You are on page 1of 21

A Taxonomy of Manufacturing Strategies

Jeffrey G. Miller Aleda V. Roth


School of Management, Boston University, 621 Commonwealth Avenue, Boston, Massachusetts 02215 Kenan Flakier Business School, University of North Carolina, Chapel Hill, North Carolina 27599

his paper describes the deveiopment and analysis of a numerical taxonomy of manufacturing strategies. The taxonomy was developed with standard methods of cluster analysis, and is based on the relative importance attached to eleven competitive capabilities defining the manufacturing task of 164 large American manufacturing business units. Three distinct clusters of manufacturing strategy groups were observed. Though there is an industry effect, all three manufacturing strategy types are observed in various industries. The two main dimensions along which the manufacturing strategy groups differ are the ability of the firms in them to differentiate themselves from competition with their products and services, and the scope of their product lines and markets, A general method for mapping manufacturing strategies on these dimensions is described. For each manufacturing group, the relationships between the competitive capabilities (which describe the manufacturing task), the business context (the business unit strategy), manufacturing activities (manufacturing strategy choices), and manufacturing performance measures are explored and compared.
{Manufacturing; Strategy; Taxonomy; Typology)

1. Introduction
The strategic view of manufacturing as a competitive weapondatesbackat least to Miller and Rogers (1956). They did not differentiate between a business strategy and a manufacturing strategy. Rather, they saw manufacturing policies as necessary ingredients of business strategy. The notion of manufacturing strategy as a separate but related functional component of a business unit strategy is of more recent vintage (Skinner 1978, Hayes and Wheelwright 1984), Two core elements are central to the definition of a manufacturing strategy as a functional substrategy. The first element is a statement of "what the manufacturing function must accomplish" (Skinner 1978). This statement, commonly referred to as the "manufacturing task," is defined in terms of the capabilities the manufacturing unit must have in order for the firm to compete given its overall business and marketing strategy. Lists of critical competitive capabilities typically include quality, cost/efficiency, delivery/responsiveness, and flexibility. Recent lists also include innovation and customer service as important capabilities (Giffi et al. 1990).
0025-1909/94/4003/0285$01.25
Copyriphl i 1994, The Institule of Management Sciences

Hill (1989) has linked the manufacturing task to customer needs by defining the manufacturing task in terms of those capabilities that are critical to winning customer orders. The second element of a manufacturing strategy is defined by the pattern of manufacturing choices that a company makes (Hayes and Wheelwright 1984, Wheelwright 1984, Hayes etal. 1988, Hill 1989). Hayes and Wheelwright (1984) classify these strategic manufacturing choices into two categories. The first category refers to structural or "bricks and mortar" decisions about facilities, technology, vertical integration, and capacity. The second category of choices is concerned with major decisions about the manufacturing infrastruchire, such as organization, quality management, workforce policies, and information systems architecture. The central theme that links the two elements of a manufacturing strategy is the notion that the pattern of choices followed by manufacturing must be congruent with the manufacturing task (Anderson et al. 1989, Buffa 1984, Cohen and Lee 1985, Fine and Hax 1985, Wheelwright 1978, 1984, Hayes and Wheelwright 1984,

MANAGEMENT SciENCE/Vol, 40, No, 3, March 1994

285

MILLER AND ROTH


A Taxonomy of Manufacturing Strategies

Schroeder et al. 1986, Skinner 1969, 1978, 1985, Roth et al. 1989, Hill 1989, Roth and Miller 1990, Stobaugh and Telesio 1983, Swamidass and Newell 1987). The demand that manufacturing choices and manufacturing tasks be linked follows from the presumption that good designs (such as those specified by the manufacturing choices) meet appropriate design criteria (as defined by the manufacturing task). The implication of this presumption is that superior performance will follow for those firms that develop congruence between their business and manufacturing strategies. Evidence to support this implication is growing (Skinner 1978, Buffa 1984, Swamidass and Newell 1987, Roth 1989, Roth and Miller 1990, 1992), Though the basic elements of a manufacturing strategy are commonly accepted, important questions remain about how to operationalize them, and about the nature of the connections between task and choices. Batten et al. (1978), in their description of functional area research on operations strategy and performance noted, "In general these models deal with existing operations and are not strategic in outlook, i.e., they are not concerned with changing the existing operations, nor do they take a global encompassing view of thefirm."Ten years later, a literature review of manufacturing strategy undertaken by Anderson et al. (1989), describes a growing body of research that attempts to take a more comprehensive view. However, they conclude that knowledge about key relationships among manufacturing tasks, manufacturing choices, and business strategies remains disappointingly small. The research described in this paper has two related purposes. The first purpose is to identify strategic groups of manufacturers with similar manufacturing tasks, that is, with similar sets of competitive capabilities. Determining the types of strategies extant has significant value for the emerging field of manufacturing strategy research. Taxonomies provide parsimonious descriptions which are useful in discussion, research and pedagogy. Moreover, finding sets of manufacturers with common profiles may reveal insights into underlying structures of competition as viewed from the perspective of the manufacturing function. Are the manufacturing strategies of firms shaped by market or other forces? To what extent do manufacturing managers in different industries and markets share similar views on the manufacturing task?
286

The second purpose of this research is to explore the central theme in the manufacturing strategy literature by determining and comparing how members of manufacturing strategy groups typically define their business strategies, manufacturing choices, and performance measures. This analysis illuminates the extent to which the Ht between strategy, actions, and measures can be observed for groups. This paper is organized into five sections, thefirstof which is this introduction. Section 2 provides a review of the application of taxonomies in strategy research as background for this inquiry. In 3, the specific variables forming the basis for the classification scheme are identified, and a description of the data and methods used in the analysis is presented. Here we report on the strategy groups identified by cluster analysis procedures and also outline the results of a canonical discriminant analysis. Canonical analysis aids in interpreting the underlying factors that separate the manufacturing strategy groups. In 4, we report on how the strategy groups systematically differ from one another in terms other than those used to define them in the first place. ANOVA techniques are employed to identify the contextual variables, manufacturing strategy choices, and performance measures which correspond to the manufacturing strategy groups. We conclude in 5 with a summary of findings and conclusions, as well as opportunities for future research.

2. Background
The determination of homogeneous groups of firms based upon taxonomies has been an important research theme in the general strategic management and organization literature. (Hambrick 1983a, Fahey and Christensen 1986, and McGee and Thomas 1986 provide useful reviews). Most of the research has recognized that firms can be classified by multiple variables into groups which are best characterized by the "gestalt" of the communalities they share (Miller and Friesen 1977). A variety of different classification variables or taxons have been employed to form the diverse taxonomies that have been developed. In some cases, the classification schemes are based on the context of the firm, cast in terms of environmental, technological, or product descriptors. For example, the stages of a product's life cycle have been used by many researchers to discrimMANAGEMENT SciENCE/Vol. 40, No. 3, March 1994

MILLER AND ROTH A Taxononuf of Maniifactiiriug Strategies

inate among various types of strategies and to explain behavior (Utterback and Abernathy 1975, Thietart and Vivas 1984, Horwitch and Thietart 1987), The typologies of organizational theorists have been based on such factors as the degree of uncertainty or complexity in the environment (Duncan 1972, Galbraith 1973, Lawrence and Lorsch 1969, Thompson 1967) and on technology type (Woodward 1965, and others). Hambrick and Lei (1985) provide an overview and prioritization of the contingency variables which have been used as taxons, as well as those that might be used. Another line of taxonomical research attempts to categorize firms in terms of strategic decision variables. The "strategic group" research of Harrigan (1985), Hatten et al, (1978), and Cool and Schendel (1987), and others has discriminated between firms on the basis of the scope of their business, as well as their resource allocation patterns. Others, such as Porter (1980), and Kim and Lim (1988) have classified firms on the basis of classical industrial organization constructs such as the degree of competition, and bargaining power. Miles and Snow (1978) have categorized firms on the basis of their behavior with respect to competitors (e,g,, differentiators, focusers, defenders) or new opportunities (prospectors vs, reactors). Clearly, the basis for strategic group formation has varied considerably from one study to another (McGee and Thomas 1986, Fahey and Christensen 1986), Pegels and Sekar (1989) conclude that the most appropriate way to form strategic groups depends on what the researcher intends to accomplish. We believe that the competitive capabilities which define the manufacturing task are the appropriate grouping criteria for the purposes of this research. The manufacturing task as revealed through ratings of these capabilities indicates the "strategic intent" (Hamel and Prahalad 1989) of manufacturing, and provides a basis for testing whether the business strategy and manufacturing strategy choices are consistent with this intent, Stobaugh and Telesio (1983), who have developed the only other taxonomy of manufacturing strategy groups that we have found, have also used the manufacturing task to define their strategic groups. Their taxonomy can best be referred to as a "conceptual" taxonomy, or typology, since it was inferred from their review of case studies describing over 100 multinational firms. They hypothesized the existence of cost, techMANAGEMENT SCIENCE/VOI.

nology, and market driven manufacturing strategy groups on the basis of their a posteriori empirical review of these cases. The cost-driven group emphasized the capability to produce at low cost, the technology-driven firms emphasized flexibility to introduce new products, while market-driven firms focused on quality and delivery. For each manufacturing strategy group, Stobaugh and Telesio identified critical manufacturing decisions and choices. In contrast to the work of Stobaugh and Telesio, the taxonomy developed in this paper borrows techniques from the biological and social sciences to develop a "numerical" taxonomy using primary as opposed to secondary sources of information. The taxonomy is developed by applying multivariate statistical procedures and grouping algorithms to measures of the perceptions of manufacturing managers. The research reported in this paper goes beyond the formulation of a numerical taxonomy by attempting to identify the constructs that underlie its formation, and by observing the apparent relationships between group membership, business context, manufacturing choice, and manufacturing performance measures.

