ArticlePDF Available

Sales Contest Effectiveness: An Examination of Sales Contest Design Preferences of Field Sales Forces

Authors:

Abstract and Figures

Sales contests, a widely used form of sales force special incentives, receive considerable attention in the trade and academic press. While understanding salespersons’ preferences for various contest designs is a critical first step for understanding how sales contests motivate salespeople to pursue contest goals, a knowledge gap exists in understanding design preferences. With expectancy theory serving as a theoretical basis, the authors develop hypotheses about preferences for sales contest components. Following tests of hypotheses using survey and conjoint data provided by field sales forces from three companies, exploratory analyses of how individual, supervisory, and sales setting characteristics may affect preferences suggest potential boundary conditions for initial findings. The results lead to an improved awareness of the determinants of contest design preferences as well as insights and implications for sales managers seeking to design effective contests.
Content may be subject to copyright.
http://jam.sagepub.com
Journal of the Academy of Marketing Science
DOI: 10.1177/0092070303261582
2004; 32; 127 Journal of the Academy of Marketing Science
William H. Murphy, Peter A. Dacin and Neil M. Ford
Forces
Sales Contest Effectiveness: An Examination of Sales Contest Design Preferences of Field Sales
http://jam.sagepub.com/cgi/content/abstract/32/2/127
The online version of this article can be found at:
Published by:
http://www.sagepublications.com
On behalf of:
Academy of Marketing Science
can be found at:Journal of the Academy of Marketing Science Additional services and information for
http://jam.sagepub.com/cgi/alerts Email Alerts:
http://jam.sagepub.com/subscriptions Subscriptions:
http://www.sagepub.com/journalsReprints.navReprints:
http://www.sagepub.com/journalsPermissions.navPermissions:
http://jam.sagepub.com/cgi/content/refs/32/2/127 Citations
at SAGE Publications on May 20, 2009 http://jam.sagepub.comDownloaded from
10.1177/0092070303261582 ARTICLEJOURNAL OF THE ACADEMY OF MARKETING SCIENCE SPRING 2004Murphy et al. / SALES CONTEST EFFECTIVENESS
Sales Contest Effectiveness:
An Examination of Sales
Contest Design Preferences
of Field Sales Forces
William H. Murphy
University of Wisconsin–Madison
Peter A. Dacin
Queen’s University, Canada
Neil M. Ford
University of Wisconsin–Madison
Sales contests, a widely used form of sales force special in-
centives, receive considerable attention in the trade and
academic press. While understanding salespersons’ pref-
erences for various contest designs is a critical first step
for understanding how sales contests motivate salespeo-
ple to pursue contest goals, a knowledge gap exists in un-
derstanding design preferences. With expectancy theory
serving as a theoretical basis, the authors develop hypoth-
eses about preferences for sales contest components. Fol-
lowing tests of hypotheses using survey and conjoint data
provided by field sales forces from three companies, ex-
ploratory analyses of how individual, supervisory, and
sales setting characteristics may affect preferences sug-
gest potential boundary conditions for initial findings. The
results lead to an improved awareness of the determinants
of contest design preferences as well as insights and
implications for sales managers seeking to design effective
contests.
Keywords: Sales contests; motivation; sales force man-
agement; incentives
We’re adding a little something to this month’s sales
contest. You all know first prize is a Cadillac Eldo-
rado. Anybody want to see second prize? Second
prize is a set of steak knives. Third prize is you’re
fired. You get the picture?
—Alec Baldwin in Glengarry Glen Ross
Glengarry Glen Ross. 1992.
New Line Cinema, USA.
Sales managers spend considerable time, energy, and
resources attempting to influence salesperson effort
through sales system components including compensa-
tion, job design, organizational climate, and special incen-
tives. Of these, special incentives have an important role,
whether as a formal means to celebrate achievement (via
recognition programs) or to reward particular accomplish-
ments (via sales contests). Special incentives, including
recognition programs, sales contests, and other special
performance incentives (Colletti et al. 1988) are widely
used across industries and companies: expenditures in the
1980s reached more than $4 billion annually (Chrapek
1989), an amount more than doubling by the late 1990s
(Nolan and Alonzo 1997).
Contests, used to motivate salespeople to attain specific
goals, receive considerable attention in the trade press
(e.g., Personnel Journal, Sales & Marketing Management,
Journal of the Academy of Marketing Science.
Volume 32, No. 2, pages 127-143.
DOI: 10.1177/0092070303261582
Copyright © 2004 by Academy of Marketing Science.
at SAGE Publications on May 20, 2009 http://jam.sagepub.comDownloaded from
Selling Power) and by researchers (see Murphy and Dacin
1998 for a recent review). Across all reports, there is a
common observation—well-designed contests motivate
salespeople. Thus, selecting an appropriate contest de-
sign is a crucial issue for sales managers and for contest
success.
Given widespread use of sales contests and a growing
literature surrounding their use, Wildt, Parker, and Harris
(1980-1981) took a “what we know and need to know”
look at the literature, summarizing the knowledge base as
primarily descriptive and lacking answers to many impor-
tant questions concerning contest design and implementa-
tion. Their call for research led to a host of studies high-
lighting sales contest design possibilities. Still, nearly 20
years later, Murphy and Dacin (1998) revisited the collec-
tive body of research and found only a limited understand-
ing of many important contest design issues, including
salespeople’s contest design preferences, a crucial factor
in motivating the pursuit of contest goals.
In this article, we take up the call and empirically ad-
dress some fundamental questions concerning sales con-
test design in a field sales force setting. For our theoretical
foundation, we use expectancy theory to develop rationale
and hypotheses for why, in general, particular designs
may be preferred by salespeople. We then use data ob-
tained from the field sales forces of three companies to test
these hypotheses. Following this, we extend our investiga-
tion to explore how different contest design preferences
relate to individual-level characteristics, supervisor-to-
salespeople relationships, and sales setting characteristics.
In essence, we investigate potential boundary conditions
for our initial findings.
Our approach contributes to current knowledge in sev-
eral ways. First, we examine why various sales contest
designs may be more or less preferred, and we ground this
discussion in theoretical and practical perspectives. Sec-
ond, we identify and test variables that may affect design
preferences. Finally, we provide insights for sales manag-
ers seeking to design effective contests.
Sales Contests: A Brief Overview
Most sales managers regularly develop special incen-
tives to motivate salespeople to pursue performance goals
outside the range of performance generated by pay and/or
compensation packages (Beltramini and Evans 1988).
Sales contests, a class of special incentives designed to
gain increased effort on short-term objectives (Churchill et
al. 2000), have long been popular (Colletti et al. 1988;
Haring and Morris 1968; Haring and Myers 1953; Nolan
and Alonzo 1997). Positive impacts of contests include
increasing motivation and morale of salespeople, leading
to greater effort, and resultant performance improvements
on targeted goals, revenue generation, and contributions to
sales and profits (Urbanski 1986; Wildt et al. 1980-1981;
Wotruba and Schoel 1983).
A sales contest represents a collection of several com
-
ponents widely discussed in the literature including goal,
competitive format, award type, contest duration, and
award value. Murphy and Dacin (1998) discussed these
contest components and suggested that a salesperson’s
perceptions about these components affect his or her atti-
tude, behavioral intent, and ensuing behaviors toward the
contest. Since the purpose of a sales contest is to increase
motivation (Beltramini and Evans 1988), understanding
how various designs affect preferences and ensuing
motivation is an important research aim.
The sales contest literature provides two detailed calls
for research (Murphy and Dacin 1998; Wildt et al. 1980-
1981), with the most recent call identifying several limita-
tions including a lack of theoretical rationale (see Hart
1984 for a notable exception
1
) with an over reliance on (1)
case studies, typically describing a specific contest in
detail but with limited generalizability; (2) input-output
analyses, necessitating the assumption that if a sales con-
test is associated with behavioral change and/or financial
gains, then salespeople must have positive attitudes
toward the design; (3) sales managers rather than sales-
people for reporting contest design preferences; and (4)
frequency of use of particular contest designs to infer pref-
erence. The authors conclude that considerable uncer-
tainty remains about salespeople’s preferences for sales
contest designs.
In this article, we address several of these issues, in-
cluding the lack of theoretical rationale. We do this by
incorporating one of the more prevalent theoretical bases
in the study of sales force motivation, expectancy theory
(Vroom 1964), as a basis for approaching our expectations
for sales contest design preferences.
Expectancy Theory
Motivating salespeople is a major task of sales manag-
ers, leading to an ongoing need to understand how sales-
people are likely to respond to organizational directives
such as sales contests. Expectancy theory, introduced to
sales management research in the 1970s (Oliver 1974), led
to a shift in research from descriptive reports of sales force
motivation schemes to using theory as a basis for identify-
ing variables and processes involved in influencing behav-
ior. Briefly, expectancy theory posits that an individual’s
motivation is determined by his or her expectancy, instru-
mentality, and valence estimates and that higher motiva-
tion results in increased effort. Expectancy represents an
individual’s estimate of the extent to which increased
effort will lead to greater performance. Instrumentality
represents an individual’s estimate of the extent to which
greater performance will lead to additional rewards.
128 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE SPRING 2004
at SAGE Publications on May 20, 2009 http://jam.sagepub.comDownloaded from
Valence is an individual’s estimate of the attractiveness
of these additional rewards. The theory posits that motiva-
tion increases as an individual’s expectancy and instru
-
mentality estimates, along with valence for rewards, are
heightened.
Expectancy theory research suggests that many vari-
ables affect expectancy, instrumentality, and valence esti-
mates. For example, variables affecting expectancy esti-
mates include the effects of current and past performance
(DeCarlo, Teas, and McElroy 1997; Johnston and Kim
1994) and task conflict (Tyagi 1982). Variables affecting
instrumentality estimates include feedback, task variety,
and task significance (Teas 1982); identification with the
organization; and perceived concern of management
(Tyagi 1982). In addition, many environmental, organiza-
tional, and/or individual-level factors including supervisor
consideration and performance feedback, tenure, and self-
esteem affect both expectancy and instrumentality esti-
mates (Evans, Margheim, and Schacter 1982; Teas 1981).
Finally, variables including pay, promotion, and perceived
sense of accomplishment affect valence estimates (Ford,
Churchill, and Walker 1985).
Valence receives extensive attention in the literature
due to management’s need to identify and provide rewards
that influence effort (Chonko, Tanner, and Weeks 1992).
Valence discussions usually involve intrinsic/extrinsic
rewards and/or needs hierarchies. Theory suggests sales-
people will typically not have high valences for intrinsic
rewards (motivated by the value of the work itself) until
they are satisfied with extrinsic rewards (motivated by
incentives offered by management) (Alderfer 1969;
Herzberg, Mauser, and Snyderman 1959; Maslow 1943).
In general, increases in extrinsic rewards (such as pay)
appear to be of particular importance (Chonko et al. 1992;
Ford et al. 1985; Oliver 1974), but there is also some evi-
dence suggesting intrinsic rewards may be just as impor-
tant (Tyagi 1985). Valences for intrinsic and extrinsic
rewards may also vary by individual characteristics
(Churchill, Ford, and Walker 1979), although some stud-
ies find the effects of these characteristics to be weak or not
predictive of valence (Chonko et al. 1992; Ingram and
Bellenger 1983).
