ISSN 2683-345X

journal homepage: http://ijiemjournal.uns.ac.rs/

International Journal of Industrial

Engineering and Management

Volume 11 / No 4 / December 2020 / 275 - 287

Original research article

The cumulative capability models:

A contingency analysis

M. Vilkas *, J. Duobiene, R. Rauleckas

Kaunas University of Technology, Kaunas, Lithuania

ABSTRACT ARTICLE INFO

Cumulative capability development models imply that there is a universal sequence in the de- Article history:
velopment of performance capabilities. Despite the critique that such an approach does not
Received September 22, 2020
align with contingency theory, efforts to identify which internal organizational characteristics
affect the sequence of capabilities development have been limited. The study employs PLS- Revised December 21, 2020
SEM multigroup analysis and a sample of 500 production firms to reveal whether internal Accepted December 22, 2020
organizational factors result in contrasting capability development sequences. The findings Published online December 28, 2020
reveal that strategic priority choices result in significantly different capability development
paths. The study suggests that strategic priority is an important contingency factor for explain-
ing whether an organization will follow the sand cone sequence of capability development. Keywords:
Cumulative capability development;
Competitive priorities;
Contingency theory

*Corresponding author:
Mantas Vilkas
mantas.vilkas@ktu.lt

1. Introduction the core of the operation strategy field [3], [4] and are
extensively tested and extended in various contexts
The research on the sequence of development of [5], [6].
performance capabilities is centered around cumula- The parsimony of the cumulative capability sand
tive capability development models. The proponents cone model and its universality claims were among
of the cumulative capabilities approach [1], [2] argue the most important reasons for its prevalence. Schol-
that organizations can excel at multiple dimensions ars suggest [7] that the model is based on three in-
of performance; however, the capabilities develop- tertwined propositions related to universality. First,
ment sequence matters. One of the most accepted the model assumes that companies compete on the
sequences of capability development – the sand cone same sets of competitive capabilities. Second, the
model [2] – suggests that organizations should start model implies that companies must succeed in all
improving quality performance, concentrating on de- capabilities to achieve success. Finally, there is one
livery, flexibility, and, finally, cost-effectiveness. Cu- best way to develop capabilities. Indeed, the propo-
mulative capability development models belong to nents of the cumulative capability approach argued

Published by the University of Novi Sad, Faculty of Technical Sciences, Novi Sad, Serbia. DOI: http://doi.org/10.24867/IJIEM-2020-4-271
This is an open access article distributed under the CC BY-NC-ND 4.0 terms and conditions
276 Vilkas et al.