3. Methods
3.1. The Sample The data for this study were obtained from the 1987 Manufacturing Futures Project (MFP) Survey (Miller and Roth 1988), The annual Manufacturing Futures Project Survey is specifically designed to gather information on competitive factors pertinent to leading manufacturing business units. The underlying logic of the survey instrument was established in thefirstManufacturing Futures survey administered at Boston University in 1981 (Miller 1982). Since 1983, the survey has also been conducted internationally on an annual basis (DeMeyer et al. 1989, Ferdows et al. 1987, Miller et al. 1989, Roth et al. 1989). The duster analysis reported in this paper is restricted to the 1987 North American respondents. A manufacturing business unit (MBU) is defined by the level in the organization where a manufacturing strategy was formulated, ln the 1987 study, the dominant type of manufacturing business unit was a division or group. However, not all firms are divisionalized. In some firms, business unit considerations are pulled to287

40, No, 3, March 1994

MILLER AND ROTH


A Taxmomxi of /Urim^/zn/Krujy Stratcgit

gether at the plant, or firm level. Table 1 highlights descriptive statistics characterizing the 1987 manufacturing business units. The 1987 sample was pooled from two sources: a longitudinal panei of businesses whose responses were obtained in previous years, and a probability sample of businesses derived from the 1986 Diin'i^ Business Rankings Director}/. This sample was stratified to ensure that all types of manufacturing business would be represented. The longitudinal panel was originally drawn from the 1980 Fortune 500 listing of the largest industrial corporations, supplemented by various other lists of senior manufacturing executives. (See Roth and Miller 1987 for sample details.) In 1987, the response rate for the longitudinal panel was about 40 percent; and from the probability sample, 23 percent. An analysis was performed to determine whether there were any significant differences between the longitudinal panel

Table 1 Industry Mix' Machinery Electronics Consumer Industrial Basic

Respondent Profile Statistics '

27.7% 23.9%
12.8% 21.8% ' 13.8 % 100.0 %

Type of Manufacturing Unit Division/group Entire company Plant Other .* r-'i.' Overall Performance . 1 ' ' ' '

54.3 % 30.3 % .. 13.8% 1.6% 100.0 % :-* . ,ii

Annual sales revenue (median) Net protit (mean % of sales) 1986 unit growth rate (mean) Market share of primary product (mean) Manutacturing Manufacturing costs (mean) Capacity utilization (mean) Direct labor employees (mean) Number cf plants (mean)

$200 million 8.3 % of sales 9.9 % growth 33.4 % share

57.4 % of saies 69,7 % of capacity 2800 people 9.9 plants

* Unless otherwise indicated."%" refers to the percent of the sample respondents. , ' . :

and the probability sample. The hypothesis that the distribution of manufacturing tasks of the two panels was drawn from different populations was easily rejected at the 0.05 level of significance. Of the 195 surveys returned in 1987, 188 useabJe surveys were initially eniployed in this analysis. Those excluded were duplicate responses and substantially incomplete questionnaires. A comparison of the financial and market performance of the respondents with aggregate industry data suggests a nonresponse bias. Respondents were more likely to be market share leaders and financially mt)re successful. 3.2. The Respondents A single respondent from each sample MBU, typically holding the title of Vice President or Director of Manufacturing, completed the survey instrument. Social scientists have long puzzled over the problem of "common method variance" due to the bias associated with using a single informant. Asking a single informant to make complex social judgements nbuut organizational characteristics may increase the subjective propensity of respondents to seek out consistency in their responses and increase random measurement error, It is not clear whether these random error components result from the reporting process, knowledge deficiencies, inadequate measures, or some combination of these and other factors. It is generally concluded that strong assessments of convergent or discriminent validity cannot be made when there is a single informant. However, the cost associated with gaining both participation and consensus from several individuals from large numbers of organizations is very high. Numerical taxonomies are best developed from large samples. Therefore, we have employed data from single respondents while attempting to minimize the extent of common method variance. Research suggests that greater attention to informant selection can help to overcome the common method variance problem when practical considerations require single respondents. Phillips (1981) indicates that "high ranking informants tend to be more reliable sources of information than their lower ranking counterparts . . . and . , , large multidivisional firms may have more well-established competitive intelligence systems than small organizations." The high level of the respondents in the MFP survey and the size of their business units help to moderate the mono-respondent problem.
MANACEMEhJTSciENCE/Vol. 40, No. 3, March 1994

288

MILLER AND ROTH


A Taxoiwiiiu of Miuiufticturin;^ Strategies

We also took several other steps to reduce response bias and measurement error due to mono-informants. The survey instrument was long and variables in each section of the survey were randomly placed, reducing the chance that respondents could cross check for their own internal consistency. Internal and external checks for the reliability and validity of taxons, as described in the next section, were also made. 3.3. The Instrument The questionnaire focused on four broad categories of questions. The first category determined the profile of the company or business unit. The second category addressed the competitive capabilities the respondents planned to pursue, and thus the manufacturing task. In the third section, the respondents were questioned on the performance measures employed in manufacturing, the business unit and in the company overall. The fourth and largest category of questions probed for the key action programs the respondents and their business units intended to invest in over the ensuing two years. This fourth category of questions attempted to tap the pattern of choices in the manufacturing strategy. The 1987 survey instrument consisted of over 100 questions and over 300 separate data elements. Some of the questions relating to contextual data sought objective answers, such as the percent market share or profit margin. The majority, including those which sought to determine the relative importance of the eleven competitive capabilities and those indicating the importance of various action programs or performance measures, required subjective estimates. The following list describes the 11 competitive capabilities delineated in the survey that were used as taxons in this research. Manufacturing executives in the sample were asked to rate each competitive capability measure separately on seven-point, self-anchoring scales. They indicated the relative importance attributed to each capability that the manufacturing firm chose to emphasize in appealing to customers and competing in the marketplace, where "1 = not important" and "7 = critically important." Thefirsteight capabilities clearly have an impact on manufacturing, and are similar to those suggested by Buff a (1984), Skinner (1969, 1985), and Hayes and Wheelwright (1984). The last three caMANACEMENTSCIENCE/VOI, 40. No. 3, March 1994

pabilities have been added to provide some insights into the relationship between the first eight manufacturing capabilities and marketing capabilities, and because development of these capabilities can have important implications for manufacturing choices.
'
*

Taxons Competitive Capability Low Price Design Flexibility Volume Flexibility Cunformance Performance Speed Dependability After Sale Service Advertising Broad Distribution Broad Line

Defined as: The capability to compete on price. The capability to make rapid design changes and/or introduce new products quickly. The capability to respond tu swings in volume. The capability to offer consistent quality. The capability to provide high performance products. The capability to deliver products quickly. The capability to deliver on time (as promised). The capability to provide after sale service. The capability to advertise and promote the product. The capability to distribute the product broadly. The capability to deliver a broad product line.

The industry pooling problem referred to by Hatten et al. {1978), was avoided by wording the questions so that respondents prioritized their capabilities relative to their particular competitive situation, rather than using industry in general as a reference set. It is important to note that the internal reliabilities of the capability questions in the survey have been shown to exceed 0.90 (Huete and Roth 1987). Other researchers have also reported high levels of measurement reliability and validity using similar competitive capability variables with different populations and using different research methods (Nemetz 1990, Wood et al, 1990). 3.4. Identifying Strategy Types Cluster analysis was employed to identify the manufacturing strategy types from the respondent capability profiles. The SAS AceCLUS (Approximate Covariance Estimation for CLUstering) procedure, a variation of the Art et al. (1982) algorithm, was used to obtain es289