With respect to needs hierarchies (Maslow 1970), the-
ory and research suggest satisfaction with lower order
needs (associated with more pay, higher monetary
rewards, job security, etc.) reduces valence for rewards tar-
geted at those needs and increases valence for rewards tar-
geted at higher order needs (associated with recognition,
personal growth, etc.) (Ford et al. 1985), although some
rewards associated with lower order needs (compensation)
never seem to lose attractiveness, a finding consistent
across numerous variables (Chonko et al. 1992) including
gender (Dubinsky, Jolson, Michaels, Kotable, and Lim
1993).
Overall, expectancy theory has had an enormous im
-
pact on the study of sales force motivation and effort, yet it
rarely appears as a conceptual basis for understanding
sales contest design issues. The next section addresses this
theoretical gap using expectancy theory to develop
hypotheses concerning preferences for components of
sales contest design in a field sales force setting.
Preferences for Sales
Contest Design Components
In this section, we focus on general tendency hypothe-
ses rather than individual difference hypotheses. Later, we
use individual-level analyses to identify boundary condi-
tions under which general tendencies may not hold. We do
this for two reasons. First, sales managers usually must de-
sign sales contests that appeal to a broad array of salespeo-
ple; positing hypotheses as general tendencies capture
broad preference expectations. Second, as Berlyne (1968)
stated,
It is perfectly obvious that human beings are differ-
ent from one another in some respects but alike in
other respects. The question is whether we should
first look for statements that apply to all of them or
whether we should first try to describe and explain
their differences. (P. 640)
Searching for general tendencies followed by the inclu-
sion of individual differences to refine generalizations
(Weiner 1986) recognizes that numerous individual,
supervisory, and sales setting variables could be related to
design preference and ensuing motivation (Evans et al.
1982; Teas 1981), although these differences may not nec-
essarily improve predictive ability (Chonko et al. 1992;
Ingram and Bellenger 1983).
In developing hypotheses, we focus on sales contest
components discussed by Murphy and Dacin (1998).
These components are contest goal, competitive format,
award type, contest duration, and award value. Briefly,
contest goal is the performance expectation required to
win. As we later discuss, these expectations can be out-
come, process, or combined outcome/process based.
Competitive format refers to two factors. First, contests
can have either individual or team formats. Second, the
format can allow everyone an opportunity to win or restrict
the number of winners. Award type refers to whether con-
tests offer cash, merchandise, or travel awards. Duration
relates to the length of time during which the sales contest
is held. Award value refers to the “price tag” of an award.
Although not specified in Murphy and Dacin (1998), con
-
test theme does represent an additional component. Novel
exciting themes are generally associated with greater per
-
formance on targeted goals (Wotruba and Schoel 1983),
making them a relevant issue when preparing a contest
Murphy et al. / SALES CONTEST EFFECTIVENESS 129
at SAGE Publications on May 20, 2009 http://jam.sagepub.comDownloaded from
launch. However, we exclude themes from the present
study due to their situation-specific nature. Following, we
use expectancy theory to develop our general hypotheses
for each contest component.
Contest goal can be one of three types: process based,
outcome based, and combined outcome and process
based. Process-based goals require ongoing supervisory
involvement either through observation or other mecha-
nisms (e.g., keeping tabs through weekly field reports) to
assess goal attainment. These goals emphasize improve-
ments in product knowledge, selling skills, and/or demon-
strating appropriate levels of effort on targeted activities
(e.g., product demonstrations, dealer shows, working with
distributor representatives, etc.). In retail settings where
supervisory attention to salespeople is ongoing, process-
based goals have proven effective in gaining increases
in targeted behaviors (Luthans, Paul, and Baker 1981;
Luthans, Paul, and Taylor 1985).
For field salespeople, process-based goals can be prob-
lematic—field salespeople typically have high autonomy.
The average span of control for field sales managers of 8.2
salespeople (Heide 1994) results in very few management
days spent in the field with any given salesperson. Conse-
quently, from an expectancy theory perspective, uncer-
tainty as to whether management fully and accurately
tracks process (effort) is likely to follow, reducing con-
fidence in one’s effort-to-performance assessment. Con-
sistent with Anderson and Oliver’s (1987) discourse on
behavior-based and outcome-based control systems and
related literature raising concerns about behavior-based
controls (Cravens, Ingram, LaForge, and Young 1993;
Oliver and Anderson 1995), Oliver and Weitz (1991)
asserted that this uncertainty “reduces both the expectancy
associated with sales effort and the instrumentality of sales
generated” (p. 4). Lower expectancy and instrumentality
estimates associated with this uncertainty will, in turn,
lead to lower motivation to pursue contest goals.
Outcome-based goals, often revenue or volume based
but also including improvements in indices such as cus-
tomer satisfaction scores, are frequently used goal formats
across many field sales force settings (Colletti et al. 1988;
Wildt et al. 1980-1981; Wotruba and Schoel 1983). These
goals focus sales force energies on gaining targeted results
while not requiring much supervisory involvement during
a contest. To a salesperson, reduced supervisory attention
to process (effort) allows greater behavioral leeway in
choosing the effort needed to gain targeted goals, po-
tentially enhancing expectancy estimates (Oliver and
Anderson 1995). Also, outcome-based goals are “objec-
tive” in that there is little uncertainty as to the performance
required or ensuing accuracy of measurement. As a result,
performance-to-rewards estimates are also likely to be
heightened. In effect, well-chosen outcome-based goals
should lead to higher expectancy and instrumentality
estimates and thus higher motivation to participate in a
contest.
Combined outcome/process-based goals are also in
frequent use (Colletti et al 1988). The objective of these
goals is to maintain a sales force’s focus on achieving tar-
geted results while simultaneously ensuring that good
outcome-based performance only leads to winning if tar-
geted process-based requirements are also met. While
possessing the strengths and weaknesses of each of the
above designs, in a field sales force setting, the process-
based component again can be problematic, thus lowering
expectancies and instrumentalities. Thus, from a field
sales force perspective, this design should also be inferior
to the outcome-based goal format. In sum, based on
expectancy theory, we expect the following:
Hypothesis 1: In general, field salespeople will have the
greater preference for outcome-based goals over
either process-based or combined outcome- and
process-based goals.
Competitive format: Number of potential winners. The
number of potential winners can span from one salesper-
son to all salespeople in a field sales force. The most fre-
quently used contest design allows everyone to win by
setting individual performance goals, although this is fol-
lowed closely by designs limiting the number of win-
ners (Colletti et al. 1988). With limited winners, common
practice is to have from 20 percent to 40 percent of a sales
force win (Churchill et al. 2000; Colletti et al. 1988), al-
though some contests have fewer and some have more
winners.
Based on expectancy theory, the primary effect of num-
ber of potential winners on contest design preferences will
be through its counterbalancing effects on instrumentality
and valence estimates. In terms of instrumentality, cer-
tainty about the performance-to-reward relationship will
be higher when high numbers are given the opportunity to
win, resulting in higher instrumentalities (Vroom 1964).
On the other hand, for valence, when lower numbers are
given the opportunity to win, salespeople will feel that the
higher order rewards (recognition) for winning will be
greater, resulting in higher valence estimates. Conse-
quently, it appears that higher preference is likely to occur
for designs that create a balance between these estimates.
In effect, the optimum design may need to reduce num-
bers of winners to heighten valence estimates while be-
ing sufficient to avoid overly dampening instrumentality
estimates.
In determining a “balanced” number of winners, the ex-
tent that the higher valence estimates provided by limited
numbers of winners overcomes dampened instrumentality
estimates due to limited numbers of winners is uncertain
(Ford et al. 1985). However, some reports suggest that
130 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE SPRING 2004
at SAGE Publications on May 20, 2009 http://jam.sagepub.comDownloaded from
contests should not have more than half of the participants
given the opportunity to win (Kalra and Shi 2001;
Moncrief, Hart, and Robertson 1988). Consequently,
based on our expectancy theory-related call for a balanced
number of winners, it appears that a midrange number of
winners is preferred. At midrange numbers, any dampen-
ing effects on instrumentalities may be limited (still suffi-
cient numbers to keep salespeople confident), even being
offset by increased valence for higher order rewards pres-
ent due to limited numbers, leading to greater preference
for these designs. Therefore,
Hypothesis 2: In general, field salespeople will have a
greater preference for sales contests designed with
midrange numbers of winners over those with all-
can-win or overly constrained numbers of potential
winners.
Award type. A variety of award types are available for
sales contest design including cash, merchandise, and
travel. Cash or cash-equivalent awards, used in more than
70 percent of sales contests, are the most frequently used
award (Colletti et al. 1988; Urbanski 1986), perhaps due to
the belief that cash always motivates (Ford et al. 1985). In
expectancy theory terms, preference for award type seems
most likely to be understood through the effects of valence
estimates, particularly as a satisfier of lower order needs
(through award value) and higher order needs (achieved
through factors including praise and recognition). Of the
award types, cash awards may be best suited for address-
ing lower order needs—the “real” value of the award is ap-
parent, and there is control over how to use the award. At
the same time, cash likely has little effect on higher order
need satisfaction. This suggests that other awards may be
better suited for providing both lower and higher order
needs satisfaction.
Moncrief, Hart, and Robertson (1988), in discussing
merchandise awards, suggest that giving winners some-
thing tangible to show for their winning provides better
higher order needs satisfaction (as opposed to cash, which
is often used simply to pay bills or put into savings). In
addition, Haring and Myers (1953) suggested that for sim-
ilar costs to the company, the face value of a merchandise
reward can be greater than a cash award since management
accesses merchandise at discounted prices while offering
them to salespeople at face value. From an expectancy the-
ory perspective, this suggests that merchandise, in addi-
tion to higher order need satisfaction, can also provide
lower order needs satisfaction, providing a strong basis for
motivation (Moncrief et al. 1988).
As with merchandise, travel awards can also address
both lower order and higher order needs. However, from
an expectancy theory perspective, travel may have an ad
-
vantage over merchandise. Whereas travel can satisfy
lower order needs through high face value, it can also pro
-
vide substantially more prestige and recognition for win
-
ning compared to merchandise, especially when manage-
ment accompanies winners. Consistent with this rationale,
there is descriptive evidence suggesting that salespeople
may value travel awards more than other awards
(Hastings, Kiely, and Watkins 1988) and that travel awards
may lead to the greatest gains in sales goals (Caballero
1988). Consequently, based on expectancy theory logic
and these findings, travel awards better satisfy both lower
and higher order needs and, in so doing, lead to higher va-
lence estimates (Ford et al. 1985). As result, these awards
will be more preferred by members of a field sales force
than either merchandise or cash:
Hypothesis 3: In general, for awards of comparable mon-
etary value, field salespeople will have the highest
preference for travel awards, followed by merchan-
dise, then cash.
Contest duration. Contests can run the gamut from a
few days or weeks to a year or more. Although 3 months
duration appears most common (Colletti et al. 1988), du-
ration has not been a specific focus of sales contest re-
search. Following the rationale of Murphy and Dacin
(1998), duration might be appropriately thought of as a hy-
giene factor, or dissatisfier (Herzberg et al. 1959)—while
duration cannot motivate behavior, inappropriate duration
can reduce motivation. In terms of expectancy theory,
preference for duration is most likely to result from its ef-
fects on expectancy estimates. For example, if a salesper-
son perceives too little time to plan and implement
strategies for goal attainment, his or her perception of the
effort-to-performance relationship will be dampened.
