 28 Jul 2016 02:10:11 UTC
All use subject to http://about.jstor.org/terms
Fall 1995 9

Sell/non-sell time ratio (%) 1.BC1 0.50 0.20 0.39


2. BC2 0.51
3. OC 0.50
Selling technique ("Smarter") 1.BC1 5.35 3.23a 2.51·
2. BC2 5.36
3.OC 5.15
Selling technique ("Harder") 1.BC1 4.43 2.88 1.03
2. BC2 4.71
3. OC 4.45

Performance
Relative performance scale
(outcome) 1.BC1 6.69 5.45" 1.16
2. BC2 7.16
3. OC 6.77
Relative performance rank (outcome) 1 . BC1 3.47 7.76" -0.59
2. BC2 4.06
3. OC 3.84
Pursuing sales and profit (outcome) 1.BC1 4.34 15.07b 2.25a
2. BC2 4.97
3. OC 4.40
Information gathering (input) 1.BC1 4.93 2.35 1.30
2. BC2 5.17
3. OC 4.89
Sales expense control (input) 1 . BC1 5.85 2.21 2.07a
2. BC2 5.86
3. OC 5.67
Sales presentation planning (input) 1.BC1 5.37 15.42" 3.72"
2. BC2 5.80
3. OC 5.21
Job satisfaction 1.BC1 5.81 8.97" 3.94b
2. BC2 5.92
3. OC 5.53

Participative decision-making 1.BC1 5.16 48.72" 9.87"


2. BC2 5.01
3. OC 3.82

Pay as a control mechanism 1.BC1 2.79 13.27" -3.35"


2. BC2 3.43
3. OC 3.54

Organizational culture
Innovative 1.BC1 2.60 18.51" 5.13"
2. BC2 2.81
3. OC 2.36
Bureaucratic 1.BC1 1.75 3.27" 0.16
2. BC2 2.10
3.OC 1.91
Supportive 1.BC1 3.02 31.46" 7.93"
2. BC2 2.95

•p < .05, "p < .01

This content downloaded from 132.174.255.223 on Thu, 28 Jul 2016 02:10:11 UTC
All use subject to http://about.jstor.org/terms
10 Journal of Personal Selling & Sales Management

Table 4
Summary of Order Effects in the Between-Cluster Findings
(Significant Relationships Only)

Effect "Linear Effect Stepped Hybrid High


Dimension

Cognitions/capabilities
Sales expertise/competence * *

Affects/attitudes
Prganizational commitment * *
Accepting of authority/direction * *
Accepting of cooperation as part
of a sales team * *
Accepting of performance
reviews * *

Motivation
Nature of motivation (Intrinsic) *
Nature of motivation (Extrinsic) *a
Direction of motivation (Agency) *

Behavioral strategy
Planning *
Selling technique ("Smarter") *

Performance
Relative performance scale
(outcome) *
Relative performance rank (outcome) *
Pursuing sales and profit (outcome) *
Sales expense control (input) *
Sales presentation planning (input) *

Job satisfaction *

Participative decision-making * *

Pay as a control mechanism *a *a

Prganizational culture
Innovative * *
Bureaucratic *

•OC high for this variable

Discussion chotomy presented in A&O. The results also sug-


gest that intermediate positions between these two
The results presented here suggest that there philosophies may be taken to strike a balance be-
may indeed exist two basic philosophies of sales tween the organization's circumstances and its goals.
control which follow the behavior vs. outcome di- Moreover, the primary intermediate position found

This content downloaded from 132.174.255.223 on Thu, 28 Jul 2016 02:10:11 UTC
All use subject to http://about.jstor.org/terms
Fall 1995 11