MILLER AND ROTH A Taxononiy of Manufacturing; Strategies

timates of the pooled within-duster covariance matrix of firms, using the competitive capability scores. The pooled within-cluster covariance matrix from AceCLUS was processed by the SAS Fastclus algorithm. Fastclus is particularly well suited for large data sets (i.e., n > 100). The Fastclus procedure uses a variation of the fc-means method for nonhierarchical cluster analysis called nearest centroid sorting. Limitations of the k means procedure include its sensitivity to outliers and the requirement for fairly stable seed points (Milligan 1980). Of the 188 observations entered into the clustering procedure; six were dropped due to missing data; 18 other observations were not assigned to a cluster. An a priori analysis of the frequency distribution of the distances of each point to the nearest cluster seed indicated that these observations would have a distorting impact on cluster definition. One thorny problem with cluster analysis is the determination of the most appropriate number of clusters. Three criteria were employed to determine the final number of clusters to be used in subsequent analysis. First, we were guided by Lehmann's (1979) suggestion that the number of clusters be limited to between n/ 30 to n/60, where n is the sample size. Thus, only models with between three and six clusters were considered. Second, we looked for pronounced increases in the tightness of the clusters, as measured by the R^ and pseudo-F statistic {Milligan and Cooper 1985). The three cluster model provided the best fit. Third, we sought managerial interpretability of the clusters on the defining variables using (a) ANOVA and (b) the Scheffe pairwise comparison tests of mean (centroid) differences (Harrigan 1985). The three cluster model best satisfied these criteria. An overall multivariate test of significance using the Wilks Lambda criterion and the associated F statistic indicated that the null hypothesis that the three clusters are equal across all defining variables could be rejected (p < 0.0001). The three resultant manufacturing strategic groups are described in Table 2 in terms of their respective group centroid (mean) scores and their relative ranking in the set of 11 competitive capability variables. The probability that one or more of the cluster means differed from another is also depicted for each competitive capability. The clusters differed from each other on six

of the eleven variables at the 0.05 level of significance or less. The interpretation of the three manufacturing strategic groups, which we have named "caretakers," "marketeers," and "innovators" respectively, is given below. These interpretations are predicated upon: (a) whether there are significant differences on the cluster means of the competitive capability variables at the 0.05 level or less, and (b) the relative ranking of the importance of a competitive capability within a cluster. It is possible that a high ranking capability within a cluster may actually exhibit a relatively low numerical score. Cluster 1: Caretakers. We label cluster 1 the "caretakers" because their low relative emphasis on the development of competitive capabilities appears to prepare them for the minimum standards for competition. Price was not significantly different in importance among the groups. However, based upon its relative rank, price appears to be the dominant competitive capability for the members of cluster 1. The relative ranks of the timebased competitive capabilities of delivery dependability and speed are high (2nd and 4th respectively), though the ability to meet delivery schedules and the ability to make fast deliveries is also important to the members of other clusters. Conformance quality, while rated significantly below the importance given by both cluster 2 and 3 members, is the third most important competitive priority for the caretakers. On the other hand, cluster 1 members ascribe significantly less importance to after sales service and high performance products. The 18 members of cluster 1 represent 11 percent of the cases in all three strategic groups. Cluster 2: Marketeers. The marketeers distinguish themselves from their cluster 1 and 3 counterparts on several key market oriented competitive capabilities. They seek to obtain broad distribution, to offer broad product lines and to be responsive to changing volume requirements. Top ranked priorities within the marketeer cluster were conformance quality, dependable deliveries, and product performance. On each of these top-ranked competitive capabilities, except for dependability, the magnitude of importance was markedly higher in comparison to the caretakers but were not distinguishable from cluster three scores. Furthermore, marketeers reflected some price consciousness. The

290

MANAGEMENT SCI ENCE/Vol. 40, No. 3, March

1994

MILLER AND ROTH


A Taxonomy of Mamifacluniig Slralegies

Table 2

Competitive Capabilities By Group* Manufacturing Strategic Group Caretakers {n = 13) Marketeers {n = 31) Group 2 Innovators (n = 65) Group 3

Competitive Capabilities Low Price Cluster mean* .,,, Rank** Standard Error"' . Design Flexibility Cluster mean" Rank" Standard Error'" Volume Flexibility Cluster mean* Rank** Standard Error*** Conformance Cluster mean' Rank"* Standard Error""' Pertormance Cluster mean" Rank" Standard Error'" Speed Cluster mean' Rank** Standard Error'"" ,...,. Dependability Cluster mean' Rank" Standard Error""" After Sale Service Cluster mean' Rank"" Standard Error'" Advertising Cluster mean* Rank*' Standard Error'*' Broad Distribution Cluster mean* Rank'* Standard Error*" Broad Line Cluster mean' Rank" Standard Error""'

Group 1

F = Value [p = probability)

6il6 1 0.21 5,11 5 0.35 4.17 9 0,37 5.83 3 0.17 4,17 9 QM 5,55 4 0.25 6,00 2 0,20 2,94 10 0,25 4.55 8 0.35 (2,3) {2,3) (3)

5,70 4 0,13 5,07 9 0,15 5.04 10 0.15 6,78 1 0.05 6,07 3 0.09 MM 0.12 6.11 2 0,09 5.27 8 0.16 4.30 11 0.14 5.68 5 0.13 -V 5.30 7 0,13 {3} (3)

5.46 6 0.18 5.98 4 0,12 4.06 10 '0.16 6,75 1 0,05 6.38 2 6.06 .4O 7 (1.2)

F= 1,68 , p = 0,190

, ..

F= 10,07 p < 0.0001

{3)

(2)

F = 9,42 p < 0,0001

(1)

(1)

F = 32,59 p < 0.0001

(2.3)

(1)

(1)

F - 46,76 p < 0.0001

'

F = 0,584 p = 0,666

f = 1,465 p " 0,234

' am"
(1) 5.52 5 0.17 (1) F = 25,92 p = 0.0001

4 :

F = 0,239 p = 0,787

,m
3,92 11 0.22 4,78 8 0,19 (2) F = 26,12 p < 0,0001

4.7a ,

6 0 33

F = 2,59 p = 0.078

' Represents ttie average degree of importance attactied to each competitive capability by cluster. Importance is measured on a seven point self ancfioring scale, (Interval scale 1 - 7 1 = very unimocrtant and 7 = very important), " The rank order ot importance of this competitive capability within the group. " ' T h e standard error of the estimate of the mean for the group. Note. The numbers in parentheses indicate the group numOers from which this group was significantly different at the 0.05 level as indicated by the Scheffe pairwise comparison procedure. Numbers in bold indicate the highest group centroid tor that measure. Group 1 = caretakers, group 2 = marketeers, group 3 = innovators. The observed f-statistics were derived trom one-way ANOVAs and the p-values are associated with the observed f-statistics.

MANAGEMENT SCIENCE/VOI, 40, No, 3, March 1994

291

MILLER AND ROTH A Taxonatity of Matiufaclitring Strategies

ability to offer low price was fourth in overall importance lo this group. Finally, marketeers were more likely to note after sales service as important than the caretaker group. The marketeers are the largest group, accounting for about 49 percent of the cases. Cluster 3: Innovators. Labelled the "Innovators," cluster 3 members are differentiated by the relative emphasis placed upon their ability to make changes in design and to introduce new products quickly. The innovators share certain characteristics with the marketeers. In both groups, conformance and performance quality hold top ranked spots. Dependability is also important to the innovators. There were no statistically significant differences between the innovators and the marketeers on after sales service. As with the caretakers, the innovators share lower degrees of emphasis on the ability to carry a broad product line and on volume flexibility. In comparison to the other clusters, price is least important to this group. The innovators form the second largest cluster, comprising 40 percent of the cases.

ings can be interpreted like the factor loadings in principle components analysis; i.e., they represent the correlations of the original variables with an underlying, unobserved dimension. For substantive interpretations, the canonical loadings are useful as indicators of which original variables are most correlated with each canonical variate in the spirit of factor analysis (Dillon and Goldstein 1984), The standardized canonical coefficients are analogous to beta weights in regression, and can be used to predict cluster membership. To further verify our results, multiple discriminant classification analysis and cross-validation techniques were adopted. Table 3 contains the results of the canonical discriminant analysis used to investigate the relationship between the 11 taxons and cluster membership. In these analyses, the f statistic for Wilk's Lambda = 0,19 indicated a significant overall multivariate relationship (p < 0,0001), Two significant canonical functions for defining manufacturing strategy clusters resulted (K, - 0.76, p < 0.0001 and K-2'=0,74, ;; < 0.0001), applying Rao's (1973) approximate f-statistics. Taken together, the two canonical functions retain 23 percent
Table 3 Canonical Function Correlation 1 2 Eigenvalue or Root 1,33 1,19 Canonical Correlation-flf 0,76 0,74 Canonical 1 Loadings Function Predictor Set Low Price Design Flexibility Volume Flexibility Contormance Performance Speed Delivery Dependability After Sale Service Advertising/Pro mo Broad Distribution Broad Line Function Results of Canonical Oiscriminant Analysis Significance of Canonical Correlation 0,0001 0,0001 Squared Canonical 0,58 0,55

4. Analysis and Discussion


4.1. Underlying Dimensions The 11 taxons representing the importance of the competitive capabilities succeeded in generating a numerical taxonomy somewhat consistent with that hypothesized by Stobaugh and Telesio (1983). However, the 11 taxonomic variables used to form the manufacturing strategy groups are correlated with each other. Therefore, we adopted multivariate statistical methods to extend our exploration. Multiple group discriminant analysis, with each of the taxonomic groups as criterion variables coded into 3 - 1 ^ 2 dummies and with the 11 competitive priority taxons comprising the predictor set, was performed using the SAS Candisc procedure (canonical discriminant analysis). Canonical discriminant analysis, a more general approach to discriminant analysis, is a dimension-reduction technique related to principal component analysis and canonical correlation (Green 1978, Fomell 1978, Kendall and Stuart 1968, Dillon and Goldstein 1984), With this approach, we obtained standardized estimates for both the canonical structure loadings and for the canonical coefficients. The canonical structure load-

Canonical Coefficients Function Function

1
-0.1639 0,1761 0,0608 0.6980 0.7996 -0.0347 0.1220 0.6529 -00713 -0,0665 -0,0136

2
0,0965 -0.4146 0,4346 0,1343 -0,0631 0,1095 -0,1312 -0,0042 -0,0104 0.6677 0.2390

1
-0.2954 -0,2169 -0,0652 0.6837 0.7960 -0,1002 - 0 1151 0.5788 -02799 -02204 0,2636

2
0,0262 -0.7404 0.4377 0 3944 -0,1270 0,0085 -0,3813 0,0554 0,4903 1.4484 -0.4694

Bold numbers indicate fiigh loadings (weights) in canonical functions 10,401.