Based on this rationale, it is not surprising that anecdotal
evidence suggests increased contest performance when
contests last at least one sales cycle compared with
contests of shorter duration (Wotruba and Schoel 1983).
Still, this does not mean duration can be extended
indefinitely; from an expectancy theory perspective, dura-
tions extending to multiple sales cycles can have dampen-
ing effects on contest-related motivation. For instance,
salespeople may feel that they will be unable to sustain the
perceived effort required to pursue contest goals for
extended periods. If so, dampened perceptions of the
effort-to-performance relationship will result in lower
expectancy estimates and reduced motivation to pursue
goals (Oliver 1974). In addition to effects on expectancy
estimates, long durations may also lead salespeople to feel
that managers are using a contest to allay weaknesses in
compensation and/or territory goals rather than as incen-
tives to pursue short-term goals (Wotruba and Schoel
1983).
Thus, while assigning a specific length for an ideal du
-
ration is difficult, duration relative to a sales cycle provides
Murphy et al. / SALES CONTEST EFFECTIVENESS 131
at SAGE Publications on May 20, 2009 http://jam.sagepub.comDownloaded from
a frame of reference meaningful to any field sales setting.
In particular, based on expectancy theory, we expect sales
contests with durations of one sales cycle to be the most
preferred design, with shorter durations or longer dura-
tions relative to this cycle to be less preferred. Thus, we
expect the following:
Hypothesis 4: In general, field salespeople will have
greater preference for sales contests with durations
of approximately one sales cycle over those with
durations shorter than or considerably longer than a
sales cycle.
Award value. Determining appropriate award value is
crucial. Expectancy theory clearly suggests that only val-
ued rewards gain high valence estimates (Vroom 1964).
From an expectancy theory perspective, award value helps
satisfy lower order needs and, once satiated, a salesper-
son’s attention should be directed to fulfilling higher order
needs (Maslow 1970). However, sales research finds that
increases in pay and other pay proxies are always associ-
ated with higher valence estimates (Chonko et al. 1992;
Dubinsky et al. 1993; Ford et al. 1985), a reminder that
greater award values may always be associated with
higher valences for awards (whether cash, merchandise, or
travel). Because of its effects on valence estimates and evi-
dence that lower order needs associated with value are
never fully satiated, our next hypothesis seems obvious—
higher award values are more preferred. That is,
Hypothesis 5: In general, field salespeople will prefer
sales contests designed with greater award value.
While the above hypothesis may be obvious, in prac-
tice, a crucial management consideration becomes that of
determining sufficient but not excessive award values. For
instance, if the award value is too high, it can engender
such high valence estimates that salespeople could be-
come overly distracted from other duties or become
tempted to use inappropriate behaviors while pursuing
contest goals (Hampton 1970; Murphy 2003). Currently,
management guidance on this matter seems to be limited
to suggestions that appropriate award values be approxi-
mately 2 percent to 5 percent of annual income (LaForge,
Bolger, and Englander 1992) or no less than 1 week’s pay
(Colletti et al. 1988).
In sum, we expect field sales force preferences for out-
come-based goals (Hypothesis 1), all-can-win designs
(Hypothesis 2), travel awards (Hypothesis 3), durations of
one sales cycle (Hypothesis 4), and high award values
(Hypothesis 5). Next, we discuss our methodology. Fol-
lowing the analysis of hypotheses, we adopt Weiner’s
(1986) suggestion for refining generalizations by conduct
-
ing a variety of exploratory analyses to identify deviations
from the general findings. Specifically, we reexamine the
general hypotheses in the context of individual-level
characteristics, supervisor-to-salespeople relationships,
and sales setting characteristics as discussed in a broad
spectrum existing of motivation research.
METHOD
Salespeople from three U.S.-based companies with
field sales forces spanning consumer, industrial, and
health care sectors comprised the sample (3 companies; 46
business units). Management provided names and home
addresses of salespeople, along with letters encouraging
participation promising anonymity and stressing the
study’s value. Respondents had the opportunity to obtain a
summary of the findings; more than one third requested
these findings. A total of 1,560 sets of study materials were
mailed, followed 3 weeks later by reminder letters. Seven
hundred ninety-six usable responses were returned (51%
response rate).
Study Design, Materials, and
Procedures
We collected several specific types of information in-
cluding preferences for various contest design compo-
nents, individual-level characteristics, measures of
supervisor-to-salespeople relationships, and sales setting.
To do this, we used a full-profile conjoint task and a
survey.
Conjoint task. We chose a conjoint task (and subse-
quent analysis) to arrive at preferences for individual com-
ponents because of its ability to measure trade-offs among
multiattribute objects. We initially considered obtaining
sales contest component-related measures by exposing re-
spondents to each component in isolation. However, to
gain realism, we felt that ratings of contest design profiles
should allow for trade-offs among components. Con-
sequently, a full-profile conjoint task was used, allowing
respondents to consider preferences across a range of pos-
sible sales contest components simultaneously.
To reduce the chance of contrived component levels,
we initially selected the levels for each component based
on common practices in sales contest use (Murphy and
Dacin 1998; Wildt et al. 1980-1981). Then, managers at
each firm examined our initial profiles and suggested revi-
sions to component levels so as to ensure realistic levels
across all profiles. We then revised the component levels.
For example, with respect to number of potential winners,
managers agreed with setting 20 percent of the sales force
as the lowest level, considering anything less to be too con-
straining. Also, after learning that participating firms
never used “pure” process-based goals in a contest, we
used only outcome-based and combined outcome- and
process-based goals. One caveat of this change to contest
goals is that it results in an unbalanced design. In
132 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE SPRING 2004
at SAGE Publications on May 20, 2009 http://jam.sagepub.comDownloaded from
unbalanced designs, components with more levels are
often associated with greater importance (Hair, Tatham,
Anderson, and Black 1998). Consequently, the impor-
tance of contest goals may be attenuated. Finally, although
business groups’ sales cycles tended to differ in duration,
all managers agreed that 1 month and 3 and 6 months were
acceptable durations to include.
Two pretests followed these revisions. First, salespeo-
ple at an executive development program completed the
conjoint task and provided comments on the task and
the profiles. Next, complete packages including a
management-sponsored cover letter, conjoint task, survey,
and postage-paid return envelope were mailed to 70 sales-
people from two business units of one of the participating
firms. Thirty-one (44%) returned completed study materi-
als. Pretest feedback led to further refinements in the
wording of written instructions in the survey and conjoint
task (e.g., the sorting instructions), assessments of mea-
sure validity, and verification of the use of JC Penney as an
appropriate source for the merchandise option (Table 1
contains a description of final sales contest component
levels, while the appendix contains our measures of con-
structs indicating the source of each measure and how it
was refined).
In using the conjoint task, for practical purposes, we
assumed interaction effects among the sales contest com-
ponents to equal zero (Hair et al. 1998; Louviere 1994).
We made this choice understanding that interaction effects
may account for additional explained variance but can also
decrease the predictive power of the model (Hair et al.
1998). Also, while interaction effects may be seen as a
more accurate representation of how respondents value an
attribute, they require the use of additional profiles (Hair
et al. 1998; Louviere 1994). As we were working with
actual field sales forces, we decided to use an additive
form, intentionally trading off some explained variance
for a substantially less overwhelming task.
Based on the components in Table 1, we used SPSS
ORTHOPLAN to generate a fractional factorial design.
The generated design consisted of 16 full profiles of sales
contests. To facilitate the conjoint task, each profile
appeared on a separate card.
Murphy et al. / SALES CONTEST EFFECTIVENESS 133
TABLE 1
Factors in Conjoint Task
Variable Level Descriptor
Goals 0, 1 (0) Sales performance on the highest sales volume product in the line (percentage increase over previous matching
time period).
(1) 80 percent sales based, 20 percent manager evaluation based as follows:
(a) Sales performance on the highest sales volume product in the line (percentage increase over previous
matching time period).
(b) Management evaluation of performance factors of selling effort, presentation effectiveness, overall skills.
The evaluation is used as a multiplier for the sales performance percentage.
Multipliers are the following:
(1) considerably below expectations, 0.8;
(2) somewhat below expectations, 0.9;
(3) meets expectations, 1.0;
(4) more than meets expectations, 1.1;
(5) greatly exceeds expectations, 1.2.
Example: Salesperson who has a 13 percent increase who “greatly exceeds expectations” will receive
13 × 1.2 = 15.6% performance increase.
Competitive format 0, 1, 2 (0) Top 20 percent of the sales force at the end of the contest period.
(1) Top 40 percent of the sales force at the end of the contest period.
(2) All salespeople exceeding a 15 percent increase over previous matching time period.
Award types 0, 1, 2 (0) Winners receive a merchandise award for JC Penney’s catalog items.
(1) Winners receive a vacation award that includes expenses for travel to a resort, lodging for a five-night stay, and
meal vouchers at the resort. Award includes bringing significant other on the vacation.
(2) Winners receive a cash award.
Award value 0, 1, 2 (0) One week’s pay
(1) Two weeks’ pay
(2) Three weeks’ pay
Duration 0, 1, 2 (0) One month
(1) Three months
(2) Six months
at SAGE Publications on May 20, 2009 http://jam.sagepub.comDownloaded from
Study procedure. A cover page provided task instruc
-
tions requesting respondents to review the 16 cards con-
taining the sales contest profiles and to sort the profiles in
order of preference. After completing this task, respon-
dents numbered individual cards from 1 (the most pre-
ferred)to16(the least preferred) using each number only
once. This ensured that even if the cards were not returned
in sorted order, there would be no loss of information.
Each respondent numbered and returned all cards.
The ensuing survey included several established mea-
sures; others developed for the specific requirements of
our research; and demographics including position in the
firm, years in the position, years at the firm, education,
income, age, and gender (see appendix for these measures
and their properties).
STATISTICAL ANALYSES
As discussed, we began our analyses with the testing
of our hypotheses concerning preferences for various
sales contest components. Following this, we conducted
several exploratory analyses involving individual-level
characteristics, supervisor-to-salespeople relationships,
and sales setting.
Sample Characteristics
The characteristics of respondents included the follow-
ing: 45 to 49 years old (19%), 40 to 44 (18.1%), 35 to 39
(15.8%), 50 to 54 (14.4%), and 30 to 34 (13.8%); men
69%, women 31%; high school degree (22.3%), bache-
lor’s degree (56.2%), advanced study (21.3%); income
below $50,000 (20.6%), income between $50,000 and
$70,000 (45%), and income above $70,000 (34%). We
used contingency tables to examine differences in demo-
graphics across the three industries represented in the sam-
ple (approximately 26% consumer, 29% industrial, and
25% health care, while 20% fell into more than one indus-
try). The following differences in demographics existed
across industries. Industrial and health care salespeople
were older (68.5% and 65.1% older than 40, respectively)
than consumer goods salespeople (53% older than 40
years), χ
2
(df = 16) = 47.64, p < .05. Industrial had a smaller
percentage of women (18.5%) than either consumer or
health care (each with approximately 30%), χ
2
(df =4)=
16.69, p < .05. Industrial salespeople spent more years in
sales (median category of 11-15 years) than either con-
sumer or health care (median category of 7-10 years for
both), χ
2
(df = 12) = 67.11, p < .05. Finally, industrial and
health care salespeople had higher annual incomes
(median category of $60,000-$69,999) than consumer
salespeople (median category of $50,000-$59,999), χ
2
(df
= 14) = 38.29, p < .05. We further investigate the potential
associations of these differences to contest design
preferences in our exploratory analyses following the tests
of hypotheses.
