Title: Marketing Performance Measurement Ability and Firm Performance
Authors: O"Sullivan, Don., Abela, Andrew V
Subject: Marketing & CRM
Publish: 2007
Status: full text
Source: Journal of Marketing; Apr2007, Vol. 71 Issue 2, p79-93
Preparation: Scientific Database Management Journal Articles www.SYSTEM.parsiblog.com
Abstract: Marketing practitioners are under increasing pressure to demonstrate their contribution to firm performance. It has been widely argued that an inability to account for marketing"s contribution has undermined its standing within the firm. To respond to this pressure, marketers are investing in the development of performance measurement abilities, but to date, there have been no empirical studies of whether the ability to measure marketing performance has any actual effect on either firm performance or marketing"s stature. In this study of senior marketing managers in high-technology firms, the authors examine the effect of ability to measure marketing performance on firm performance, using both primary data collected from senior marketers and secondary data on firm profitability and stock returns. They also explore the effect of ability to measure marketing on marketing"s stature within the firm, which is operationalized as chief executive officer satisfaction with marketing. The empirical results indicate that the ability to measure marketing performance has a significant impact on firm performance, profitability, stock returns, and marketing"s stature within the firm.    --Download Article

Introduction: The effective dissemination of new methods of assessing marketing productivity to the business community will be a major step toward raising marketing’s vitality in the firm and, more important, toward raising the performance of the firm itself” (Rust et al. 2004,p. 76). In response to the pressure on marketers to demonstrate their value to the firm, there have been several highprofilecalls for more research in the area of marketing performance measurement (MPM) and several conceptual andempirical research papers (e.g., Donthu, Hershberger, and Osomonbekov 2005; Lukas, Whitwell, and Doyle 2005;Rust et al. 2004). Furthermore, there have been regular calls for marketing practitioners to develop and enhance their ability to account for marketing’s contribution to firm performance (Ambler 2003; Bolton 2004). An assumption underlying these related academic and practitioner concerns is that developing and applying MPM ability leads to both greater status for marketing at the board level (see, e.g., Webster, Malter, and Ganesan 2005) and improved firm performance (Morgan, Clark, and Gooner 2002). However, to date, the relationship between MPM ability and either firm performance or marketing’s stature within the firm has not been demonstrated empirically.
The primary purpose of this article is to test whether MPM ability contributes to actual firm performance or to marketing’s stature within the firm, which we operationalize as chief executive officer (CEO) satisfaction with marketing. A secondary purpose is to explore two potentially distinct aspects of MPM ability: the ability to measure performance across a range of marketing activities (e.g., advertising, trade promotion, direct mail) and the ability to assess marketing performance using a comprehensive set of metrics (e.g., financial, nonfinancial). We focus on firms in the high-technology sector. We chose high-tech firms because of the recognition that within this sector, marketing has been under intense pressure to demonstrate its contribution to firm performance. There are two primary reasons for this pressure. First, high-tech companies tend to have more of an engineering orientation than a marketing orientation, and thus top management tends to be more skeptical about the value of marketing (Davies and Brush 1997). Second, during the period we studied (early 2000s), the sector experienced the collapse of the “technology boom,” which led to sharply increased scrutiny of marketing activities (Mohr and Shooshtari 2003). We begin by reviewing the MPM literature and generating several testable hypotheses.

Measurement and Performance
A long-standing caricature of marketing practitioners is that they love to spend money and hate to assess the results of that spending (e.g., Adler 1967). Marketers’ inability to account for the function’s contribution to firm performance is recognized as a key factor that has led to marketing’s loss of stature within organizations (Kumar 2004; Lehmann 2004; Webster, Malter, and Ganesan 2005). This is reflected in increased demand for greater accountability (Doyle 2000; Morgan, Clark, and Gooner 2002; Rust et al. 2004). In addition, there have been several high-profile calls for more research in the area of MPM. Most notably, MPM topics have been consistently listed among the Marketing Science Institute’s (1998, 2000, 2002, 2004, 2006) top priorities.