292

MANAGEMENT SciENCE/Vol. 40, No, 3, March 1994

MILLER AND ROTH A Taxonomy of Manufacturing Strategies

phasis on specialty distribution channels and design of the variance in the data as indicated by their average flexibility are likely to be assigned to the low end. One communality (Fornell 1978), The canonical correlation interpretation is that the low end competitors are comis a measure of the relative strength of the relationship between the set of competitive capabilities and the peting in specialty markets with a broad range of tastes manufacturing strategy group membership. We em- (hence the need for design flexibility), while those at the high end of this scale are competing in high volume ployed 7(, as the measure of multivariate association (MVA) {Cramer-Nicewander 1979). In the case of ca- mass markets with more stable product lines. Figure 1, derived from the canonical coefficients, nonical discriminant analysis, 7^ can be defined as the characterizes the responding MBUs strategic position average squared multiple correlation which is equivalent against the two manufacturing task constructs of market to the average squared canorucal correlations. Therefore, differentiation and market scope. The numbers on the about 56 percent of the variance in manufacturing strategy group membership is explained by the set of plot, indicate the manufacturing strategy group assignment designated by the clustering procedure (1 = care11 taxons, and vice versa, since the measure is symtakers, 2 ^ marketeers, and 3 = innovators). It seems metric. The redundancy index (Green 1978) was less clear that the caretakers are less likely to have or to appropriate because only a few of the observed variables value differentiating products and services. Therefore, determined the structure matrix (Fornell 1978) and because the criterion is a single categorical variable, i.e., it is not surprising that the caretakers compete mainly on price, and are on the low side of the market differmanufacturing strategy group membership (Cramerentiation scale. The innovators, like the marketeers, are Nicewander 1979), more apt to give weight to product attributes, and thus Our interpretation of the underlying manufacturing appear on the high end of the market differentiation strategy dimensions is based upon canonical variate analysis of the standardized canonical loadings. The canonical loadings are also helpful in labelling the derived Figure 1 Plot of Respondent Business Units and Group Centroids on canonical variates of the predictor set. Using these, we Canonical Functions interpret canonical function 1 to be a "market differLetters in circles indicate cluster centroids. The letters are: Caretakers ( C ) , entiation" dimension that depicts the perceived reMarketeers ( M ) , and Innovators ( I ) , quirement of the manufacturing firm to distinguish itself from competitors by attributes of its products and services. Here the largest correlates of cluster membership have to do with the relative importance given to product performance, conformance quality and after-sales service. The relatively large canonical coefficients for performance, conformance and service suggest that a firm placing a greater priority on these capabilities will fall at the high end of the market differentiation dimension; and those with lower emphasis will be less likely to differentiate products/services on these attributes. We call function 2 "market scope." This canonical variate reflects the magnitude of the customer base served by the business unit. Broad distribution and volume flexibility are positively correlated with function 2, while design flexibility is negatively associated. The strong positive coefficients for broad distribution and volume flexibility imply that a firm placing a high emFUMOtOHl MariiM Dtrre phasis on these capabilities will be assigned to the high end of the spectrum. Those manufacturers with an emKey: 1 = Caretakers; 2 = Marketeers: 3 = Innovators
->, I.M

M A N A G E M E N T S C I E N C E / V O I , 40, No, 3, March 1994

293

MILLER AND ROTH A Taxonomy of Mamifacturing, Strategies

scale. However, unlike the marketeers who as a group have very broad distribution channels, and thus are on the high end of the market scope scale, the innovators emphasize responsiveness to product change through design flexibility, which is associated with the low end of the market scope scale. These findings suggest that in general, the manufacturing task is a multivariate, multidimensional construct that reflects the needs of widely different market environments and relative market positions. They also suggest that these contextual factors represent deeper and more fundamental drivers of firm behavior than cost, technology, or markets as proposed by Stobaugh and Telesio (1983). Here, we see that a cost (price) emphasis is associated with a lack of differentiation, that technological change in terms of flexibility to introduce new products is associated with specialty markets, and that marketeer oriented behavior is associated with broad distribution and product and service differentiation. On the other hand, the market structure orientation of these underlying constructs reinforces the views of Skinner (1978) and Hill (1989) who relate the manufacturing task to fundamental market positioning (and order winning) strategies. The factors are also closely allied with those proposed by Porter (1980), He categorized firms by their propensity to differentiate themselves or to compete on price, and by the degree to which they focus their efforts on particular markets or products, 4.2. Statistical Cross-validation To determine the stability of the estimates, a crossloading analysis was first conducted over the canonical structure variates. The cross-loadings on the original defining variables appeared stable as shown in Table 3. Each of these loadings exceeded |0,30|, giving us no reason to change our original interpretation. Next, the performance of the discriminant taxons was evaluated by estimating the probability of misclassification in the classification of future observations through cross-validation procedures. We could not employ split-sample techniques due to the small size of cluster 1. Using a jackknife procedure, a discriminant function with the 11 taxons as predictors was computed for u - I out of n cases, and that function was used to classify the one

Table 4

Number of Observations and Percent Cross Validated 1 Caretakers Marketeers Itjnovators Total

To/From Cluster

1 ? 3

13(72%) 0(0%) 0(0%)

3(17%) 80(99%) 11(17%)

2(11%) 1(1%) 54(83%)

18(100%) 81(100%) 65(100%)

Error Rates from: Cross-validation Posterior Probabilities

0,28 0,34

001 0,05

0,17 0,28

0,10 0,18

observation held out (See Table 4). This process was repeated for each of the cases, and the proportion of hold-out observations assigned to each group was determined (Lachenbruch and Mickey 1968), Cross-validation error rates reflect the percentage of hold-out cases misclassified. Another set of error-rate estimates were computed based upon the posterior probability estimates based on cross-validation procedures described above. The cross-validation posterior prtibability estimators have been shown to possess both low bias and low variance (Hora and Wilcox 1982). The crossvalidation results are as follows. Cross-validation classification analysis suggests that the overall discrimination power of the taxons is quite good, with 72 percent of cluster 1, 99 percent of cluster 2, and 83 percent of cluster 3 correctly classified. The taxons do a superior job of predicting actual membership in cluster 2, but are more prone to misclassify members of clusters 1 and 3, 4.3. Industrial Mix Table 5 illustrates the relationships between group membership and broad categories of industry membership, where industr)' is described in terms of five groups developed for the Manufacturing Futures Survey based upon aggregation of three digit Standard Industrial ClassificaHon (SIC) codes (See Miller and Vollmann 1984 and Roth and Miller 1987). A chi-square test indicates that cluster membership is associated with the industry in which the MBU resides (;J < 0.01). For example, the electronics (computer, instrument, and electronic equipment manufacturers) and machinery (machine tools, transportation equipment, machinery