Tests of Hypotheses
The SPSS conjoint module allowed us to decompose
the rankings of the 16 contest profiles based on the contest
components (goals, number of winners, award types,
duration, award value) and arrive at individual-level utility
scores for each level of each contest component. To test the
hypotheses, we averaged these results across respondents
(N = 796).
As conjoint analysis provides no tests of significance,
we conducted within-subject Scheffé tests as a conserva-
tive test of our hypotheses using utilities as indicators of
preference. The directionality of our hypotheses led to the
use of one-tailed tests (Hays 1994). This analysis provided
support for four of the five expectancy theory–based pref-
erence hypotheses (Table 2). For goals, higher preferences
for outcome-based (utility of .688) than combined out-
come and process based (–.688) (t = 12.66, p < .05), sup-
ported Hypothesis 1. For number of potential winners,
there was a higher preference for designs with a midrange
number of potential winners (i.e., 40%) (utility of .199)
compared with 20 percent can-win (–.272) (t = 6.51, p <
.05) or all-can-win designs (.081) (t = 3.10, p < .05). For
award type (Hypothesis 3), we hypothesized the most pre-
ferred award type to be travel, followed by merchandise,
then cash. However, cash was clearly the most preferred
award (utilities of 1.630 for cash, .307 for travel, and
–1.973 for merchandise). Thus, Hypothesis 3 was not
supported.
Clearly, the duration hypothesis (Hypothesis 4) is
dependent on the length of the sales cycle, and this differs
across industries (consumer sales typically have shorter
duration sales cycles than either health care or industrial
sales). Consequently, we tested this hypothesis across in-
dustries, finding that consumer salespeople clearly prefer
shorter duration contests than industrial salespeople, with
health care salespeople tending toward longer duration
contests as well. In particular, consumer salespeople pre-
ferred 3 months over either 1 month (utilities of .61 and
–.30, respectively) (t = 3.98, p < .05) or 6 months (.61 and
–.31, t = 4.56, p < .05. Meanwhile, industrial salespeo-
ple preferred 6 months over either 1 month (.85 and
–1.51, respectively) (t = 8.22, p < .05) or 3 months (.85 and
.65, t = 1.99, p < .05). At the same time, health care sales-
people also preferred 6 months over 1 month (.70 and
–1.35, t = 5.69, p < .05), with nearly identical utilities for 6
months and 3 months (.70 and .60), a nonsignificant differ-
ence (t = 0.43, p < .05). Overall, these findings support
Hypothesis 4. Finally, for award value, we found a higher
preference for 3 weeks’ pay (utility of 1.03) than either 2
weeks’pay (.34) (t = 11.04, p < .05) or 1 week’s pay (–1.38)
(t = 23.65, p < .05). This finding supports Hypothesis 5.
134 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE SPRING 2004
at SAGE Publications on May 20, 2009 http://jam.sagepub.comDownloaded from
In sum, tests of hypotheses revealed preferences for
outcome-based goals, limiting numbers of winners to 40
percent of the sales force, 3 months duration (with excep-
tions as noted by industry), and with cash awards at high
value levels (3 weeks pay). To provide additional evi-
dence as to how widespread sales force preferences were
for this “preferred design, we examined each respon-
dent’s utilities to identify the percentage whose utilities
matched this combination. Approximately 18 percent of
respondents’ utilities indicated this would be their most
preferred combination. In addition, about 7 percent of re-
spondents’ utilities indicated a sales contest design prefer-
ence deviating from this design only in terms of duration
(6 months rather than 3 months). Thus, these two design
combinations accounted for approximately 25 percent of
the respondents.
Exploring Individual, Supervisory,
and Sales Setting Characteristics
Following hypothesis testing, we explored whether
component preferences differed across industries and by
individual, supervisory, and sales setting characteristics—
variables found in existing motivation research. Being
exploratory, we used two-tailed tests of significance for
these analyses.
By industry. We investigated whether design prefer-
ences differed on an industry basis (for this analysis, we
excluded respondents falling into multiple industries). Be-
yond the preferences for duration, conjoint analysis by in-
dustry revealed a near mirroring in terms of preferences
for outcome-based goals, cash awards followed by travel,
and 3 weeks’followed by 2 weeks’value. In terms of num-
ber of winners, consumer and health care each preferred
40 percent, followed by all-can-win, while industrial had a
slight preference for all-can-win, followed closely by 40
percent. For all three, 20 percent can-win was the least
attractive (Table 2).
Continuing with industry, we examined whether
responses to measures (appendix) differed across indus-
tries. The results indicated that only relationship to super-
visor differed between industries—health care salespeo-
ple (5.83) were closer to their supervisors (p < .05) than
consumer salespeople (5.42), while industrial (5.74) was
not significantly different from health care or consumer.
By individual, supervisory and sales setting character-
istics. We also explored how design preferences varied
due to individual, supervisory, and sales setting variables
on a component-by-component basis. To do this, we in-
vestigated utility scores for each level of each component.
For example, for the goals component, we used each re-
spondent’s utility scores to determine preferences for out-
come or combined outcome- and process-based goals. For
goals, 65.3 percent preferred outcome and 34.7 percent
preferred combined outcome- and process-based goals.
Using the explanatory variables (appendix), we ran a se-
ries of Scheffé tests and Pearson chi-square tests (for cate-
gorical measures) to determine which variables were
associated with preferences toward one or the other goal.
Then, we repeated the analysis for each component. Only
significant results from these analyses (p < .05) appear in
Table 3.
We also explored the relationship between component
importance (from the conjoint analysis) and the various
individual, supervisory, and sales setting variables. In par-
ticular, in terms of the importance of individual design
components in determining contest attitudes, significant
relationships (p < .05 using two-tailed tests) were found
for several variables. First, with respect to the goal-type
component, we found that as satisfaction with supervisor,
relationship with supervisor, supervisor effectiveness, and
the frequency with which the supervisor works with the
salesperson increased, the importance of goal type
decreased. Furthermore, these supervisory relationship
variables were unrelated importance variations for any
other component. With respect to award value importance,
Murphy et al. / SALES CONTEST EFFECTIVENESS 135
TABLE 2
Conjoint Analysis:
Aggregate and Industry-Based Solutions
Industry Based
Aggregate Consumer Industrial Health
(n = 796) (n = 211) (n = 224) (n = 200)
Goal
Outcome 0.688 0.79 0.61 0.65
Combined –0.688 –0.79 –0.61 –0.65
Importance 0.130 0.18 0.12 0.13
Number of winners
All can win 0.081 –0.01 0.15 –0.04
40 percent can win 0.199 0.18 0.13 0.21
20 percent can win –0.272 –0.17 –0.29 –0.17
Importance 0.138 0.04 0.04 0.04
Award type
Merchandise –1.937 –1.85 –1.95 –1.84
Travel 0.307 0.25 0.23 0.17
Cash 1.630 1.61 1.73 1.67
Importance 0.315 0.41 0.36 0.36
Award value
One week’s pay –1.38 –1.27 –1.44 –1.43
Two weeks’ pay 0.34 0.30 0.39 0.36
Three weeks’ pay 1.03 0.97 1.04 1.06
Importance 0.181 0.26 0.24 0.26
Duration
One month –0.899 –0.30 –1.51 –1.35
Three months 0.670 0.61 0.66 0.70
Six months 0.229 –0.31 0.85 0.65
Importance 0.251 0.11 0.23 0.21
NOTE: Respondents reporting more than one industry are not included in
the table. We conducted an ANOVA for each of the average utilities
across industries and found no significant differences (in all cases, p <
.05) except for duration as reported in the text. The difference for duration
is a result of different length sales cycles among the industries. Results for
respondents reporting more than one industry were not significantly dif-
ferent from the aggregate results.
at SAGE Publications on May 20, 2009 http://jam.sagepub.comDownloaded from
we found that as status aspiration-based need for achieve
-
ment, competitiveness-based need for achievement
advancement desire, perceived control over own perfor
-
mance, and relative perceived performance increased, so
136 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE SPRING 2004
TABLE 3
Component-by-Component Exploratory Analyses
Goal Outcome Based Combined
Percentage preferring 65.3 34.7
Individual, supervisory, and sales setting characteristics
Satisfaction with supervisor 5.34
a
5.78
b
Relationship to supervisor 5.53
a
5.90
b
Supervisor effectiveness 4.36
a
4.90
b
Advancement desire 3.80
a
4.24
b
Supervisor time with rep (categorical) < 1 day a month > 2 days a month
Gender (categorical) 70 percent of men 55 percent of women
Competitive Format 20 Percent Can Win 40 Percent Can Win All Can Win
Percentage preferring 34.4 38.9 26.7
Individual, supervisory, and sales setting characteristics
Advancement desire 4.12
a
4.03 3.76
b
Commitment 6.22
a
6.12 6.09
b
Competitive-based need for achievement 5.21
a
5.03 5.01
b
Award Type Cash Travel Merchandise
Percentage preferring 57.8 36.3 5.9
Individual, supervisory, and sales setting characteristics
Satisfaction with compensation 4.85
a
5.13
b
5.17
Satisfaction with supervisor 5.50 5.61
a
5.10
b
Supervisor effectiveness 4.96 4.82
a
4.68
b
Advancement desire 3.91 4.10
a
3.44
b
Income (categorical) < $30,000, $60,000-$69,000, $60,000-$69,000, $70,000- > $80,000
and $70,000-$79,000 $79,000, and > $80,000
Award Value One Week’s Pay Two Weeks’ Pay Three Weeks’Pay
Percentage preferring 7.5 8.2 84.3
Individual, supervisory, and sales setting characteristics
Competitive-based need for achievement 4.79
a
5.20
b
5.07
Supervisor time with rep (categorical) 1 day a week Up to 2-3 days a month Less than 2-3 days
a month
Income (categorical) < $50,000 < $50,000 > $50,000
Duration 1 month 3 months 6 months
Percentage preferring 26.8 31.8 41.3
Individual, supervisory, and sales setting characteristics
Percentage pay at risk 57.9
a
68.6
b
75.0
c
Advancement desire 4.37
a
3.72
b
3.86
b
Status-based need for achievement 5.65
a
5.48
b
5.54
Supervisor time with rep (categorical) Up to 1 day a month Up to 7-11 days Less than 11 days
a year a year
Years in sales (categorical) < 3 years 3-11 years > 7 years
Industry type (categorical) Consumer No dominant industry Health care and industrial
NOTE: Preferences for component levels were determined by individual-level conjoint-derived utilities. Both Pearson chi-square tests (for categorical
variables noted in the table) and Scheffé tests (for continuous variables) were used. Only significant differences (p < .05) are reported (n = 796). For categor-
ical variables, categories reflecting the primary differences in trends found in the contingency tables are reported. For percentage scores, we first trans
-
formed the data using an arcsin transformation. Raw percentages are reported for interpretability. In each row, different letters represent scores that are
significantly different (p < .05) based on Scheffé tests. For example, for Advancement desire, under Duration, those who preferred 1 month had signifi-
cantly different Advancement desire scores than those who preferred 3 months and 6 months. Advancement desire scores for those who preferred 3 months
were not significantly different from those who preferred 6 months.
a. Those who preferred 1 month.
b. Those who preferred 3 and 6 months.
at SAGE Publications on May 20, 2009 http://jam.sagepub.comDownloaded from
did importance of this component. In addition, importance
of the award value component decreased as age and years
in sales increased. Finally, with respect to duration impor
-
tance, we found that it increased as the percentage of pay at
risk decreased.