Marketing performance measurement is the assessment of “the relationship between marketing activities and business performance” (Clark and Ambler 2001, p. 231). Because the problem in question is the inability to account for marketing activities, our specific interest is in marketing’s ability to assess this relationship. Given that the goal of MPM research is to demonstrate the value of the marketing activities, in line with the work of Rust and colleagues (2004), our focus is on marketing not as the “underlying products, pricing, or customer relationships” (Rust et al. 2004, p. 76) but rather as the “marketing activities” themselves, which we define as marketing communication, promotion, and other activities that represent the bulk of the typical marketing budget. Marketing performance measurement research can be divided into three research streams: measurement of marketing productivity (e.g., Morgan, Clark, and Gooner 2002; Rust, Lemon, and Zeithaml 2004), identification of metrics in use (e.g., Barwise and Farley 2003; Winer 2000), and measurement of brand equity (e.g., Aaker and Jacobson 2001; Ailawadi, Lehmann, and Neslin 2002). Rust and colleagues (2004) build on the work of Srivastava, Shervani, and Fahey (1998) to describe a “chain of marketing productivity” that extends from marketing activities to shareholder value. Marketing activities influence intermediate outcomes (customer thoughts, feelings, knowledge, and, ultimately, behavior), which in turn influence financial performance of the firm. The MPM research we cited examines how marketers can measure the relationships along the chain of marketing productivity; which metrics firms use or could use along this chain, particularly financial, nonfinancial, and market-based assets; and contextual factors, particularly the firm’s market orientation (e.g., Clark and Ambler 2001).
Underlying all this work is the assumption that such measurement effort is beneficial to the firm and is not just a post hoc justification of marketers’ efforts—that improvements in marketing’s ability to account for its activities will actually raise the performance of the firm. In the face of senior management demands that marketers demonstrate their value, the desire for justification is understandable. However, overcoming the inability to account for the function’s contribution to firm performance requires that resources and management attention be expended on measurement efforts (Bonoma and Clark 1988). Incurring such cost assumes that the firm will benefit, and testing this assumption is the primary purpose of this article.

Hypotheses
We develop hypotheses based on a theoretical framework that links MPM ability to firm performance and CEO satisfaction with marketing. We begin by hypothesizing that MPM ability has an effect on actual firm performance. Sevin (1965) argues that the implementation of robust performance measures should result in greater marketing and firm performance. Several arguments that link MPM to improvements in marketing and firm performance have been advanced (see, e.g., Rust et al. 2004). First, anticipation of the scrutiny of marketing efforts will encourage greater attention to the activities that will be measured. The idea that “what gets measured gets done” is well founded in the management literature (see, e.g., Ouchi 1979) and is assumed within the MPM literature. Second, Webster, Malter, and Ganesan (2005) contend that marketing’s contribution to the achievement of strategic goals is underrepresented in firms that do not measure marketing performance and that the performance of such firms may suffer as a result. Third, it has been argued that MPM should lead to learning, which should enable improved marketing decisions and, consequently, performance (Morgan, Clark, and Gooner 2002). Fourth, MPM offers performance feedback, and performance feedback has consistently been found to influence both managerial attitudes and behavior (Audia, Locke, and Smith 2000; Curren, Folkes, and Steckel 1992; Greve 1998; Miller 1994). Finally, feedback relative to goals has been demonstrated to produce strong effects (e.g., Locke and Latham 1990). Thus:

H1: MPM ability positively influences firm performance. It has long been recognized that the marketing function typically plays a limited role in the process of strategy formulation (Anderson 1982; Day 1992; Webster 1992). Srivastava, Shervani, and Fahey (1998) argue that an important reason for this is that marketers struggle to measure and communicate to top management the impact of marketing activities on firm performance. Lehmann (2004) and Webster, Malter, and Ganesan (2005) observe that marketing has the greatest influence and stature in firms in which there are clear measures of marketing’s contribution. Accordingly, H2: MPM ability is positively associated with CEO satisfaction with marketing.