M A N A G E M E N T S C I E N C E / V O I . 40, No. 3, March

1994

MILLER AND ROTH


A Taxonomy of Manufacluri/ig Strategies

Table 5

4.4. Context After developing the clusters with the procedures described above, the factors associated with strategic group Frequency/% Marketeers Innovators Caretakers membership were systematically examined to fulfill the second purpose of the study. Using ANOVA and 11 3 20 6 Consumers Scheffe post hoc pairwise comparisons of group mean 12% 37 7 17 13 differences, industry, environmental, and contextual Industrial 23% variables, as well as manufacturing strategy choices and 4 23 3 16 Basic performance measures, were explored to see if they 14% showed significant differences across clusters. In making 46 20 25 Machinery 1 statements about relationships, we have designated an 28% a priori level of statistical significance of 0,05 or less. 20 38 1 17 Electronic 23% Table 6 describes how each cluster of manufacturers differs across the context variables. This table is typical 164 65 81 Frequency 18 of several that follow. With the exception of the Scheffe 40% 100% 11% 49% Percent multiple comparisons tests, no adjustments were made for the problem of simultaneously testing multiple hyx ' = 24,78 d,f, = 8 p < 0 , 0 1 , potheses on the same data set due to the exploratory groups) industries are more likely to populated by in- nature of our inquiry. Thus, the true ;?-value may actually be much greater than the nominal level shown novators. Conversely, the industrial (parts, components, in the tables, and future research is required to confirm intermediate goods producers), basic (chemical, paper, the associations found. primary metals firms) and consumer packaged goods In Table 6 we see evidence that the clusters tend to (foods, cosmetics, pharmaceutical manufacturers) are more likely to be marketeers. The caretakers are most differ both in terms of the stage of their product life cycle, and by the degree of product standardization. likely to be found among the industrial goods and conThe caretakers derive significantly more revenues from sumer packaged goods firms. "maturity stage" sales, as a percent of total sales, than Despite these broad industry associations with maneither of their counterparts. The marketeers lie in beufacturing strategic group membership, examination at tween the caretakers and innovators in terms of both a more detailed level showed that it was common for growth and maturity stage revenues. The innovators at least one competitor in a particular three digit SIC have the least standardized (most customized) products industry to compete on a substantially different basis than its primary competitors. We could identify 17 in- while the caretaker's products are most heavily standardized. stances in which two or more of the respondents to the As might be expected the innovators invest more survey were competing in the same industry. In 12 of heavily in R&D than their counterparts, especially the these instances, at least one competitor was in a different caretakers. They were also more apt to report a stronger manufacturing strategic group from the other(s). For example, in the micro computer industry, three of the functional influence by Engineering/R&D in setting the long run goals and strategies of the business unit. Incompetitors in the sample were innovators, but two respondents were clearly classified as marketeers. In an- novators place more emphasis on increasing market other industry, three competitors each occupied a dif- share by developing new products for both old and new ferent strategic position; one was a caretaker, another markets. Marketeers may develop new products for existing markets but they are less apt to emphasize ena marketeer, and the third an innovator. These examples tering new markets with new products. are consistent with findings from other strategic group Further scrutiny of the data shows some evidence research that demonstrates that a broad range of stratthat the clusters exhibit the "product-process" matrix egies is available to competitors within an industry (for characteristics suggested by Hayes and Wheelwright example. Porter 1980).
indusUy Representation By Strategic Groups (Number of Respondents)

MANAGEMENT SciENCE/Vol, 40, No. 3, March 1994

295

MILLER AND ROTH A Taxonomy of Manufacturing Strategies

Table 6

Strategy and Context By Group


Manufacturing Strategic Group Caretakers {n - 18) Marketeers [n = 81) Group 2 Innovators (n = 66) Group 3

II

Context/Strategy Variable
R&D/Sales % Cluster mean" Standard Error" Export Sales % Cluster mean' Standard Error" Product Standardization Cluster mean*** Standard Error** Maturity Stage Revenues % Cluster mean* Standard Error** T Engineerifjg/R&D Influence Cluster m e a n " ' . . . , . . . . . . ,, Standard Error" Increase Mkt. Share ,,

F=

Value

Group 1

probability)

2,41 0,86

(3)

3,88 0,40

4.97 0.60

(1)

F = 3,06 P = 0,051

1.91 0,82

(2,3)

14.89 2,04

(1)

14,88 2,52

(1)

F = 4 24 P = 0.017

4,44 0.41

(2,3)

4,15 0,19

(1)

3,30 0,21

(1)

= 5,66

P- c 0.004

73.44 6,18

(3)

57,59 2.84

, ,

55,15 3,04 . . , 5.31 0.15

(1)

F-= 3,91
- 0,022

3,83 0,37

(3)
1

4.61

(1)

(1.2)

F = 10.92 p < 0.0001

O.t5
k

Cluster m e a n ' " . .,. Standard E r r o r "


New Products/Old Mkts. Cluster m e a n ' " Standard Error*' 1 : _ New Products/New Mkts. Cluster mean*** Standard Error**

5,00 0,32

(3J

5.58 ' 0.1?


5.75 0.11

5.78 0,14

(1)

= 3,10

P- -- 0 048

4,72 0,27

(2,3)

(1)

5,39 0.16

(1)

7 6 4

1 ,

P =- 0,001
*'

3,72 0,44

(3)

4,43 0,19

4.95 0,22

(1)

F -. 4,06 P = 0,019
1

* Represents the mean value of the context measure reported by each strategic group, ** The standard error of the estimate of the mean for each group. " * Represents the average values measured on seven point self anchoring scales (interval scaie 1-7). See Roth and Miller 1987 (or a detailed definition of the variables, ' ' '

Note Ttie numbers m parentheses indicate the group numbers (rom which this group was significantly different at the 0,05 level as mdicaled by the Scheffe pairwise comparison procedure. Numbers in bold indicate the highest group centroid for that measure. The observed f-statistics were derived from one-way ANOVAs and the p-values are associated with each of the observed f-statistics.

(1984). We asked the executives to rate the degree to which the business unit's products are standardized, on a scale from "1 = highly customized" to "7 = highly standardized," The product/process theory would
2% "

suggest that the lower the standardization score, the more likely the process is characterized as a job shop. The higher the score, the more likely the process is continuous. The caretakers display a strong proclivity for
M A N A G E M E N T SciENCE/VoI. 40, N o . 3, March 1994

MILLER AND ROTH


A Taxonomy of ManiifacturiDg Strategic?

producing more standardized products, in contrast to the innovators and marketeers. Weaker differences in the degree of product customization were reported by the innovators and marketeers. The mean values for these two groups indicate greater customization for the innovators and higher standardization for the marketeers. We asked a question concerning the dominant production process of the manufacturing unit, v^'here a score of " 1 " indicates a continuous flow process and a score of "5," a job shop with no dominant flows. While the evidence is weaker {p = 0.08), the data suggest the predicted pattern. The innovators are at the near end of the job shop continuum, while the caretakers are at the opposite extreme (more continuous processes); the marketeers lie in between. 4.5. Action Programs In the survey, respondents were given a list of 36 key action programs to improve the effectiveness of their operations over the following two years. (See Miller and Roth i988 and Roth and Miller i 987 for a discussion and complete list of the action programs in the survey.) The manufacturing executives were asked to rate from "1 = unimportant" to "7 = very important" the degree of emphasis that would be placed on each action program over the next two years. The action programs indicate the intended emphasis on the respondent manufacturing choices. Roth et al. (1989), Roth and Miller (1990, 1992), and Ward etal. (1988) have shown that these action plans tap important underlying structural and infrastructural directions in a manufacturing strategy. In testing the common theme In manufacturing strategy that actions (choices) should be congruent with the manufacturing task, ten strategy choice variables were found to be significant at the 0.05 level in our data. The innovators' manufacturing strategy choices place significantly more emphasis on programs that promise to shorten total product cycle times. Innovators also focus on computer aided design (CAD), and emphasize developing new processes for their new products. And, innovators plan to embark upon manufacturing programs that reduce their manufacturing lead-times. This pattern of choices appears to be congruent with a manufacturing task that prizes the ability to change designs and introduce new products.
MANAGEMENT SciENCE/Vol. 40, No. 3, March 1994

Table 7 suggests that the marketeers plan on strengthening their manufacturing operations through infrastructural changes, particularly those that will cut costs and improve quality. They intend to emphasize changing labor/management culture and streamlining their workforces through headcount reductions and plant closings. Besides reducing the size of the workforce and improving productivity, the marketeers hope to attack quality problems through zero defect programs and statistical process control. These infrastructural changes appear to be congruent with their more standardized products and continuous processes. They are also consistent with the relatively high priority they attached to the quality, dependability and price capabilities. Notably, the caretakers, who are competing first on price, tend to place relatively lower emphasis than their counterparts on each improvement program measured. Recall that this group had the most mature products. At worst, it appears that the caretakers, as a group, have no manufacturing strategy and are not renewing their manufacturing function. At best, their manufacturing strategy is passive. Does the caretaker group represent the end of the life cycle? The data suggest that this may be so. However, there are several among the caretaker group that have high profits, market shares and economic returns. Further work must be done to examine how a firm can use manufacturing to be successful in mature markets.