DISCUSSION
The results associated with our hypotheses suggest sev-
eral field sales force contest design preferences that are
consistent with expectancy theory–based predictions.
These preferences are for outcome-based goals, a limited
number of winners (40%), durations of approximately a
sales cycle (as appropriate for various industries), and
higher award values (3 weeks’ pay). The only result our
expectancy theory–based hypotheses did not anticipate
was the preference for cash awards.
All told, the emergent preferred design represented the
preferred combination of component levels for nearly one
fifth of the sample and, with duration allowed to be either
3 or 6 months, the identified preference design fully
accounts for approximately one fourth of the respondents
preferences based on conjoint derived utilities. While rep-
resenting a preferred design for substantial numbers of
field salespeople, these tendencies do not provide com-
plete guidance to sales contest planners seeking to design
effective sales contests. However, our combined general
results and our component-by-component exploratory
analyses using individual, supervisory, and sales setting
characteristics do provide deeper insights into potential
boundary conditions in contest design considerations.
Although, on average, goal type is the least important
component determining preferences (not unexpected
given our previous caveat about unbalanced conjoint
designs), both its preference and importance seem related
to a salesperson’s relationship to supervisor. While its
importance decreases with closer relationships, stronger
supervisor-to-salesperson relationships are associated
with a higher preference for combined outcome- and
process-based designs. This is consistent with our expec-
tancy theory perspective. As we argued earlier, if sales-
people feel distant from their supervisors, they are likely
to lack certainty as to whether management fully and ac-
curately tracks process, which in turn could reduce con-
fidence in their own effort-to-performance assessments
(dampening expectancy and instrumentality estimates).
Furthermore, the use of this goal type would magnify the
effects of this component in determining contest attitudes,
hindering the ability of contest planners to gain positive
attitude and ensuing motivation to pursue contest goals.
Management attuned to these relationship-to-supervisor
issues can better choose appropriate instances to use com
-
bined outcome- and process-based goal formats, with the
accompanying advantage of being able to make certain
that winning can be recognized as being due to both meet
-
ing outcome targets and good effort (reducing wins by
windfall or unacceptable behaviors; Murphy 2003). On
the flip side, the addition of a process-based goal, even if
representing just 20 percent of the overall goal, as is the
case for our combined outcome- and process-based goal
design, can severely affect expectancies and compromise
motivation toward a contest for salespeople with more dis-
tant supervisory relationships.
Combined outcome- and process-based goal designs
also seem more acceptable to women (45% of women pre-
ferred combined designs compared with just 30% of men)
and to salespeople with more frequent days spent working
with supervisors. As to the latter, this attests to the benefits
of ongoing observation, which is likely to increase sales-
people’s confidence in whether management is tracking
process, thus leading to increased expectancy and instru-
mentality estimates when process-based goals are used.
As to the former, we considered the possibility that the
relationship-to-supervisor variable could be driving this
difference. However, there are no significant differences
by gender on the supervisor relationship variables, thus no
support for this conjecture. Still, the results must be inter-
preted with care; similar to earlier gender-related findings
(Dubinsky et al. 1993; Schul, Remington, and Berl 1990),
more gender similarities than dissimilarities exist across
the contest design components here.
The number of potential winners component has fairly
low importance scores, with 34 percent preferring all-can-
win designs, 39 percent preferring 40 percent can-win
designs, and 27 percent preferring 20 percent can-win
designs. This does not vary by individual, supervisory, and
sales setting characteristics. With respect to preference,
both more competitive and higher advancement-desiring
salespeople seem to prefer limited numbers of winners. As
argued previously, this may be due to the dampening
effects on expectancies provoked by limited numbers of
winners not affecting these salespeople as much as their
peers and the counterbalancing valence effects gained by
being among a select group of winners. That is, salespeo-
ple with either of these characteristics might view limited
numbers as a positive challenge. Also, these salespeople
may have especially high valence estimates for the higher
order needs satisfaction (recognition) gained by winning;
with small numbers of winners, these estimates would be
enhanced.
The award type component is of high importance for
most salespeople and, although our expectancy theory-
based rationale and prior research (Caballero 1988;
Hastings et al. 1988) led us to expect a preference for travel
awards over cash or merchandise, cash awards are gen
-
erally the most preferred. Even so, preferred award type
varies by at least some individual, supervisory, and sales
setting characteristics. For instance, and not unexpectedly,
salespeople with less close relationships to supervisors
Murphy et al. / SALES CONTEST EFFECTIVENESS 137
at SAGE Publications on May 20, 2009 http://jam.sagepub.comDownloaded from
find travel awards accompanied by supervisory staff to
be a far distant third in award preference; these sales-
people strongly prefer cash. At the same time, closer-to-
supervisor salespeople, along with those with higher lev-
els of satisfaction with supervisors and whose supervisors
are seen as more effective, seem quite comfortable with
having travel awards accompanied by supervisory staff as
an alternative to cash. Thus, consistent with earlier re-
search suggesting that manager-salesperson relationships
affect valence estimates (Legace, Castleberry, and Ridnour
1993), supervisory factors clearly affect reward prefer-
ences. In a positive light, this means that travel awards can
be effectively used when an audit of supervisor-to-sales
force relationships suggests “good health. However, if in
doubt about the quality of these relationships, or if these
relationships vary noticeably, travel awards in which
supervisors accompany winners should be avoided—wide
variation in resultant motivation seems likely to follow.
Additional observations concerning award type sug-
gest that lower (below $30,000) as well as mid- to upper
income brackets ($60,000-$80,000) have greater prefer-
ences for cash awards. The preference for cash seems clear
for lower incomes—these salespeople would be far from
satiated in terms of satisfying their lower order needs (i.e.,
need for more pay). Thus, cash would be more likely to
gain high valence estimates. At the same time, middle/
upper income salespeople might have a higher valence for
cash (suggesting a renewed drive to satisfy lower order
needs occurring at these income levels) due to financial
demands associated with midcareer/midlife stages. Un-
fortunately, we did not collect data on financial demands.
While it is not surprising that increases in award value
are nearly always preferred, award value is seldom of high
importance relative to other components. This is an inter-
esting finding in that salespeople do not seem merely
focused on chasing sales contest goals solely to address
lower order needs. Even so, our exploratory analysis does
indicate that award value importance varies by individual
characteristics. First, value seems more important to
highly competitive salespeople. This could be due to
highly competitive salespeople having a heightened focus
on the extrinsic rewards gained by winning—they may be
“chasing the carrot” being extended by management
(Kohn 1993), and only a large enough carrot will excite
them to the chase. Second, consistent with the findings of
Cron (1984), younger salespeople, as well as salespeople
in earlier career stages, place greater importance on value.
This may be due to having less financial security, for
example, lower order needs remain unsatisfied, thereby
giving value heightened importance. From an expectancy
theory perspective, contest designers need to understand
that “undershooting value” in contest design could cause
considerable problems in gaining motivation with sales
forces composed of these types of salespeople.
As predicted by our expectancy theory-based rationale,
duration seems to require a “sufficient but not excessive”
quality; sufficient for planning and executing strategies for
contest goal attainment but short enough that difficulties in
sustaining a positive effort-to-performance relationship
are not an issue. Based on this, we predicted a preference
for duration of approximately one sales cycle. Using
industry analysis as a proxy for varying sales cycles, with
consumer salespeople having the shortest sales cycle and
industrial the longest, duration preferences clearly vary
between industries. Consistent with their shorter sales cy-
cles, consumer salespeople have a clear preference for 3
months’ duration, with substantial numbers also showing
a preference for 1 month. Meanwhile, consistent with their
longer sales cycles, both industrial and health care sales-
people clearly dislike 1-month durations, while industrial
salespeople even have a slight bias toward 6 months.
Other factors associated with duration preferences
include years in sales, frequency supervisors work with
salespeople, pay at risk, advancement desire, and status-
based need for achievement. We find that salespeople with
more years of experience are more likely to prefer longer
duration contests. In part, this might be due to experienced
salespeople having “been there before, giving them
increased patience in letting a contest run for longer peri-
ods. We also find that salespeople possessing higher
advancement desire and/or a higher status-based need for
achievement are each more likely to prefer shorter dura-
tion contests. This could be due to having higher valences
for the higher order awards associated with winning (rec-
ognition), thereby increasing a sense of urgency to bring
contests to conclusion. Finally, we find that salespeople
working in sales settings with higher base salaries (lower
pay at risk) seem to key in on duration as a vital component
affecting their attitudes, with longer duration contests
more preferred. One reason for this may be the types of
sales settings where high base salary is typically found. In
field sales force settings where territory objectives include
factors other than simply short-term sales volume and
when a salesperson’s impact on volume is hard to measure,
higher base salaries are often the norm (Churchill et al.
2000). Thus, salespeople with higher base salaries are
often more accustomed to longer time horizons. Mean-
while, higher pay-at-risk salespeople are nearly ambi-
valent as to whether a contest is 3 months or 1 month,
perhaps reflecting their focus on short-term performance
goals as a normal part of day-to-day selling efforts. Since
pay characteristics (high base salary versus high pay at
risk) are a consequence of the sales setting, contest de-
signers would be well served to make contest duration
reflect the pay/setting dynamic. Suggested earlier, dura
-
tion may only have adverse consequences on expectancy
estimates when misaligned (too short or excessive for goal
attainment).
138 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE SPRING 2004
at SAGE Publications on May 20, 2009 http://jam.sagepub.comDownloaded from
ADDITIONAL MANAGERIAL IMPLICATIONS
Our findings and discussion of results provide manage
-
ment with two paths for designing sales contests. On one
hand, management can fall back on our general findings,
constructing sales contests that reflect preferences of the
average salesperson. For example, management can use
the general findings to justify outcome-based goals in con-
test design. Outcome-based goals reduce the need for
monitoring mechanisms, thereby lessening contest admin-
istration activities. As such, these goals have advantages
not only in that salespeople tend to prefer them but also for
ease of administration. Similarly, management can always
use cash awards since these awards tend to be preferred
over travel or merchandise. Whether due to ease of admin-
istration or an awareness of the effects of these goals on
expectancies and instrumentalities, the literature consis-
tently reports a management preference for outcome-
based designs (Haring and Myers 1953; Haring and
Morris 1968; Wotruba and Schoel 1983).
On the other hand, management can “dig deeper” and
be driven by the understanding that improved contest
design is possible by attending to individual, supervisory,
and sales setting factors (i.e., factors related to motivation)
and adapting designs and communications that best reflect
the needs and circumstances of a given sales force. Many
of the variables differentiating design preferences appear
to have strong effects. For example, supervisory factors
clearly affect design preferences for goals and award type,
suggesting the need to consider the nature of supervisor-
salesperson relationships. Are they consistently “healthy”
and do salespeople feel their supervisors treat them with
respect, are eager to reward good performance, and are
active in training and coaching? If so, management might
consider taking advantage of the managerially positive
effects of including a process-based goal component.