As we noted previously, although the primary purpose of this article is to test empirically whether MPM ability contributes to firm performance or to marketing’s stature within the firm, a secondary purpose is to explore the ability to measure performance across a range of marketing activities and the ability to assess marketing performance using a comprehensive set of metrics. Although the two aspects are clearly related in that they both contribute to a firm’s MPM ability, we hypothesize that they are distinct. For example, one firm may be able to measure the performance of its
advertising or public relations (activities) only in terms of changes in awareness (nonfinancial metric), whereas nother firm may be able to measure them in terms of revenue change (financial metric) and against specific goals and
competitor performance (benchmark metric) (Ambler 2003). The focus of the broader marketing accountability literature
has been on the importance of the ability to measure disparate marketing activities (e.g., Rust, Lemon, and Zeithaml
2004; Webster, Malter, and Ganesan 2005). In addition, the discussion of MPM among practitioners has also
tended to focus on the activities dimension (e.g., McMaster 2002). However, within the existing academic literature on
MPM, the focus has tended to be on the metrics in use (e.g., Ambler, Kokkinaki, and Puntoni 2004; Barwise and Faley
2003, 2004; Lages, Lages, and Lages 2005; Lukas, Whitwell, and Doyle 2005).
We hypothesize that both the activities and the metrics aspects have separate but related effects on performance
and CEO satisfaction. Because the activities aspect precedes the metrics aspect both theoretically and logically, we est the activities aspect first:
H3: The ability to measure performance across the range of marketing activities a firm employs positively influences
firm performance.
H4: The ability to measure performance across the range of marketing activities a firm employs is positively associated with CEO satisfaction with marketing. As we noted previously, the identification of metrics in
use is one of the main streams of MPM research (e.g., Ambler, Kokkinaki, and Puntoni 2004; Barwise and Farley
2003, 2004; Lages, Lages, and Lages 2005; Lukas, Whitwell, and Doyle 2005). An assumption underlying this esearch stream is that choice of metrics matters.
Researchers in this area have concluded that in their choice of metrics, firms should employ both financial and nonfinancial
metrics (Clark 1999; Rust et al. 2004) and that they should compare these against goals and competitors
(Ambler 2003). Thus, we expect that firms that follow this guidance and are able to assess marketing performance
using a broad set of metrics (financial and nonfinancial, in relation to goals, and in relation to competitors) should outperform hose that lack this ability. It has previously been noted that the academic community’s focus on metrics in
use has had little impact on practicing marketers (Clark 1999). Reflecting this, we are interested in isolating the impact of metrics ability on performance and CEO satisfaction beyond that which is accounted for by activities ability.
Thus:
H5: The ability to provide a comprehensive set of metrics positively influences firm performance.
H6: The ability to provide a comprehensive set of metrics is positively associated with CEO satisfaction with marketing.
Dashboards are a variation of a balanced scorecard (Kaplan and Norton 1992) and are used as a means to report
key metrics to senior management from the array of information generated by corporate information systems (Paine
2004; Wind 2005). Ambler (2003) describes a dashboard as a refined set of marketing performance data, usually presented together, which communicate an overview of strategic
performance. Two important elements of dashboards are that they provide automated or (close to) real-time reporting
(Iyer, Lee, and Venkatraman 2006; Wind 2005) and that they enable users to “drill down” to program-level details
(Miller and Cioffi 2004). It has been noted that dashboards,
which are increasingly popular among marketing practitioners,
have received only limited attention in marketing research (Rogers 2003). Recently, Srivastava and Reibstein
(2005) have called for more research on the role of dashboards in managing marketing productivity.
Dashboards are viewed as a means by which information  can be summarized and readily communicated to senior decision makers (Srivastava and Reibstein 2005). It is argued that this distilling of data increases the perceived
value and managerial use of information (Peyrot et al.
2002), which in turn creates a closer link between marketing
activities and firm goals (McGovern et al. 2004; Miller
and Cioffi 2004). Therefore, the use of a marketing dashboard
is hypothesized to act as a moderator in the relationships
between ability to measure and performance and
between ability to measure and CEO satisfaction.
H7: The greater the use of a marketing dashboard, the more
positively MPM will influence firm performance and
CEO satisfaction.
The study controls for firm size and firm age because
both variables have previously been shown to affect performance
(e.g., Ahuja and Lampert 2001; Miles, Covin, and
Heeley 2000). In controlling for firm age, the study follows
previous research on high-tech firms (e.g., Hill and Naroff
1984). Firm age is accepted as influencing performance
through the ability to learn in the customer relationship and
on competitive advantage outcomes (Zahra, Ireland, and
Hiltt 2000). We summarize the relationships outlined in this
section in Figure 1.