4.6. Measures Measuring manufacturing performance is becoming an area of concern for U.S. manufacturers. In particular, there is a need to understand how nonfinancial measures of performance are viewed in the manufacturing unit. Nannietal. (1988), and McDougall (1988) have called for performance measures relating to the business strategy and the key improvement programs in manufacturing, Richardson et al. (1985) have suggested that measures should correspond specifically to the strategic capabilities which define the manufacturing task. Roth (1989) has shown that the content of a manufacturing strategy is correlated with manufacturing perfonnance measures that have a business unit impact. From a list of 29 manufacturing performance indicators, the respondents to the 1987 Manufacturing Futures Survey were asked to indicate on a self-anchoring
297

MILLER AND ROTH A Taxonomy of Maiiufactunug Strategies

Table 7

Future Improvement Programs By Group Manufacturing Strategic Group Caretakers ("=18) P^Qfl^^^s Group 1 ' 4.28 0.37 . ' 4.28 0.39 (2,3) 5.43 0.17 (1) i;1 d.1 ' 5.35 0.1S
I-

Marketeers (/)-81) Group 2

Innovators (/I = 65) Group 3 " '" (3) 479 0.22 . (jj i ^ = 5 53 p = 0,005 F - Value (^^probability)

Lat)or/Mgmt Relationships Cluster mean- .,, ,.,. .,... Standard Error" Zero Defects Ciuster meanStandard Error" MFG Lead Time Reduction Cluster meanStandard Error" C40 Cluster mean' Standard Error" . , , ., ,. 3.50 0.38 4.33 0.36

(1)

f = 4 35 p = 0.014

(3j

5,16 0,16

5.S6 0.15

(1)

f=585 p ^ 0,004

(2.3)

4,99 0^0

(1)

5.14 0.21

(1)

f=656 p = 0,002

New Process/New Product Cluster meanStandard Error" Closing Plants Ciuster mean' Standard Error" . . . . . . . . SPC (Process) Cluster mean- . . . . . . i..v Standard Error" SPC {Product) Cluster mean* Standard Error" New Product Introductions Cluster meanStandard Error" Reducing Workforce Size Ciuster mean- , . . . . . . . . , 3,83 4.83 (3) 3,89 (2) f = 6 61 Standard Error" ;. 0,47 0,20 0,22 p = 0,004 Represents the average degree of importance attached to each program. Importance is measured in a seven point self anchoring scaie where 1 = very unimportant and 7 = very important, " The standard error of the estimate of the mean for each group. Note. The numbers in parentheses indicate the group numbers from which this group was significantly different at the 0,05 ievei as indicated by the Scheffe pairwise comparison procedure. Numbers in bold indicate the tiighesi group centroid lor that measure. The observed f-statistics were derived from one-way ANOVAs and the p-vaiues are associated with each of the observed f-statistics. 376 0,44 (2,3) 5,06 o,2O (1) 5.42 0,20 (1) f = 6 56 p = 0.002 4,39 0,36 (2) 5.43 0,16 (i) 4,89 0 20 f = 4 53 p = o'oi2 4,61 0,36 2.00 0.33 . (2) 5.79 0,15 (1,3) ' 2.91 0,15 (3) 211 0.20 "' 5,11 0,20 (2) f=654 p = 0.002 (2) f=386 p = 0.023 4.0$ 0,* ,' ' tm 0.18 5.05 0,i8 ' f = 316 p = 0,045

298

M A N A G E M E N T SciENCE/Vol, 4 0 , N o , 3, M a r c h 1994

MILLER AND ROTH


A Taxouottiy of Mafnifiicturiiig Stratef^ics

Table 8

Pertormance Measures By Group .


Caretakers {n - 18) Manufacturing Strategic Group Marketeers {n - 81) Group 2 Innovators [n = 85) Group 3 F = Value (p = probabiiity)

Performance Measures
Manufacturing Lead Time Cluster mean' Standard Error" Changeover Setup Time Cluster mean' Standard Error" Headcount Cluster mean' Standard Error" Ratio White/Blue Collar Cluster mean* Standard Error" Outgoing Quality ' , Cluster mean' Standard Error" Number of Grievances Ciuster mean" Standard Error" % New Products Ontime Cluster mean' Standard Error"

Group 1

4.88 0,30

(3)

5,51 0,12

5.63 0,13

(1)
p = 0,039

4,58 0.35

5,35 0,15

4.83 0,20 p = 0,039

4,83 0.41

5,54 0,12

521 0,14

F = 3,39 p - 0.036

4.44 0.30

4.94 0,15

(3)

4,37 0.18

(2)

F - 3.32 p - 0,039

5,83 0.29

(2, 3)

6.63 0,06

(1)

6.78 0.06

(1)

f = 15,38 p = 0,0001

'
)

3,72 0,29

4.42 0.21

(3)

3,61 0.06

(2)

f = 5,38 p = 0,006

4,33 0,23

(2, 3)

5,24 0,14

(1)

5.64 0.15

(1)

F = 8,39 p = 0.0003

* Represents the average degree of importance attached to each measure, importance is measured on a seven point self anchoring scaie where 1 = very unimportant and 7 = very important, " The standard error of the estimate of the mean for each group. Note. The numbers in parentheses indicate the group numbers from which this group was significantly different at the 0,05 ievei as indicated by the Scheffe pairwise comparison procedure. Numbers in bold indicate the highest group centroid for that measure. The observed f-statistics were derived from one-way ANOVAs and the p-values are associated with each of the observed f-statistics.

scale ranging from 1 to 7, the relative degree of importance the manufacturing management team placed on each measure, (See Roth and Miller 1987 for a complete list: of performance measures.) Eight performance indicators separated the strategy groups at the .05 level and are shown in Table 8. The innovators differ from their counterparts by attaching more importance to performance measured by the percent of new products/models introduced on
M A N A G E M E N T S C I E N C E / V O I . 40, No. 3, March 1994

time. They are less apt to be concerned with personnel problem indicators, such as the ratio of white collar to blue collar workers, headcounts, or the number of grievances. Personnel measures are more important to the marketeers as are changeover and setup times, and work in process inventories. Innovators place about the same relative degree of importance as the marketeers on manufacturing lead times, and the level of outgoing quality. Interestingly, the caretakers tend to give lower
299

MILLER AND ROTH


A Taxonomy of Manufacturing Strategies

scope. Innovator firms may not be simply "driven" by technology, per se, as suggested by a univariate perspective. There appears to be an interplay between focusing on markets where technology offers a greater opportunity to differentiate and on markets whose nar5. Summary and Conclusions There is consensus among researchers ab.out the key row and more specialized customer base creates pressure to change products frequently. More research, however elements of a manufacturing strategy. The common is required to determine whether technology drives intheme that permeates the growing literature is that the novator firms into niche markets or whether the remanufacturing task, as measured by the importance quirements of specialized markets propels the firms togiven to competitive capabilities like quality, flexibility, ward technology. The marketeers may be motivated to delivery, cost, must be linked to manufacturing strategy maintain market share through their broad distribution choices, and that both should link to the business stratsystems, and respond to demand opportunities with egy. This research has identified three types of manutheir volumefiexibility.The lack of differentiation could facturing strategies, and has empirically addressed how be behind the caretaker's primary focus on price. this common theme is played out in the choices and Fourth, we have developed evidence that there is a business strategies of each, relationship between broadly defined industry types and the manufacturing task. However, at a more detailed 5.1. Manufacturing Taxonomy level, we see that this effect does not preclude variations Three distinct types of manufacturers can be identified in the manufacturing task between the firms competing by the importance they place on competitive capabilities; in the same industry. caretakers, marketeers, and innovators. Our taxonomy The taxonomy developed here has much in common provides analytical evidence that generally supports the with other taxonomies developed with entirely different typology hypothesized by Stobaugh and Telesio (1983), taxons and by researchers from substantially different but it is different in some significant and intriguing ways. disciplines. The similarity to Miles and Snow's (1978) Our marketeers and innovators are similar to their martaxonomy is apparent, though ours yields fewer cateket and technology-driven groups. However, there is a gories. Clearly, their "prospectors," "differentiators" substantial difference in their conception of a "costand "defenders," are similar to the innovators, mardriven" group and our caretakers. While both the costketers, and caretakers here. Their "focusers" and "redriven firms and caretakers emphasize price competiactors" are likely to be variants of the large mass martion, their cost-drivenfirmsare assumed to have adapted keter and innovator categories. To the extent that their this as part of a coordinated market, manufacturing and taxonomy identifies the same general categories of strabusiness strategy. Our caretakers are notable for the tegic behavior, this research provides evidence on how low levels of importance they ascribe to manufacturing manufacturing strategies can be linked to the business capabilities and choices, and for their seeming lack of unit. Similarly, we see a reflection of the industrial orcongruence between these capabilities and choices. ganization oriented taxonomies of Porter (1980) and Second, we have shown that the underlying structure Kim and Um (1988), the product process theories of of the manufacturing task consists of multivariate conHayes and Wheelwright (1984a), and the technology structs. Combinations of capabilities provide better intheory oriented taxonomies of Hambrick (1983) and sight into the manufacturing task than statements about Horwitch and Thietart (1987), in this taxonomy of individual capabilities. manufacturing strategies. These similarities suggest the Third, we also found that more than one multivariate strength of the underlying competitive factors which dimension is necessary to determine the manufacturing seem to explain much of industrial behavior. They also task. The differences between categories of strategy were suggest that the further development of manufacturing shown to be related to two underlying market characstrategy research will be enhanced by synthesizing teristics: the degree of market differentiation and market findings and approaches from a number of disciplines.
300

ratings overall to each manufacturing performance indicator. This is further evidence that the caretakers are playing out an endgame in the product Hfe cycle.