Table 5
The Three Control System Groups

Characteristic

Supervision High Very high Very low

Importance of attitude and effort


in evaluating performance High Very high Very low

Emphasis on quantitative results


in evaluating performance Low Very high High

Philosophy that results are more


important than behaviors Low Neither High

Importance of complete, accurate


paperwork Very low High Low

Average percent salary in annual


compensation 62% 55% 46%

Size of group in sample 26% 39% 35%

here apparently does involve seemingly opposing heavily on a "paper trail" to track their salespeople's
governance strategies. A summary of the charac- behavior.
teristics of the three governance modalities appears Interestingly, salespeople in the "pure-form" out-
in Table 5. come control group perceive that management pays
The results showed evidence for a relatively "pure- some attention to whether they display favorable
form" behavior control group, as the A&O discus- job attitudes and whether they seem to be trying
sion would predict. The description of this group hard, but this concern is less pronounced than in
sounds much like one would expect from a behavior the other two clusters. Additionally, this group
control system, and indeed these salespeople, as a scores very low on the extent of supervision, high
group, score the highest on the behavior control on bottom-line orientation, and very low on subjec-
index. However, there is one exception to the hy- tive evaluation of attitude and effort. Salespeople
pothesized behavior control profile, which may be in this group, on average, derive 46 percent of their
peculiar to the firm sample used here. Compared to annual income from salary, the lowest percentage
salespeople from the other two groups, those in among the three groups.
straight behavior control systems report their man- Generally, this description sounds much like one
agements pay minimal attention to paperwork. They would expect from an outcome control system, and
apparently perceive that management is not par- indeed these salespeople, as a group, score the low-
ticularly concerned about whether their paperwork est on the behavior control index. Again, however,
(call reports and the like) is accurate, complete, and there is an exception. Compared to salespeople from
timely. Salespeople may think that, in these sys- the other two groups, those in straight outcome
tems, management relies more on its own subjec- control systems reported that their managements
tive impressions of a salesperson's effort and atti- were somewhat concerned about whether their pa-
tude rather than inferring these from the perwork is accurate, complete and timely.
salesperson's reports. One can only speculate that Additionally, the outcome control group appears
this relative lack of emphasis on paperwork may be to think that management is somewhat more con-
an anomaly to this sample because logic would dic- cerned about paperwork than in the straight be-
tate that behavior control systems should rely havior control group. This belies the common image

This content downloaded from 132.174.255.223 on Thu, 28 Jul 2016 02:10:11 UTC
All use subject to http://about.jstor.org/terms
12 Journal of Personal Selling & Sales Management