As another example, contest planners contemplating
the number of salespeople eligible to win need to realize
that constraining the potential number of winners too
severely will likely harm preferences and motivation to-
ward sales contests. At some point, the dampening effects
on expectancy and instrumentality estimates likely be-
comes too great to be offset by increased valences for
higher order rewards provided by designs with limited
numbers of winners. Here again, best practices are
affected by both individual and supervisory differences.
For instance, if supervisors are actively involved in train-
ing and coaching and are eager to reward good perfor-
mance, salespeople seem to prefer limits placed on the
number of potential winners. In effect, this can be due to
the fact that these salespeople come to develop greater
confidence in their abilities while also feeling that a
design with limited numbers of winners has been designed
fairly.
LIMITATIONS AND DIRECTIONS
FOR FUTURE RESEARCH
A number of limitations introduce future research op-
portunities. The first concerns the nature of the sample.
Although the sample included a diverse group of business
units spanning consumer, industrial, and health care, only
three companies participated, potentially limiting the
generalizability of the findings. There may also be some
concern for self-selection bias; salespeople with a greater
vested interest may have responded in greater numbers. In
addition, our focus was on field sales forces only. Today’s
highly dynamic markets have sales conducted by a wide
array of salespeople (e.g., telesales; sales engineers; etc.)
in numerous settings (e.g., business-to-business; business-
to-consumer); at present, our findings cannot be extended
to these contexts.
In terms of the methodological issues, future research
should include a balanced design as a means to clarify
ideal attribute levels. Although possibly very cumber-
some, in future research, the ability to detect interaction
effects is also important. In this study, respondents were
asked to focus on a high sales volume part of their line as
the contest goal. Future studies should also include other
offerings including newly launched, slow-moving, and/
or discontinued products. As noted, our research did
not examine competitive format in terms of team versus
individual-based designs. With some growing emphasis
on teams in many sales settings, this would be an impor-
tant component to study in future research. With respect to
the award-type component, we chose our specific levels
based on pretests involving both management and sales-
people from the organizations we surveyed. While appro-
priate for this study, a broader approach to this component
would be beneficial for a deeper understanding.
Finally, although areas of known potential impact
include effects on motivation and morale of salespeople
(Wildt et al. 1980-1981), our study only alludes to the
effects of contests on relevant constructs including com-
mitment and satisfaction, as well as the possibility of con-
ditioning salespeople to “chase carrots” (Kohn 1993); the
effects of sales contests on these outcomes must be
explored if the full impact of sales contests are to be under-
stood. Along these lines, the risks/concerns of using
extrinsic motivators such as sales contests must also be
given attention. Although unintended consequences have
been addressed (Hampton 1970; Murphy 2004), research
is needed in terms of the kinds of cheating behaviors that
might occur and the ethical or legal implications of these
behaviors.
In sum, the importance of sales contests is clear. Fur
-
thermore, the potential of expectancy theory as a legiti
-
mate basis for deriving an understanding of sales con
-
test design preferences and motivation appears very
Murphy et al. / SALES CONTEST EFFECTIVENESS 139
at SAGE Publications on May 20, 2009 http://jam.sagepub.comDownloaded from
promising. What remains is for researchers to further
advance understanding of this widely used incentive tool
and to make the direct connection between preferences
motivation and effort in this context. Until then, descrip-
tive reports provide most of the available guidance,
although Alec Baldwin has added to the discourse, leaving
little doubt about how sales contests can even be used to
bludgeon salespeople toward greater motivation. You get
the picture?
NOTES
1. Hart (1984) grounded her sales contest study in goal theory (Locke
1968). Goal theory proposes that effort expended by a goal-focused indi-
vidual is greater than effort expended by an individual without goals and
that goal characteristics affect performance outcomes (see Locke and
Latham 1990 for a review). In sum, researchers suggest that acceptable,
clear, specific, and difficult goals lead to higher performance by increas-
ing goal-related effort (Hollenbeck and Klein 1987; Latham and Locke
1979). Hart found support for most of these expectations.
2. All award types (cash, merchandise, travel) likely provide similar
baseline levels of higher order need satisfaction through the public an-
nouncement of winners. These announcements often occur at meetings/
banquets where every member of the sales force, along with manage-
ment, attends and the winners are given accolades. The focus of this dis-
cussion concerns the above baseline effects of particular award types.
REFERENCES
Alderfer, C. P. 1969. An Empirical Test of a New Theory of Human
Needs. Organizational Behavior and Human Performance 4:142-
175.
Anderson, E. and R. L. Oliver. 1987. “Perspectives on Behavior-Based
Versus Outcome-Based Salesforce Control Systems. Journal of
Marketing 51 (4): 76-88.
Beltramini, R. F. and K. R. Evans. 1988. “Salesperson Motivation to Per-
form and Job Satisfaction: A Sales Contest Participant Perspective.
Journal of Personal Selling & Sales Management 8 (2): 35-42.
Berlyne, D. E. 1968. “Behavior Theory as Personality Theory. In Hand-
book of Personality Theory and Research. Eds. E. F. Borgetta and
W. W. Lambert. Chicago: Rand McNally, 629-690.
Caballero, M. J. 1988. A Comparative Study of Incentives in a Sales
Force Contest. Journal of Personal Selling & Sales Management
8 (1): 55-58.
Cassidy, L. and R. Lynn. 1989. “A Multifactorial Approach to Achieve-
ment Motivation: The Development of a Comprehensive Measure.
Journal of Occupational Psychology 62:301-312.
Chonko, L. B., J. F. Tanner, and W. A. Weeks. 1992. “Selling and Sales
Management in Action: Reward Preferences of Salespeople.Jour-
nal of Personal Selling & Sales Management 12 (3): 67-75.
Chrapek, D. 1989. “Sales Incentive Expenditures Top $7.8 Billion Mark
in 1988.Premium Incentive Business, June, p. 3.
Churchill, G. A., N. M. Ford, and O. C. Walker Jr. 1979. “Personal Char-
acteristics of Salespeople and the Attractiveness of Alternative
Rewards.Journal of Business Research 7:25-50.
, , , M. W. Johnston, and J. F. Tanner Jr. 2000. Sales
Force Management: Planning, Implementation, and Control. 6th ed.
Homewood, IL: Irwin.
Colletti, J. A., D. J. Cichelli, S. D. Linser, J. F. Martin, D. S. Schattinger,
and G. S. Tubridy. 1988. Current Practices in Sales Incentives.New
York: The Alexander Group, Inc.
Cravens, D. W., T. N. Ingram, R. W. LaForge, and C. E. Young. 1993.
“Behavior-Based and Outcome-Based Salesforce Control Systems.
Journal of Marketing 57 (4): 47-59.
Cron, W. L. 1984. “Industrial Salespeople Development:A Career Stages
Perspective.Journal of Marketing 48 (4): 41-52.
DeCarlo, T. E., R. K. Teas, and J. C. McElroy. 1997. “Salesperson Per
-
formance Attribution Processes and the Formation of Expectancy
Estimates. Journal of Personal Selling and Sales Management 17
(3): 1-17.
Dubinsky, A. J., M. A. Jolson, R. E. Michaels, M. Kotabe, and C. U. Lim.
1993. “Perceptions of Motivational Components: Salesmen and
Saleswomen Revisited. Journal of Personal Selling & Sales Man-
agement 8 (4): 25-37.
Evans, K. R., L. Margheim, and J. L. Schacter. 1982. A Review of
Expectancy Theory Research in Selling. Journal of Personal Selling
& Sales Management 2 (2): 33-40.
Ford, N. M., G. A. Churchill Jr., and O. C. Walker Jr. 1985. “Differences
in the Attractiveness of Alternative Rewards Among Industrial Sales-
people: Additional Evidence. Journal of Business Research 13
(April): 123-138.
Hair, J., R. Tatham, R. Anderson, and W. Black. 1998. Multivariate Data
Analysis. 5th ed. Englewood Cliffs, NJ: Prentice Hall.
Hampton, D. R. 1970. “Contests Have Side Effects Too. California
Management Review 12 (4): 86-94.
Haring, A. and M. R. Morris. 1968. Contests, Prizes, Awards for
Sales Motivation. New York: Sales and Marketing Executives
International.
 and R. H. Myers. 1953. “Special Incentives for Salesmen. Jour-
nal of Marketing 18 (October): 155-159.
Hart, S. H. 1984. An Empirical Investigation of Salespeople’s Behavior,
Effort and Performance During Sales Contests. Unpublished Doc-
toral Dissertation. Texas A & M University, College Station.
Hastings, B., J. Kiely, and T. Watkins. 1988. “Sales Force Motivation
Using Travel Incentives: Some Empirical Evidence. Journal of Per-
sonal Selling & Sales Management 8 (2): 43-51.
Hays, W. L. 1994. Statistics. 5th ed. Fort Worth, TX: Holt, Reinhart &
Winston.
Heide, C. P. 1994. Dartnell’s 28th Sales Force Compensation Survey,
1994-95. Chicago: The Dartnell Corporation.
Herzberg, F., B. Mauser, and B. Snyderman. 1959. The Motivation to
Work. 2nd ed. New York: John Wiley.
Hollenbeck, J. R. and H. J. Klein. 1987. “Goal Commitment and the Goal
Setting Process: Problems, Prospects, and Proposals for Future
Research.Journal of Applied Psychology 72:212-220.
Ingram, T. N. and D. N. Bellenger. 1983. “Personal and Organizational
Variables: Their Relative Effect on Reward Valences of Industrial
Salespeople.Journal of Marketing Research 20 (May): 198-205.
Jaworski, B. J. and A. K. Kohli. 1993. “Market Orientation: Antecedents
and Consequences.Journal of Marketing 57 (July): 53-70.
Johnston, W. J. and K. Kim. 1994. “Performance, Attribution, and Ex-
pectancy Linkages in Personal Selling. Journal of Marketing 58 (4):
68-81.
Kalra, A. and M. Shi. 2001. “Designing Optimal Sales Contests: A Theo-
retical Perspective.Marketing Science 20 (2): 170-193.
Kohn, A. 1993. “Why Incentive Plans Cannot Work.Harvard Business
Review 5 (September-October): 54-63.
LaForge, A., B. Bolger, and T. Englander. 1992. “Budgeting and Evaluat-
ing Awards and Incentives.Personnel Journal, June, pp. 131-135.
Latham, G. P. and E. A. Locke. 1979. A Theory of Goal Setting and Task
Performance. Englewood Cliffs, NJ: Prentice Hall.
Legace, R. R., S. B. Castleberry, and R. E. Ridnour. 1993. “An Explor-
atory Salesforce Study of the Relationship Between Leader-Member
Exchange and Motivation, Role Stress, and Manager Evaluation.
Journal of Applied Business Research 9 (4): 110-119.
Locke, E. A. 1968. “Toward a Theory of Task Motivation and Incentives.
Organizational Behavior and Human Performance 3 (May): 157-189.
 and G. P. Latham. 1990. A Theory of Goal Setting and Perfor
-
mance. Englewood Cliffs, NJ: Prentice Hall.
140 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE SPRING 2004
at SAGE Publications on May 20, 2009 http://jam.sagepub.comDownloaded from
Murphy et al. / SALES CONTEST EFFECTIVENESS 141
APPENDIX
Individual, Supervisory, and Sales Setting Variables Included in the Survey
Reliability
Measure Scale Items (Coefficient α)
Affective Organizational Commitment seven-item, 7-point
scale adapted from Jaworski and Kohli (1993) reduced to
six items following pretest
1. I am fond of this organization.
2. I am happy to make personal sacrifices for this organization if it
is important for the organization’s well-being.