Method
Sample and Procedure
A survey of senior marketers in high-tech firms about MPM
ability, CEO satisfaction with marketing, and aspects of
firm performance produced the primary data for our
research. We used the membership list of the CMO Council
as the sample frame for our study. The CMO Council is a
U.S.-based, not-for-profit organization for senior marketers
in high-tech firms. The council’s membership is global,
though at the time of study, it was heavily skewed toward
North American firms. The membership list contains names
and background information (title, firm, and contact details) or all members. We collected survey responses through an
online, structured survey. We supplemented the primary
data captured through the survey with secondary data on
aspects of firm performance.
The study sought responses from key informants.
Because the CMO Council’s membership is limited to
senior marketers, we included in the sample all members
other than those who worked in marketing services
providers, such as advertising agencies. We subsequently
analyzed the responses to ensure that the respondents had
senior marketing responsibilities (job title) before we
included them in further analysis. The views of key informants
are widely used within the marketing literature (see,e.g., Day and Nedungadi 1994; Moorman and Rust 1999;
Narver and Slater 1990).
Before constructing the questionnaire, we conducted
preliminary in-depth interviews with 17 chief marketing
officers (CMOs). These discussions focused on the interviewees’
understanding of and motivations for measuring marketing
performance. A strong functional orientation was
apparent; interviewees were most interested in measuring the performance of the marketing function as opposed to
the broader marketing performance of the firm. In addition,
respondents were interested in assessing performance
impact at the firm level. In short, MPM was viewed as an
assessment of the marketing function’s contribution to firm
performance. The interviews provided a basis for the development
of our survey questionnaire.
The questionnaire was divided into three sections that
contained questions related to MPM ability, firm performance,
and respondent profile. The questionnaire included
a 15-item scale to quantify the ability to measure performance
across a range of marketing activities and a 4-item
scale to measure the ability to assess performance using a
comprehensive set of metrics. These scales reflected the
views captured from our interviews with CMOs and a
review of the literature. To test for comprehension, relevance,
and completeness, we pilot-tested the questionnaire
with ten senior marketers from the CMO Council. Participants
in the pilot phase were asked to identify any problems
they encountered with the e-mail invitation, the content of
the questionnaire, or the process of completing it online.
Participants were also asked to evaluate the clarity of the
questions and the response formats. No major difficulties
were identified, though we clarified some of the response
options and revised the questionnaire accordingly.
The survey was administered online between February
and March 2004. A total of 810 marketers received e-mail
notification of the survey. This was followed 14 days later
by a reminder e-mail to nonrespondents. Each e-mail contained
an embedded link to the survey. We received 214
usable response, for a response rate of 26.4%. This response
rate was highly satisfactory given that rates ranging from
12% to 20% are considered acceptable for cross-sectional
samples (Churchill 1991). We tested for nonresponse bias
using time-trend analysis (Armstrong and Overton 1977).
We selected two subsamples from early and late respondents.
Because these did not differ in terms of respondent
profile or the variables of interest, we concluded that nonresponse
bias was not a significant concern.
We collected survey responses over a four-week period.
After that time, we made the survey available through several
additional channels, most notably a BusinessWeek
research panel. This produced an additional 98 qualified
respondents, for a total of 312 responses. Subsequent analysis
of these additional 98 respondents indicated that they
were not materially different from the first 214 respondents
in terms of job title and sector. Furthermore, their responses
to the key issues under consideration in the study were
similar to those of the original 214 respondents. Consequently,
we included them for further analysis. In total, we
included 312 responses in subsequent analysis.
The job titles of respondents represented the range of
possible senior marketer titles: 17% were CMOs, 40% were
vice presidents of marketing, and 15% were marketing
directors. Of the 27% who answered “other,” most were
senior managers with titles such as president or vice president
of sales and marketing. Respondent firms were drawn
from a cross-section of information technology–related sectors:
36% were software providers, 35% provided Internetrelated
services, 3% provided components, 3% provided
computer systems, and 3% provided networking products
and services. Peripherals and integration accounted for 2%
and 1%, respectively. Of the 17% that responded “other,”
most were application service providers, information technology
consulting services, or telecommunications-related
products and services. Firm age varied greatly among respondents, and most (>90%) were headquartered in North
America.