that salespeople in outcome control systems don't ceived to be making very heavy demands (proper
have to put "pen to paper" as long as they generate paperwork, good attitude and effort, good sales re-
results. Apparently, the firms in the sample using sults) and to be supervising heavily in order to see
outcome control ask for something of a paper trail their demands met. Thus, support for Ouchi's (1979)
from their salespeople to compensate for their rela- third governance mode corresponding to an appar-
tive lack of direct observation. This, too, is a specu- ent understanding of transformation knowledge and
lation and is contrary to the profile of an outcome- a perceived ability to measure outcomes accurately
control group as defined by A&O. Researchers are is provided.
encouraged to probe into this "paperwork anomaly" This hybrid group scores midrange on the behav-
to further understand the meaning and significance ior control index, and salespeople in this cluster, on
of this finding. average, derive 55 percent of their annual income
The existence of the large BC2 cluster would seem from salary. Hence, in these hybrid systems, man-
to suggest that a large proportion of behavior con- agement insulates salespeople from some market
trol sales offices are employing hybrid evaluation risk but still puts, on average, 45 percent of their
strategies. The data suggest that this "mix" of strat- income at risk, thereby obliging salespeople to share
egies employs high levels of input and outcome moni- management's concern with turning in quantifiable
toring. This appears to be behavior control with results.
greater emphasis on results than would be expected Despite these differences, the hybrid group ap-
from a "pure" behavior control sales office. It sug- pears more closely aligned with the BC1 group than
gests that management may believe that inputs with the OC group. This is apparent from the Table
such as effort and attitude are the first require- 4 results where more score "matches" were evident
ments of sales results and that measurement of for the two BC groups than for BC2 and OC. Even
outcomes is necessary to calibrate proper levels of when the BC2 group appeared unlike either BC1 or
these inputs. Managements may wish to consider OC, the BC2 scores were usually closer to those for
this "heavy evaluation" strategy as a viable alter- BC1 than for OC.
native to pure behavior control.
Thus, as in behavior control, salespeople in the System Preferences
hybrid group reported experiencing high levels of
supervision and something of a management men- It may be that all three systems represent defen-
tality that emphasizes what they do, not just the sible, coherent management philosophies and styles.
results they gain. Also, as in behavior control, these All three offer advantages and drawbacks. To date,
salespeople apparently think management stresses no evidence exists to suggest that one system is
attitude and effort heavily in evaluating perfor- inherently better than another. Each entails trade-
mance. Yet, as in outcome control, these salespeople offs which are explored based on analysis of the
think management is highly concerned with their responses in this study.
quantifiable results. In fact, this group thinks man- It would be helpful to review Table 1 again, show-
agement cares more about the results numbers than ing the characteristics of outcome and behavior con-
do salespeople under straight outcome control. trol. As management moves from right to left on
Our explanation for this heavy emphasis on re- the continuum (from outcome control, toward be-
sults in the hybrid group is that management may havior control), the thoughts, feelings and motiva-
perceive that they have a vested stake in the re- tion of salespeople in each system differ. Bear in
sults, given the great amount of effort put into be- mind that these are general tendencies, as there
havior monitoring. Thus, they may be overly sensi- are always specific exceptions. Discussion is re-
tive to the fruits of their efforts, expecting that stricted to those variables which showed signifi-
inputs should result in outcomes (sales). In the more cant differences, either in terms of BC1 vs. OC or in
"laissez faire" culture of outcome control offices, terms of the BC1 and BC2 combination vs. that of
managers may view performance monitoring with OC. It should also be emphasized that these find-
greater dispassion. Definitive conclusions on this ings are specific to the present study and to the
interpretation await further research. particular study measures used to represent the
In addition to the results emphasis, salespeople constructs in the A&O framework. Additional re-
in the hybrid system think their managers are search using different samples and
highly concerned that their paperwork be accurate, operationalizations will be needed to verify the con-
complete and timely. In short, management is per- clusions drawn here.

This content downloaded from 132.174.255.223 on Thu, 28 Jul 2016 02:10:11 UTC
All use subject to http://about.jstor.org/terms
Fall 1995 13