MANAGEMENT SciENCE/Vol. 40, No, 3, March 1994

MILLER AND ROTH


A TaxQitonni of Mainifacturnt^ Sfruffjjies

5.2. Consistency of Purpose Going beyond the questions examined in previous research, we used the taxonomy to explore the central theme in the manufacturing strategy literature, that is, how major competitive orientations are related to the manufacturing task as revealed through key capabilities and context, manufacturing strategy choices, and manufacturing performance measures. We have shown in this paper that significant differences exist in certain capabilities, actions, and measures among the strategy groups. Moreover, these relationships are generally in the directions one would anticipate from the underlying theory of life cycles, product / process hfe cycles, and other existing theories about production. However, we cannot conclude that linkage between manufacturing task, choice measures and business unit strategy are a universal phenomenon. Our research supports the notion that a manufacturing strategy is related to product life cycles. The innovators, with relatively short product lives, demonstrate the characteristics of start-up firms in product life cycle theory. For example, the manufacturing strategy of these firms is heavily influenced by engineering and research and development functions, These firms plan frequent modification of the business unit's production processes and new product introduction capabilities with a special emphasis on reduced manufacturing and product development lead times (Utterback and Abernathy 1975), The marketeers with well established products and markets, pursue manufacturing strategies characteristic of a business in more mature phases of the life cycle. Marketeer manufacturing strategy choices are oriented towards improving the reliability of the manufacturing process. Quality control programs were especially prevalent, as are attempts to reduce the size of the workforce. The key performance measures of the marketeers are congruent with their manufacturing strategy as reflected in their focus on productivity and quality issues, The caretakers, on average, demonstrate the characteristics of businesses in the declining stages of the life cycle. The fact that this group has no clear cut pattern of action programs or performance measures related to the manufacturing task violates the assumption of the one common theme in manufacturing strategy research. There are several plausible explanations. One is that the sample size for this group (the smallest) is simply

too small to yield a reliable result. The second is that the preoccupation of these firms with price competition in mature/declining markets leads them to think short term, and to merely react to the current situation rather than to take any coherent strategic steps for renewal. Further research will be required to determine which, if any, of the above reasons is closest to the truth, and to understand how the best performers in this group achieve success. This research will require a larger sample of caretaker firms to yield significant results. 5.3. Future Research This study has several limitations which future manufacturing strategy researchers should consider. First, as noted above, the cases examined represent a biased sample of U.S, industry. As a result, the proportion of firms in the caretaker group is probably underestimated, prohibiting any attempt to understand the distinction between subgroups in it. Future research should attempt to obtain a broader representation of caretaker businesses. Another limitation is the correlated error problem that derives from the use of one respondent for all of the data gathered from each firm. Future lines of inquiry, should consider using multiple sources of information and methods to reduce this problem. The study provides important clues for better understanding the relationships between manufacturing strategy and performance. Nevertheless, a critical shortcoming is that it provides no visibility into the causal relationships between manufacturing strategy and performance outcomes. For example, what factors characterize good and bad marketeers or innovators? Future research pertaining to the causal linkages between manufacturing task, manufacturing choices and performance outcomes is best studied via a longitudinal study, and cannot be ascertained from a cross-sectional study without significant prior theory as guidance. Finally, an important line of future research is to test the stability of this taxonomy globally, and over time. The "laws" of business that seem to define the strategic positioning and competitive behavior in this and other taxonomies have all been based on observations of the business environment during a short 20 year span. If we suppose that intelligent competitors will use their knowledge of the "normal" rules of battle to better develop new principles of competitive warfare, then we

MANAGEMENT SciENCE/Vol, 40, No- 3, March 1994

301

MILLER AND ROTH A Taxonomy of Manufacturing Strategies

may anticipate that new manufacturing strategic groups will be formed over time, and in different parts of the world.* ,^^ ,
' The authors wish to thank two anonymous reviewers for their Ihoughtful reviews of this paper.

Green, P. E,, Analyzing Multivariate Data, Dryden Press, Hinsdale,. IL, 1978. Hambrick. D, C . "Some Tests of the Effectiveness and Functional Attributes of Miles and Snow's Strategic Types," Academy of Management}.. 1 (1983a), 5-25, , "An Empirical Typology of Mature Industrial- Product Environment," ^fdiit'my I'/Manayciru'ii//,, 2 (1983b), 213-229. and D- Lei, "Toward and Empirical Prioritizahon of Conhngency Variables For Business Strategy," Academy of Management}., 28, 4(1985), 763-788Hamel, G, and C. K- Prahalad, 'Strategic Inleni," Harvard Business Revieiv, i}7. 3 (May-June 1989), 64-68Hdrrigan, K, R., "An Application of Clustering for Strategic Group Analysis," Strategic Management } . , 10 (1985). 55-73, Hartigan, J, A,, Clustering Algorithms. John Wiley and Sons, New York, 1975, , "Statistical Theory in Cluster," /- Classtfiiation. 2 (1983). 6 3 76Hatten. K, J . D, E, Schendel and A, C, Cwper. "A Strategic Model of the US, Brewing Industry: 1952-1972," Academy of Management } . , 4(1978), 592-609, Hax, A- C- and N- S. Majluf, "The Corporate Strategic Planning Process," Interfaces. 14 (1984), 47-60. and . "The Concept of Strategy and Strategy Formation Process," Interfaces. 18 (May-June 1988), 99-109. Hayes, D. R, and S. C. Wheelwright, Restoring Our Competitive Edge. John Wiley and Sons. New York, 1984. Hayes, R, H,, S, C, Wheelwright and Kim Clark, Dytiamic Manufacturing. Free Press, New York, 1988, Hill, T,. Manufacturing Strategy: Text and Cases, R, D, lrwin. Homewood, IL, 1989, Hofer. C, W,. "Towards a Contingency Theory of Strategy," Academy of Management I.. 18. 4 (1975). 784-810, and D, Schendel. Strategy Formulation: Analytical Concepts, West Publishing, New York. 1978. Hora, S, C, and J B, Wilcox, "Estimation of Error Rates in Several Population Discriminant Analysis." /- Marketing Research, XIX (1982), 57-61, Horwitch, M, and R, A, Thietart, "The Effect of Business Interdependencies on Product R&D-Intensive Business Performance," Mattagenient Sci.. 2 (1987), 178-197, Huete, L and A. V, Roth, "Linking Manufacturing Capabilities with SBU Strategic Directions." Proceedings of the 1987 Decision Sciences Institute. 1987, Kendall, M. G, and A. Stuart, The Advanced Theory of Statistics. 3, Hafner Publishing, New York, 1968, Kim, L. and Y, Lim. "Environment, Generic Strategies, and Performance in a Rapidly Developing Country: A Taxonomic Approach," Academy of Management /,, 31, 4 (1988), 802-827, Lachenbruch, P. A. and M, A. Mickey. "Estimation of Error Rates in Discriminant Analysis," ri'f/ifi(iHit'/)(,-s, 10(1968). 1-10. Lawrence, P, R- and J, W, Lorsch. Organization and Environment, lrwin, Homewood, IL, 1969,

References '

- '

Anderson, J. C . G, Cleveland and R, G, Schroeder, "Operations Strategy: A Literature Review." /. Operations Management. 8 (April 1989), 133-158. Ansoff, H, 1,, Corporate Strategu: An Anah/tical Approach to Business Policy for Growth ami E.-(pansion. McGraw-Hill. N'ew York, 1965Art, D.. R, Gnanadesikan and R. Kettenring. "Data-based Metrics for Cluster Analysis," Utilitas Mafhematica, 21A (1982), 75-99. Buffa, E,, Meeting the Competitive Challenge, Dow Jones-lrwin, Homewood, IL 1984, Cohen, Morris A, and 11 L, Lee, Manufacturing Strategy: Concepts and Methods," In Paul R. Kleindorfer (Ed.), The Management of Productivity and Technology in Manufacturing. Plenum Press, New York, 1985, Cool K, O, and D, Schendel, 'Strategic Group Formation and Performance: The Case of the U.S, Phannaceutical Industry, 1963\982," Management Sci.. 9 (1987), 1102-1124, Cramer, E. and W, A, Nicewander, Some Symmetric Invariant Measures of Multivariate Association," Psifchometrica. 44, 1 (March 1979), 43-54. DeMeyer, A., ], G, Miller. J. Nakane and K. Ferdows, "Flexibility: The Next Competitive Battle," Strategic Management j . . 10, 2 (1989). 135-144, Dess, G, G, and P. S, Davis, "Porter's (1980) Generic Strategies as Determinants of Strategic Group Membership and Organizational Performance,"/4c(J(ieH)i/ of Management /., 27 (1984). 467-488. Dillon, W, R. and M, Goldstein, Muitivariate Analysis: Methods atid Applications. John Wiley and Sons, Now York, 1984 Duncan, R, B., "Characteristics of Organizational Environments and Perceived Environmental Uncertainty," Administrative Sci. Quarterly, 17(1972). 313-327, Fahey. L, and H, K. Christensen. "Evaluating the Research on Strategy Context." /, Matia^enifiit. 12. 2 (1986). 167-183, Ferdows, K,, J, G, Miller, J. Nakane and T, Vollmann, "Evolving Manufacturing Strategies," liUentaiional}. Operations and Production Management. (January 1987), Fine, C. H, and A, C, Hax. "ManufacturingSh-ategy: A Methodology and an Illusion." Interfaces. 15, 6 (1985), 28-46Fomell, C , "Three Approaches to Canonical Analysis," /. Marketing Research Society. 20, 3 (1978), 166-181, Galbraith. J,, Designing Complex Organizations. Addison Wesley, Reading, MA, 1973, Giffi, C , A. V, Roth and G, Seal, Competing in World Class Manufacturing: America's 21s/ Century Challenge, Business One lrwin, Homewood, IL, 1990, Ginsberg, A, and N, Venkatraman, "Contingency Perspectives of Organizational Strategy: A Critical Review of the Empirical Research," Academy of Management Review, 10 ( 1985), 421-434.