3. The bonds between this organization and me are weak. (R)
4. In general, I am proud to work for this organization.
5. I often go above and beyond the call of duty to ensure this
organization’s success.
6. I have little commitment to this organization. (R)
.82
Relationship to Supervisor eight-item, 7-point Likert-type
scale adapted from Tyagi (1985) reduced to six items
following pretest
1. My supervisor is eager to recognize and reward my performance
when it is good.
2. My supervisor treats me with respect.
3. I find my supervisor friendly and easy to approach.
4. My supervisor is usually attentive to what I say.
5. I usually trust statements made by my supervisor.
6. My supervisor is usually willing to listen to my problems.
.83
Status Aspiration-Based Need for Achievement seven-
item, 7-point Likert-type scale adapted from Cassidy and
Lynn (1989), reduced to four items following pretest
1. I want to have a position in the firm where I can have prestige.
2. I like to be admired for my achievements.
3. I want to be an important person at this firm.
4. I find satisfaction when I can influence others in this firm.
.78
Competitiveness-Based Need for Achievement seven-
item, 7-point Likert-type scale adapted from Cassidy and
Lynn (1989), reduced to four items following pretest
1. I try harder when I am in competition with other people.
2. To be a real success, I have to do better than other salespeople.
3. It is important for me to do better than others in the sales force.
4. I judge my performance on whether I do better than others.
.78
Supervisor Effectiveness four-item, 7-point Likert-type
scale developed for this study
1. My supervisor provides me with insights to improve my selling
efforts.
2. My supervisor demonstrates active training and coaching for me.
3. My supervisor holds effective meetings (I leave knowing how to
do job better).
4. My supervisor handles the job such that I consider him or her
effective
.93
Advancement Desire single-item, 7-point Likert-type
scale
I view sales as primarily a stepping stone to other positions.
Control Over Own Performance single-item, 7-point
Likert-type scale
For the most part, I control my performance in my territory.
Satisfaction with Supervisor and Compensation single-
item, 7-point scale anchored by strongly satisfied and
strongly dissatisfied
Please indicate the extent of satisfaction with the following areas
of your job
Your compensation . . . Your supervisor . . .
Supervisor Span of Control categorical response How many salespeople is your sales manager responsible for?
Less than 5
5to6
7to8
9to10
11 to 12
13 or more
(continued)
at SAGE Publications on May 20, 2009 http://jam.sagepub.comDownloaded from
Louviere, J. J. 1994. “Conjoint Analysis.” In Advanced Methods of Mar-
keting Research. Ed. R. P. Bagozzi. Cambridge, MA: Blackwell Busi-
ness, 223-259.
Luthans, F., R. Paul, and D. Baker. 1981. “An Experimental Analysis of
the Impact of Contingent Reinforcement on Salespersons’ Perfor-
mance Behavior.Journal of Applied Psychology 66:314-323.
, , and L. Taylor. 1985. “The Impact of Contingent Rein-
forcement on Retail Salespersons’ Performance Behaviors: A Repli-
cated Field Experiment. Journal of Organizational Behavior
Management 7 (Spring/Summer): 25-35.
Maslow, A. H. 1943. A Theory of Human Motivation. Psychological
Review 50 (July): 370-396.
. 1970. Motivation and Personality. 2d ed. New York: Harper &
Row.
Moncrief, W. C., S. H. Hart, and D. Robertson. 1988. “Sales Contests: A
New Look at an Old Tool. Journal of Personal Selling & Sales Man-
agement 8 (November): 55-61.
Murphy, William H. 2004. “In Pursuit of Short-Term Goals: Anticipating
the Unintended Consequences of Using Special Incentives to Moti-
vate the Sales Force.Journal of Business Research.
 and P. A. Dacin. 1998. “Sales Contests: A Research Agenda.
Journal of Personal Selling & Sales Management 18 (1): 1-17.
Nolan, P. and V. Alonzo. 1997. “Incentive Magazine’s Review of the 1997
Incentive Federation State of the Industry Report. Retrieved from
www.incentivemag.com/facts_n_figures.html.
Oliver, R. L. 1974. “Expectancy Theory Prediction of Salesmen’s Perfor-
mance.Journal of Marketing Research 11 (August): 243-253.
 and E. Anderson. 1995. “Behavior- and Outcome-Based Sales
Control Systems: Evidence and Consequences of Pure-Form and
Hybrid Governance.Journal of Personal Selling & Sales Manage-
ment 15 (4): 1-15.
 and B. A. Weitz. 1991. “The Effects of Risk Preference, Uncer
-
tainty, and Incentive Compensation on Salesperson Motivation.
Working Paper. Report Number 91-104. Marketing Science Institute,
Cambridge, MA.
Schul, P. L., S. Remington, and R. L. Berl. 1990. Assessing Gender Dif
-
ferences in Relationships Between Supervisory Behaviors and Job-
Related Outcomes in the Industrial Sales Force. Journal of Personal
Selling & Sales Management 10 (Summer): 1-16.
Teas, R. K. 1981. An Empirical Test of Models of Salespersons’ Job
Related Expectancy and Instrumentality Perceptions. Journal of
Marketing Research 18 (May): 209-226.
. 1982. “Performance-Reward Instrumentalities and the Motiva
-
tion of Retail Salespeople.Journal of Retailing 58 (Fall): 4-26.
Tyagi, P. K. 1982. “Perceived Organizational Climate and the Process of
Salesperson Motivation.Journal of Marketing Research 19 (May):
240-254.
. 1985. “Relative Importance of Key Job Dimensions and Leader-
ship Behaviors in Motivating Salesperson Work Performance. Jour-
nal of Marketing 49 (Summer): 76-86.
Urbanski, A. 1986. “Incentives Get Specific. Sales and Marketing Man-
agement 136 (April): 98-102.
Vroom, V. H. 1964. Work and Motivation. New York: John Wiley.
Weiner, B. 1986. An Attributional Theory of Motivation and Emotion.
New York: Springer Verlag.
Wildt, A. R., J. D. Parker, and C. E. Harris Jr. 1980-1981. “Sales Contests:
What We Know and What We Need to Know. Journal of Personal
Selling and Sales Management 1 (Fall/Winter): 57-64.
Wotruba, T. R. and D. J. Schoel. 1983. “Evaluation of Salesforce Contest
Performance.Journal of Personal Selling and Sales Management 3
(May): 1-10.
ABOUT THE AUTHORS
William H. Murphy (wmurphy@bus.wisc.edu) is a senior lec-
turer in marketing at the University of Wisconsin–Madison and a
visiting scholar at the China Europe International Business
School (CEIBS) in Shanghai. Previously, he was an assistant pro-
fessor at Babson College and a senior lecturer at the University of
Auckland, New Zealand, where he taught in the Executive
Diploma in Business Program. He has developed distance learn-
ing marketing strategy courses for executives, taught across
numerous marketing topics, consulted for numerous firms in
sales and key account management, and provided guidance for
entrepreneurial ventures in both traditional and internet markets.
He has published in a number of journals, including the Euro
-
pean Journal of Marketing, Industrial Marketing Management,
the Journal of Personal Selling & Sales Force Management,
Marketing Research, and the Journal of Higher Education for
Business and has written cases for a leading sales force manage
-
ment textbook.
142 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE SPRING 2004
Relative Perceived Performance categorical response If you were to gauge your overall performance compared to
other salespeople, where would your performance put you
relative to the rest of the sales team?
Bottom 20% (among bottom performers)
21% to 40%
41% to 60%
61% to 80%
Top 20% (among the top performers)
Supervisor Works with Rep categorical response How often does your boss work with you?
1 day a week or more
2 or 3 days a month
1 day each month
7 to 11 days each year
3 days each year or less
APPENDIX (continued)
Reliability
Measure Scale Items (Coefficient α)
at SAGE Publications on May 20, 2009 http://jam.sagepub.comDownloaded from
Peter A. Dacin (pdacin@business.queens.ca) is a professor of
marketing at Queen’s University in Kingston, Ontario, Canada.
He received his Ph.D. from the University of Toronto. His pri
-
mary teaching and research interests lie in consumer/managerial
judgment formation, brand equity/dilution, corporate reputation,
and research methods and design. He has also published in the
area of sales force management. His research has appeared in
several leading journals including the Journal of Marketing, the
Journal of Marketing Research, and the Journal of Consumer
Research. In addition, he has published in numerous conference
proceedings. He is currently the chair of the American Marketing
Association’s Consumer Behavior Special Interest Group, serves
on the Academic Council of the American Marketing Associa-
tion, and is cofounder of the Corporate Identity/Associations
Research Group.
Neil M. Ford is a professor emeritus of marketing at the Univer-
sity of Wisconsin–Madison. He received his B.S. degree in mar-
keting from Southern Illinois University in 1958. He received his
M.S. (1961) and Ph.D. (1966) in marketing from the University
of Illinois–Urbana. He has been a member of the Marketing fac
-
ulty in the School of Business at the University of Wisconsin
since 1966, serving as chair of the department from 1984 through
1988 and 1996 through 1998. Along with his coresearchers, he
was the 1981 recipient of the William O’Dell Award for the Most
Outstanding Article published in the Journal of Marketing
Research and was selected as the “Marketer of the Year” in 1992
by the Madison Chapter of the American Marketing Associa-
tion. He has received research grants from the Sperry and Hutch-
inson Foundation, Avon Products, the University of
Wisconsin–Madison Graduate School, and the Marketing Sci-
ence Institute. He has authored numerous articles and has
coauthored several books, including the most recent Sales Force
Management: Planning, Implementation and Control (6th ed.).
Murphy et al. / SALES CONTEST EFFECTIVENESS 143
at SAGE Publications on May 20, 2009 http://jam.sagepub.comDownloaded from
... Finally, valence is the desirability of a particular reward. Expectancy theory has been applied to a large number of settings, including incentive to become an entrepreneur (Renko et al., 2012), motivating professional salespeople (Nasri and Charfeddine, 2012;Oliver, 1974) and designing sales contests (Murphy et al., 2004); hence, there is strong precedent for applying it to recruiting salespeople in entrepreneurial ventures. ...
... Expectancy is the amount of effort required to obtain a certain level of performance (Murphy et al. 2004). Expectancies are thus influenced by (1) the amount required effort to achieve a level of performance and (2) the certainty of the relationship between effort and performance. ...
... Da mesma forma que o interesse pela gamificação vem crescendo, poucos estudos têm demonstrado seus efeitos negativos (Seaborn; Fels, 2015; Andrade;Mizoguchi;Isotani, 2016). Algumas barreiras poderão ser evidenciadas nas competições embasadas em incentivos, já que é necessário o acompanhamento e avaliação dos impactos, riscos e das implicações éticas ou legais observadas nos comportamentos gerados Ford, 2004) e nas atitudes dos vendedores para participar e perseguir objetivos Dacin, 2009). Estudos sobre o uso da gamificação para estimular a partilha de conhecimento mostraram que a recompensa baseada em grupos não é apenas menos eficiente do que a recompensa individual, mas também sujeita a um problema potencial de produtividade, no qual os trabalhadores mais produtivos podem não participar na partilha de conhecimento (Lee; Ahn, 2007). ...