As one moves left (to behavior control), sales- There are also differences in how salespeople per-
people tend, on average, to possess higher levels of ceive their pay. Those toward the outcome control
sales expertise and competence than those on the side tend to see compensation as a lever manage-
right. They are more accepting of authority and ment uses to control them (which may foster re-
direction (a compatible trait under behavior con- sentment). As the percentage of full commission
trol), more willing to cooperate as part of a sales reps in the sample was modest (23%), it appears
team (less "lone wolf mentality), and more accept- that this form of control is exerted with the non-
ing of performance reviews (both in terms of the commission elements of the pay package (e.g., bo-
concept and the feedback from such reviews). nus). Alternatively, management may change the
Note, however, that in all of these examples, the commission structure by allocating lower commis-
pattern of results reflected the "stepped" response sions to the easier-to-sell products in the line. In
function where the hybrid group actually scored contrast, those toward the behavior control side
higher than the BC1 group. Thus, this profile may tend to see compensation as providing feedback
be more illustrative of the more intense manage- about their performance and less as a way for man-
ment demands of this group. It may also be a func- agement to control them, perhaps because salary
tion of experience; the hybrid group scored highest adjustments are typically made annually.
on average years of sales experience. This may ex- Respondents were also asked whether their firms
plain the greater level of sales competence and will- have a distinctive organization culture. In general,
ingness to cooperate with management reflected in salespeople closer to the behavior control side (BC1
these study measures. and BC2) saw their companies as having a more
Also, the more one moves to the left, the more innovative atmosphere in which participative deci-
salespeople feel committed (loyal) to their employ- sion-making (versus top-down management) is more
ers, and the more they are motivated to further the prevalent. They also viewed their firms as warm,
company's interests, frequently to the point of mak- nurturing and supportive. Interestingly, only re-
ing tradeoffs to benefit their employers at the ex- spondents reporting a hybrid system rated their
pense of their own interests and those of their cus- offices as more bureaucratic than the others. This
tomers. In contrast, salespeople on the right are would appear to be consistent with the aduaT man-
less committed to their employers and, when mak- agement style of this combination of control phi-
ing trade-offs, may tend to more often favor their losophies.
own anchor their customers' interest over that of The various differences observed here as one
their employers. This does not imply that outcome moves from outcome to behavior control are consis-
control reps are not committed to management; it tent with the A&O framework. By assuming risk
means simply that they do not feel as loyal as do (via salary), offering considerable feedback, man-
the behavior control reps. agement contact and guidance, and holding sales-
It was also observed that salespeople put a greater people more accountable for behavior than results,
emphasis on "working smarter" (devising and imple- a management builds cooperation, teamwork and
menting adaptive strategies, "diagnosing" the cus- loyalty. We note that these trends do not appear
tomer) as management systems moved toward be- solely at the behavior control extreme, but tend to
havior control. This may be because salary systems appear as one moves into the intermediate (hybrid)
give salespeople the luxury of strategizing and in- positions. It is not until the outcome control ex-
formation gathering, whereas commission systems treme is reached that they fall off markedly.
may force salespeople to move more quickly from
call to call.
Highlights of the Hybrid Group
Salespeople operating to the left (behavior con-
trol) report higher levels of job satisfaction with As noted, among the three control system groups,
their current job (the measure did not tap satisfac- there are several cases where the hybrid group
tion with the selling profession, per se). In contrast, stands out as an exception to the previously stated
outcome controlled salespeople are less satisfied generalizations. Compared to straight outcome and
with their current jobs than either BC group. This straight behavior groups, salespeople in the hybrid
pattern is similar to the commitment findings dis- group:
cussed earlier as it is known that satisfaction and
• are motivated by the intrinsic satisfaction of
commitment are strongly related (Mathieu and doing a job well;
Zajac 1990).

This content downloaded from 132.174.255.223 on Thu, 28 Jul 2016 02:10:11 UTC
All use subject to http://about.jstor.org/terms
14 Journal of Personal Selling & Sales Management