302

MANAGEMENT SciENCE/VoI, 40, No. 3, March 1994

MILLER AND ROTH A Taxonomy of Manufacturing Strategies

Lehmann, D, R., Market Research and Analysis, lrwin, Homewood, IL, 1979, McDougall, D,, "Learning (he Ropes: How to Tell When You've Found an Effective Performance Measurement System for Manufacturing," Operations Management Review, 6, 1 (1988), 38-48, McGee, J, and H, Thomas. "Strategic Groups: Theory. Research and Taxonomy," Strategic Management /,, 7 (1986), 141-160, Miles, R. E. and C. C. Snow, Organizational Strategy. Structure, and Process, McGraw-Hill. New York, 1978, Miller, D, and P, H, Friesen, "Strategy Making in Context: Ten Empirical Archetypes," /, Management Studies, 14 (1977), 253-280, and, "Porter's (1980) Generic Strategies and Performance: An Empirical Examination with American Data (Part I ) , " Organizational Studies, 7 (1986a). 37-55, -and . "Porter's (1980) Generic Strategies and Performance: An Empirical Examination with American Data (Part II)." Organizational Studies, 7 (1986b), 255-261. Miller, J, G,, "Report on the 1982 Manufacturing Futures Survey," Manufacturing Roundtabie Research Report Series. Boston University. Boston, MA, 1982, , A, Amano, A, DeMeyer, K. Ferdows, J, Nakane and A. Roth. "Closing the Competitive Gaps," In K, Ferdows (Ed.). International Manufacturing. North Holland. Amsterdam, 1989, 153168. , A. DeMeyer and J, Nakane, Benchmarking Global Manufacturing. Business One-Irwin, Homewood, !L, 1993. and A, V, Roth, "Manufacturing Strategies," Operations Management Review, 6, 1 (1988). 8-20, Miller. J, and T, Vollmann, Manufacturing Futures Sumey Report 1984, Boston University Manufacturing Roundtabie Research Report, 1984, Miller, S, and D, Rogers, Manufacturing Policy, irwin, Homewood, IL, 1956, Milligan, G, W,. "An Examination of the Effect of Six Types of Error Perturbation on Fifteen Clustering Algorithms." Psychonictrika, 45(1980), 141-160, and M, C, Cooper. "An Examination of Procedures for Determining the Number of Clusters in a Data Set." Psi/chometrika, 2 (1985), 159-179. Mintzberg, H,, "Patterns in Strategy Formulation," Management Sci., 9 (May 1978), 934-948, Nanni. A,. ]. G, Miller and T. E. Vollmann, "What Shall We Account For?" Management Accounting (January 1988). 42-48, Nemetz, P,, "Bridging the Strategic Outcome Measurement Gap in Manufacturing Organizations," In ), Ettlie. M, C. Burstein and A. Fiegenbaum (Eds,). Manufacturing Strategies, Kluwer Academic Publishers. Boston. MA, 1990. 63-74, Pegets, C, C, and C, Sekar, "Determining Strategic Groups Using Multidimensional Scaling," Interfaces. 19, 3 (May-June 1989). 47-57, Phillips. L, W,. "Assessing Measurement Error in Key Informant Reports: A Methodological Note on Organizational Analysis in Marketing," /, Marketing Research. 18 (November 1981), 3 9 5 415,

Porter, M,. Competitive Strategy: Techniques for Analyzing Industries and Competitors, Free Press, New York. 1980, Rao, C, R,. Linear Statistical Inference, John Wiley and Sons, New York, 1973, Richardson. P. R,. A. J, Taylor and J, R, M. Gorden. "A Strategic Approach to Evaluating Manufacturing Performance," interfaces, l 5 ( N o v - D e c 1985). 15-27, Romesburg, H. C , Cluster Analysis for Researchers. Lifetime Learning Publications. BelmonE, CA. 1984. Roth, A. V,. "Differentiated Manufacturing Strategies for the Competitive Advantage: An Empirical Investigation." Manufacturing Roundtabie Research Report Series, Boston University, 1987, presented at the Annual Meeting of the Decision Sciences Institute, Las Vegas, NV, 1987, , "Linking Manufacturing Strategy and Performance: An Empirical Investigation," Manufacturing Roundtabie Research Report Series. Boston University, presented at the Academy of Management Annual Conference. Washington, D . C , 1989, , A. DeMeyer and A. Amano, "International Manufacturing Strategies: A Comparative Analysis," In K. Ferdows (Ed.), International Manufacturing. North Holland, Amsterdam, 1989, 187-211. and J. G. Miller, 1987 North American Manufacturing Futures Survey Fact Book, Manufacturing Roundtabie Research Monograph. Boston University, Boston. MA, 1987, and , "Manufacturing Strategy, Manufacturing Strength, Managerial Success, and Economic Outcomes," In J. Ettlie, M. C. Burstein and A, Fiegenbaum (Eds.), Manufacturing Strategies. Kluwer Academic Publishers, Boston, MA. 1990, 97-108. and J, G, Miller. "Success Factors in Manufacturing." Business Horizons. 35, 4 (1992), 73-81User's Guide: Statistics. Version 6 Edition. SAS Institute Inc., Cary, NC, 1985. Schendel. D, and C, W, Hofer (Eds,). Strategic Management: A New Vieic of Business Policy and Planning, Little, Brown & Co,, Boston, MA, 1979. Schroeder. R, G,. J. C. Anderson and G, Cleveland. "The Concept of Manufacturing Strategy: An Empirical Study." /, Operations Management (August 1986). 405-415, Skinner, W,. "Manufacturing: Missing Link in Corporate Strategy," Harvard Business Review. (May-June 1969), 136-145. . Manufacturing in the Corporate Strategy, John Wiley and Sons, New York, 1978, , Manufacturing: The Formidable Competitive Weapon. John Wiley and Sons, New York, 1985. Stobaugh, R. and P. Telesio, "Match Manufacturing Policies and Product Strategy," Han'ard Business Review (March-April 1983), 113-120, Swamidass. P, and W, Newell, 'Manufacturing Strategy. Environmental Uncertainty, and Performance: A Path Analytic Model," M<3Jm,(s'('/i'ii(Sci,, 33 (1987), 509-524, Thietart. R, A, and R, Vivas, "An Empirical Investigation of Success Strategies for Business along the Product Ufe Cycle," Management Sci.. 12(1984). 1405-1423.

MANAGEMENT SciENCE/VoL 40, No, 3, March 1994

303

MILLER AND ROTH A Taxonomy of Manufacturing Strategies

Thompson. J, D,. Organizations in Action. McGraw Hill, New York, 1967, Utterback, J, M, and W, J, Abemathy, 'A Dynamic Model of Process and Product Innovation," Omega, 3 (1975), 639-656. Voss, C. A, (Ed.), Manufacturing Strategy. Content and Process. Chatham & Hall London, 1992, Ward, P., J. Miller and T, Vollmann, 'Mapping Manufacturers Concerns and Actions Plans," International}. Operations and Production Management (Fall 198S), 5-18. Wheelwright, S. C , "Reflecting Corporate Strategy in Manufacturing Decisions," Business Horizons (February 1978), 57-66,

, "Strategy, Management and Strategic Planning Approaches," Interfaces. 14. 1 (1984). 19-33. ; "Manufacturing Strategy: Defining the Missing Link," Strategic Management!., 5. 1 (1985), and R, H, Hayes, 'Competing Through Manufacturing," Harvard Business Review (January-February 1985), 99-109. Wood. C , L. Ritzman and D, Sharma. "Intended and Achieved Competitive Priorities: Measures, Frequencies and Financial impact," In J, Ettlie, M, C, Burstein and A, Fiegenbaum (Hds), Manufacturing Strategies, Kluwer Academic Publishers, Boston, MA. 1990, 225-232. Woodward, ].. Induslriat Organizathn, Theory and Practice, Oxford University Press, London. 1965,

Accepted by L. Joseph Thomas, former Departmental Editor; received April 7. 1989. This paper has been with the authors 33 mcitiths for 3, revisions.

304

MANAGEMENT SciENCE/Vol. 40, No, 3, March 1994

You might also like