Article
Full-text available
Este artículo analiza el uso de la Gamificación en la formación de vendedores de calzado durante cuatro años, a través de la investigación-acción. La implementación del programa de capacitación se llevó a cabo durante los ciclos de implementación, generando mejoras y aprendizajes para la empresa. En general, los resultados de la investigación permitieron minimizar futuros problemas de desempeño en el área de ventas. Los hallazgos de la investigación indican que la Gamificación mejoró el desempeño de los vendedores principalmente en las siguientes dimensiones: argumentos de venta, conocimiento del producto y alternativas para cerrar la venta. El proceso de Gamificación implementado indicó oportunidades para mejorar el contenido de la formación, el análisis del proceso de aprendizaje y la percepción de la realidad que brindan los juegos. Este estudio contribuye a señalar beneficios y requisitos para la aplicación de la Gamificación en la formación de vendedores. La elección de la investigación-acción como metodología y el uso de ciclos de mejora y aprendizaje aumentaron la eficiencia global de la formación, permitiendo una toma de decisiones más rápida, generando mejores resultados y permitiendo la replicación de buenas decisiones y la corrección de errores entre ciclos.
... En cuanto a los planes de carrera, Kappia et al. (2007) y Porfeli et al. (2012, enuncian que no siempre conllevan al crecimiento hacia puestos directivos y profesionales y que la exploración y la carrera en profundidad afectan negativamente la evitación del trabajo. Mientras tanto, cuando se basan en los métodos de recompensas tales como financieras, de reputación y satisfacciones intrínsecas, Lam (2011) y Murphy et al. (2004), utilizaron estos incentivos para rejuvenecer la fuerza de trabajo en las ventas y conducir a una mayor conciencia en las motivaciones intrínsecas y extrínsecas de sus empleados para lograr proyectos con mayor eficacia. De ahí que, DiPietro et al. (2014), Sledge et al. (2008) y Miles andSledge (2009), inclinados por estudios en el sector hotelero, han obtenido resultados que apoyan la teoría de Herzberg, concluyen que uno de los factores de motivación más importantes es: el reconocimiento por un trabajo bien hecho, como la cultura y que éstos influyen en el grado de satisfacción laboral. ...
Article
Full-text available
Este trabajo tiene como propósito analizar la vigencia conceptual sobre los factores de la motivación a partir de la teoría bifactorial propuesta por Herzberg and R. (1963) realizando un acercamiento en dichos factores y perspectivas que imperan en la actualidad; se utilizó como metodología el enfoque cualitativo con base en la revisión de literatura sistematizada desde la base de datos Scopus y el Scimago Journal Rank (SJR). Se observa el análisis de los resultados con base en la actualidad de las investigaciones y los factores de la motivación a partir de una caracterización de los artículos. Así se evidencio el estado de los factores en los últimos años y se encontró que este modelo conductista es vigente e influye en el desarrollo organizacional.
... Fun may allow employees to take momentary time off from their tasks, recharge, and thus be more engaged when on task. Finally, in the context of fun activities, fun could facilitate goal achievement (Murphy, Dacin, & Ford, 2004). ...
Article
Full-text available
Workplace fun can have positive individual and organizational implications. Academic studies have linked workplace fun with job satisfaction. The changing in workforce, as well as the flattening of organizations and a more casual work environment have led to a simplistic assumption that more fun is always better. Different generational cohorts namely Baby Boomers, Generation X and Millennials might have different views of workplace fun which affect their work outcome. This paper investigates how workplace fun of academicians in UiTM Kedah influences their job outcomes, which include job satisfaction, organizational commitment and task performance, and how different generations respond to workplace fun which in the end affects their job outcome. The Statistical Package for Social Science (SPSS) 21 was used to analyse the results.The findings revealed that all generational cohorts tested agreed that workplace fun did affect job outcome in general. However, Generation Y showed a slight difference where workplace fun does not have a significant relationship with job satisfaction. These results, their implications, and directions for future research are discussed.
Article
The present research focuses on the study of the factors of sales optimization in a tobacco company. The reasons for undertaking this study were to identify and analyze the explanatory factors of the optimization of the sale in the commercial company in general and in a tobacco company in particular, in order to highlight their influence in the optimization of the productivity of this one and to have an idea on the evolution of the sales of this company in order to release the gaps which exists between the realized turnover compared to the various expenses related to the marketing and the distribution The results of this study proved that the factors of optimization of the sale of the GMM/Isiro would be the implementation of an effective marketing policy which would bring the product closer to the consumers and a good management of its sales force by starting with the recruitment, the training, the remuneration, the animation, the control and the evaluation. The correlation coefficient is 0.4873 > 0 : therefore there is a presumption of a positive correlation between turnover (Yt) and marketing expenses (X_t), i.e. the two variables move in the same direction : if marketing expenses increase, turnover also increases and if marketing expenses decrease, turnover also decreases ; the coefficient of determination is 0. 2375 i.e. the model is globally explained at 23.75 % by the endogenous variable (the turnover) which depends on the exogenous variable (the marketing expenses); in other words on the 100% of the sales realized by the GMM, 23.75 % are realized according to the marketing expenses The remaining 76.25% are explained by other variables than marketing expenses such as preference, price, taste, environment, product availability or permanence, agent motivation, habit, etc.
Article
Compensation is one of the most effective methods used to align and motivate salespeople to accomplish sales and organizational objectives. For this reason, sales researchers have made considerable strides in understanding the impact that compensation structure has on salespeople and salesforce performance. In this article, we examine the theoretical foundations of the sales compensation literature. We then perform an extensive review of this literature to identify the perceptual and behavioral outcomes associated with incentive- and salary-based compensation. Finally, the limitations of the sales compensation research are identified, and future studies are proposed.
Article
Although companies are increasingly deploying inside sales units, knowledge is scarce regarding how to effectively incentivize them. In addressing this neglect, this study draws on network theory to scrutinize how various unit and individual financial and non-financial incentives affect inside sales units’ performance. In addition, the authors examine the contingent roles of two unit-level network measures, density and centralization. Using data from 366 salespeople working in 118 inside sales units, the authors empirically test the predicted relationships. Results reveal that unit incentives are positively related to unit performance, whereas individual incentives notably have a negative relationship with unit performance. Results further reveal that both density and centralization influence the incentive–unit performance relationships. Overall, findings indicate that the effects of various incentives in the inside sales unit context differ from those in other contexts, resulting in important implications for managers.
Article
The organizational frontline in retail represents the frontline for an entire supply chain. This channel structure distances the branded supplier from the end consumer and makes the supplier dependent on a retail frontline salesperson that (1) it does not control and (2) also represents competing brands. This study reveals mechanisms that the supplier may use to influence retail frontline salespeople. We demonstrate the importance of consumer marketing programs and supplier representatives in building brand identification between the manufacturer's brand and the retail frontline employee that translates into increased brand sales, while also revealing the role of rewards programs in stimulating brand-specific extra-role behaviors. Interestingly, retail frontline customer orientation, built by the retailer, diminishes the effect of brand extra-role behaviors toward the focal brand on the focal brand's sales, providing an informal control mechanism for retailers to protect its objectives. The results of this study have powerful ramifications for both retailers and suppliers in achieving both their mutual, and sometimes differing, objectives related to the retail frontline.
Article
This research uses quasi-experimental, control group design to examine the performance and behavioral impact of team design on sales performance. We assess team versus individual performance and team composition. Drawing on motivation gain literature, we evaluate the impact of team composition based on the relative ability of the team members. The research is in a field sales setting in collaboration with a major insurance provider, providing the first example of assessing Group Motivation Gain (GMG) in a sales setting through a field experiment. Our paper extends previous research in this domain by considering outcome interdependence, not merely task interdependence, evaluating the performance of both team members, assessing motivation gain in the context of a task over a longer duration, and building the related nomological network. The findings demonstrate improved overall performance for the team and the individual members of the team, the gains were particularly pronounced when members have moderate levels of difference in ability, rather than small or large differences in ability. We discuss managerial implications of our findings and suggest further research directions.
Article
Full-text available
With the movement in the U.S. economy toward a total quality environment, there will be a greater focus on relationships building within an organization. This study sought to empirically explore the association between sales manager salesperson relationships and salesperson motivation, stress, and evaluation of the manager. Results suggest that cadres (high quality relationships) are higher on extrinsic and intrinsic instrumentality, extrinsic valence, and evaluation of their manager. Cadres are lower on the role overload, role insufficiency, role ambiguity, and role conflict. Implications of these findings and suggestions for future research are offered.
Article
Full-text available
In this chapter a theory of motivation and emotion developed from an attributional perspective is presented. Before undertaking this central task, it might be beneficial to review the progression of the book. In Chapter 1 it was suggested that causal attributions have been prevalent throughout history and in disparate cultures. Studies reviewed in Chapter 2 revealed a large number of causal ascriptions within motivational domains, and different ascriptions in disparate domains. Yet some attributions, particularly ability and effort in the achievement area, dominate causal thinking. To compare and contrast causes such as ability and effort, their common denominators or shared properties were identified. Three causal dimensions, examined in Chapter 3, are locus, stability, and controllability, with intentionality and globality as other possible causal properties. As documented in Chapter 4, the perceived stability of a cause influences the subjective probability of success following a previous success or failure; causes perceived as enduring increase the certainty that the prior outcome will be repeated in the future. And all the causal dimensions, as well as the outcome of an activity and specific causes, influence the emotions experienced after attainment or nonattainment of a goal. The affects linked to causal dimensions include pride (with locus), hopelessness and resignation (with stability), and anger, gratitude, guilt, pity, and shame (with controllability).
Article
The purpose of this article is to examine the role of goal commitment in goal-setting research. Despite Locke's (1968) specification that commitment to goals is a necessary condition for the effectiveness of goal setting, a majority of studies in this area have ignored goal commitment. In addition, results of studies that have examined the effects of goal commitment were typically inconsistent with conceptualization of commitment as a moderator. Building on past research, we have developed a model of the goal commitment process and then used it to reinterpret past goal-setting research. We show that the widely varying sizes of the effect of goal difficulty, conditional effects of goal difficulty, and inconsistent results with variables such as participation can largely be traced to main and interactive effects of the variables specified by the model.
Article
This article examines the relative importance of key job dimensions and leadership characteristics in enhancing salesperson motivation and work performance. To gain a deeper insight, the relative effects of job dimensions and leadership behavior on intrinsic and extrinsic work motivation of salespersons are examined, with results indicating that key job dimensions are more instrumental in enhancing work motivation, and leadership behavior more influential in effecting extrinsic motivation. A number of major implications based on these findings are discussed.
Article
The author examines how organizational climate contributes to salespersons' intrinsic and extrinsic motivation to perform. On the basis of expectancy-valence theory of motivation, specific relationships between organizational climate and motivational components are tested using a sample of insurance salespersons. Managerial implications and future research directions are discussed.
Article
The author reports the results of a study of the motivational implications of the industrial salesperson's personal characteristics and his or her perceptions of the job, the company's organization, and selling constraints. Predictor equations for the salesperson's expectancy and instrumentality estimates are tested empirically. The results indicate the salesperson's personal characteristics and his or her perceptions of supervisory style, organizational communication, job significance and autonomy, job variety and completeness, job complexity, and selling constraints are potentially important predictors of salesforce motivation.