• spend a greater proportion of working time expected. Moreover, the percent of salary in these
on planning; groups does not approach the zero and 100% ex-
• do particularly well in terms of achieving tremes as some might expect.
company sale^profit goals; and Our response to these inconsistencies is that: (1)
• rate their offices as more bureaucratic. the managers themselves may use inconsistent gov-
Thus, there were substantial differences in how ernance strategies through policy and inertia, as
well hybrid salespeople reported that they pursued we have proposed, (2) the BCOC index is a first
sales and profit objectives, met these same sales attempt at such a categorization tool and is in need
and profit goals, and how well they did compared to of further study and revision (e.g., the role of paper-
other salespeople in their firms. This may, in part, work should be examined further), and (3) the study
be due to their greater efforts at planning and mak- results are based on salespersons' perceptions and
ing sales presentations. The role of intrinsic moti- may be subject to biases of interpretation. Thus,
vation in this process undoubtedly plays a part. It our findings would seem to suggest "nearly pure-
also appears that management plays an important form" groups that remain more heterogeneous than
role in bringing these outcomes about. would be predicted by theory. Researchers are en-
One likely explanation for the higher reported couraged to examine these variances for insights
into their sources.
performance in the hybrid group is the
"overmanagement" suggested by the respondents.
Hybrid management involves careful attention to Some Thoughts on Salesperson-
all details of the selling process, its planning, ex-
ecution, and results. Salespeople in these firms may
Control System Compatibility
feel overmanaged, but apparently this is likely to How do salespeople in each of the three systems
pay off in performance if the self-reported results of come to differ from each other? Perhaps it is a
this study can be generalized. combination of selection (a deliberate choice to work
An alternative explanation to the overmanage- in each environment) and adaptation (changing
ment assumption is potentially more plausible. It one's thoughts, feelings and motivation to fall into
may be that a myriad of management strategies line with the norms of the organization). Those
are used in various offices such that each salesper- who cannot adapt (e.g., risk averse salespeople who
son operates under a different set of managerial find themselves in outcome control systems) may
guidelines. O&A provide evidence of this where they switch to other systems, voluntarily or otherwise,
find low and nonsignificant correlations between or leave the selling profession. Those who stay
BCOC scores in the same sales office (p.59). The gradually adapt, though often unconsciously.
result of this "micro-management," whereby each Salespeople are bound to find themselves, on oc-
salesperson is treated individually by tailoring con- casion, in a system that doesn't suit them and to
trol systems to personal differences, is that any one which they cannot adapt. The turnover that results
sales office may have a number of control systems is healthy for both the employer and the salesper-
in place depending on the needs of its salespeople. son. Sales organizations experiencing zero turn-
Thus, such managements would respond to the over may be either extraordinarily good at hiring
motivational needs and experience of individual or too reluctant to make appropriate changes.
salespeople, resulting in a form of control system It would seem that many salespeople (especially
"segmentation." those who are inexperienced or those coming from
another industry) may not know their own prefer-
Variation in the "Pure-Form" ences. As they discover their preferences, they face
Clusters the choice of switching employers, adapting, or leav-
ing selling. In working out their choices, they may
While the BC1 and OC clusters are described experience emotional upheaval, including periods
here as pure form (for convenience), it is clear that of recrimination (self and employer). Further, it
the results raise a number of questions as to how should be noted that salespeople's preferences can
"pure" these groups are. It has already been noted change over time, perhaps provoking a job change
that certain findings contradict prescription. In par- or a change of functions with the same employer.
ticular, the BC1 cluster uses less paper and call- This latter solution is an attractive option in team
related monitoring than the others, and the OC selling environments where outcome risk is shared
cluster uses less output monitoring than would be among team members. Céspedes (1994) discusses

This content downloaded from 132.174.255.223 on Thu, 28 Jul 2016 02:10:11 UTC
All use subject to http://about.jstor.org/terms
Fall 1995 15

the elements of such systems in terms of sales task think and feel about their jobs. Perhaps the most
stability. Risk averse salespeople may wish to shift dramatic results occur when particular combina-
their efforts to those sales team activities with tions of behavior and outcome monitoring are
greater stability (account maintenance, specifica- achieved. To a lesser extent, a firm's control system
tions and terms) or to product lines with more pre- also influences how salespeople act and how well
dictable demand characteristics. Task stability they perform. Further study is required to deter-
would appear to argue for a behavior control ap- mine if other variations in control system philoso-
proach as sales processes are more orderly and po- phy exist and what the consequences of such sys-
tentially measurable. tems may be. Study should also proceed on varia-
tions of the hybrid system which optimize the sales
Conclusion impact of rep organizations generally.