Measurement1
We calculated MPM ability as the simple average of a
firm’s scores on the activities and metrics scales. We
assessed MPM ability using a 15-item scale based on our
in-depth exploratory interviews with CMOs. We recorded
responses on a seven-point scale anchored by “poor” and
“excellent.” These activities included above- and below-theline
promotional activities as well as marketing planning
and customer relationship management. As we noted previously,
because the issue being addressed is marketers’
inability to account for marketing activities, our specific
interest here is in marketing’s ability to assess this relationship.
Having an ability does not necessarily mean using it,
but given the demands being placed on marketers in hightech
firms at the time of this study to account for their contribution
more effectively, it seems highly unlikely that any
MPM ability would have remained latent. Therefore, we
assume that any firm in our sample that had an MPM ability
was indeed using it. Discussions with the CMO Council’s
Steering Committee and interviews with 17 CMOs during
the exploratory phase of our research indicated that this
assumption was reasonable.
In our study, metrics is a construct that consists of the
summed responses to four questions. Again, we captured
responses on a seven-point scale anchored by “poor” and
“excellent.” Over the past 40 years, ranges of metrics have
been proposed for MPM (for a review, see Clark 2001).
These include both financial and nonfinancial measures.
The inclusion of nonfinancial measures is considered an
important progression because it helps provide a more complete
deion of marketing’s contribution. Financial and
nonfinancial measurement are two of the four aspects of
metrics ability we considered. The other two aspects of
metrics we assessed are related to benchmarking. Bonoma
(1989) was one of the first researchers to argue for greater
benchmarking of marketing performance. More recently,
Vorhies and Morgan (2005) have demonstrated the impact
of the benchmarking of marketing capabilities on firm performance.
Consequently, we included the ability to benchmark
against plan and against competitors in our understanding
of metrics. The resultant scale was reliable (? =
.83).
Dependent measures. Our dependent measures were
firm performance and CEO satisfaction with marketing. We
assessed firm performance using both primary and secondary
data. Primary data were provided through our survey
of senior marketers. In the past, the most common measures
of output in firm-level marketing performance studies
have been profit, sales, market share, and cash flow
(Bonoma and Clark 1988). Financial measures, such as
sales and profit, continue to be the most important MPMs
(Clark 2000; Kokkinaki and Ambler 1999). Several studies
have suggested that managers balance profitability and sales
growth (McKee, Varadarajan, and Pride 1989; Slater and
Narver 1996), and others have considered market share a
measure of firm performance (Jaworski and Kohli 1993).
Accordingly, and in line with previous studies, we measured
performance as the mean of a respondent’s rating for
his or her firm’s sales growth, market share, and profitability
performance relative to all other competitors. We captured
responses on a five-point scale anchored by “very
poor” and “outstanding.” We measured CEO satisfaction
with marketing as the response to a single question. We captured
responses on a five-point scale anchored by “excellent”
and “poor.”
Following the work of Rust, Moorman, and Dickson
(2002), we captured secondary data on firm performance
through two measures: return on assets (ROA) and stock
returns. We calculated ROA as the firms’ overall ROA for
the 12 months subsequent to our original study, as reported
in COMPUSTAT. This time lag enabled us to determine the
direction of causality between MPM and firm performance.
Using data provided by the University of Chicago’s
Center for Research in Security Prices (CRSP), we measured
stock returns as the firms’ size-adjusted stock returns
for the 12 months subsequent to the original study. The
CRSP provides data on stocks traded on each of the major
U.S. stock exchanges: NYSE, AMEX, and NASDAQ. We
calculated returns as the difference between an individual
firm’s stock returns and the value-weighted average return
for all firms in the same size decile of the sample firm in
CRSP’s size decile portfolio for each month. Return data
were adjusted for both stock dividends and splits for each
firm by CRSP (Rust, Lemon, and Zeithaml 2004). We calculated
each firm’s return for the period, referred to as the
holding period return, as follows:...      --Download Article