It appears that the interrelated decisions about References


salesperson supervision, evaluation and compensa-
tion can be arrayed on a continuum ranging from Anderson, Erin M. (1986), "Manufacturers' Agent," in Beacham's
Marketing Reference, W. Beacham, R. T. Hise, and H. N.
behavior to outcome control. The system appears to
Tongren, eds., Washington, D.C.: Research Publishing, 484-
be somewhat non-linear, however, as behavior con- 487.
trol apparently persists well into the mid-regions Anderson, Erin and Richard L. Oliver (1987), "Perspectives on
Behavior-Based Versus Outcome-Based Salesforce Control
where it begins to take on outcome control charac-
Systems," Journal of Marketing, 51 (October), 76-88.
teristics. Contrary to the A&O predictions, the re- Bollen, Kenneth and Richard Lennox (1991), "Conventional Wis-
sponse function appears to be neither linear nor dom on Measurement: A Structural Equation Perspective,"
dichotomous. Rather, hybrid firms appear to be Psychological Bulletin, 110 (September), 305-314.
adopting both control styles, taking them on in full Céspedes, Frank V. (1992), "Sales Coordination: An Exploratory
Study," Journal of Personal Selling & Sales Management,
measure. This contrasts with a situation where
12 (Summer), 13-29.
movement in one direction or another entails a com- Cravens, David W., Thomas N. Ingram, Raymond W. LaForge,
pensatory process. and Clifford E. Young (1993), "Behavior-Based and Out-
If one were to speculate on the advisability of come-Based Salesforce Control Systems," Journal of Mar-
keting, 57 (October), 47-59.
using elements of both systems, it would only be Jaworski, Bernard J. (1988), "Toward a Theory of Marketing
necessary to examine the characteristics of the hy- Control: Environmental Context, Control Types, and Con-
brid group. This group appears both overmanaged sequences," Journal of Marketing, 52 (July), 23-39.
and micro-managed, although our interpretation Kim, Jay S. (1984), "Effect of Behavior Plus Outcome Goal Set-
ting and Feedback on Employee Satisfaction and Perfor-
leads us to conclude that the former may be more
mance," Academy of Management Journal, 27 (March), 139-
plausible. Here, management is concerned with all 149.
elements of sales generation, including the process MacCallum, Robert C. and Michael W. Browne (1993), "The Use
of Causal Indicators in Covariance Structure Models: Some
itself and the results, on an individual basis. It
Practical Issues," Psychological Bulletin, 114 (November),
appears that such managements may actually be 533-541.
over-involved in their jobs in that they may take Mathieu, John E. and Dennis M. Zajac (1990), "A Review and
much personal responsibility for the outcomes of Meta- Analysis of the Antecedents, Correlates, and Conse-
their salespeople. The data, however, suggest that quences of Organizational Commitment," Psychological
Bulletin, 108 (September), 171-194.
there is much to be said for this approach. Sales- Oliver, Richard L. and Erin Anderson (1994), "An Empirical
people in hybrid offices report the highest levels of Test of the Consequences of Behavior- and Outcome-Based
many study variables including competence, the Sales Control Systems," Journal of Marketing, 58 (Octo-
ber), 53-67.
affective variables including commitment, intrinsic
Ouchi, William G. (1979), "A Conceptual Framework for the
motivation, and many of the performance variables. Design of Organizational Control Mechanisms," Manage-
We stop short, however, of recommending this or ment Science, 25 (September), 833-847.
either of the more outcome or behavior oriented
Control: Two Functions," Administrative Science Quarterly,
systems, as numerous idiosyncratic variables will
20 (December), 559-569.
moderate the effectiveness of any one control sys- Porter, Lyman W., Edward E. Lawler, and John R. Hackman
tem. In the end, management must find its own (1975), Behavior in Organizations, New York: McGraw-
equilibrium. Hill Book Company.
Thus, moving toward one end or the other of the Terborg, James R. and Howard E. Miller (1978), "Motivation,
Behavior, and Performance: A Closer Examination of Goal
control system continuum changes the atmosphere Setting and Monetary Incentives," Journal of Applied Psy-
of the organization and influences how salespeople chology, 63 (February), 29-39.

This content downloaded from 132.174.255.223 on Thu, 28 Jul 2016 02:10:11 UTC
All use subject to http://about.jstor.org/terms

Вам также может понравиться