that the model is a universal way of the development collection procedures and methods of the study. Fi-
of capabilities: "An obvious criticism of this model is nally, we present the multigroup analysis results to
that we seem to throw all contingencies overboard. reveal which internal organizational characteristics
The model seems to suggest that there is only one affect the capabilities development sequence. Finally,
best way to achieve multiple sets of manufacturing we theorize the findings and discuss the contribution
capabilities; to some extent, this is indeed our belief” of the results..
[2, p. 174].
The model was tested extensively [8]-[12], [1],
[5]-[6], but the empirical support for the sand cone 2. Literature review and theoretical
model was mixed. Empirical tests show that 53% [1], framework
53%, and 49% (in two samples) [13] of organizations
may be classified as following a path of the sand cone The prevailing model of deliberate strategy pro-
model. The findings show that the cumulative capa- vides a framework for connecting strategic intentions,
bility development model is widely prevalent. These strategic actions, and performance (Skinner, 1969,
findings also challenged the claims of the universality Hayes and Wheelwright, 1988). Following the mod-
of the cumulative capability development model. It el, organizations are advised to select among compet-
was proposed that there are contingency factors that ing priorities. The selected priorities then guide the
guide the sequence that organizations follow [11], [1], pattern of structural and infrastructural decisions. As
[13]. Empirical evidence was provided regarding the a result, organizations develop competitive capabil-
differences in the patterns of cumulative capabilities ities. Strategic priorities represent the intentions for
between countries and industries [14]-[15], and be- which the performance dimension or pattern of di-
tween industries [11]. While environmental factors mensions are essential and will be developed in the
are associated with capabilities development differ- future. Structural and infrastructural decisions are
ences, contingency analysis of internal organizational strategic choices regarding investments in assets and
factors on capabilities development is still missing the selection of operating policies [3]. Competitive
[11]. In this article, we seek to shed light on whether capabilities are the ability to compete on the dimen-
the sand cone capabilities development sequence is sions of performance relative to primary competitors
dependent on competitive priorities and other inter- in the target markets [1]. The framework is used in
nal organizational characteristics, such as the size of different ways to explain how organizations gain com-
an organization, product complexity, type of design, petitive advantage. The research on strategic config-
manufacturing process, batch size, and extent of ca- urations argues that competitive priorities relative to
pacity utilization. competitors provide the foundation for competitive
This study is based on data collected as a part of advantage. In contrast, the resource-based view ar-
the European manufacturing survey in 2018. The gues that organizational assets, which are valuable,
data on a representative sample of manufacturing rare, nonsubstitutable, and challenging to imitate,
organizations in a country (N=500) were collected are fundamental to competitive advantage. Such as-
using a telephone-based survey method. Multigroup sets lead to superior quality, delivery, flexibility, and
analysis of partial least squares structural equation cost-competitive performance.
modeling was used to identify the differences in the Two research streams concentrate on the process
accumulation of capabilities based on the differences of developing performance capabilities: the trade-off
in competing priorities and other internal organiza- and the cumulative capabilities approach. The advo-
tional characteristics. This research contributes to cates of the trade-off approach [16] argue that com-
contingency research on cumulative capability mod- panies cannot perform well on all dimensions be-
els. The results reveal that competitive priorities are cause the improvement in a performance dimension
an important factor determining the sequence of results in the deterioration of another performance
competitive capabilities development. The study also dimension. In contrast, the proponents of the cumu-
reveals that other internal organizational characteris- lative capabilities approach [2], [19] argue that orga-
tics do not result in statistically significant differences nizations can excel at all dimensions of performance;
in the accumulation of capabilities. however, the capabilities development sequence mat-
The remainder of the paper is organized as fol- ters. The sand cone model [2] suggests that organi-
lows. First, we review the literature on sand cone se- zations should start improving quality performance,
quences and present a theoretical framework guiding concentrating on delivery, flexibility, and, finally,
our empirical research. Later, we describe the data cost-effectiveness. The model rests on the following

International Journal of Industrial Engineering and Management Vol 11 No 4 (2020)

Vilkas et al. 277

assumptions. First, quality constitutes the foundation prevalent in developing countries than the sand cone
of all capabilities [2], [10] [19]. Second, improved sequence Q->D->F->C. It was argued that emerging
quality and delivery capabilities can improve flexibil- countries have more leverage to develop cost-com-
ity and cost capabilities [10]. Third, each step in the petitive performance instead of flexibility competitive
sequence requires more learning than the previous performance in their earlier organizational maturity
step [10], [5]. Fourth, costs are treated as the effect stages. While the external contextual factors were as-
or outcome of quality, delivery, and flexibility capa- sociated with differences in the capabilities develop-
bilities [5]. ment sequence, the research on internal factors un-
The research on the sand cone model of capa- der which the sand cone sequence is supported is still
bilities development attracted much interest, as it limited. In this article, we ask whether the sand cone
promised that a universal capabilities development capabilities development sequence is dependent on
sequence might exist. The model was tested exten- competitive priorities and other internal organization-
sively [8]-[12], [1], [5]-[6], but the empirical support al characteristics, such as the size of an organization,
for the sand cone model was mixed. Empirical tests product complexity, type of design, manufacturing
show that 53% [1], 53%, and 49% (in two samples) process, batch size, and extent of capacity utilization
[13] of organizations may be classified as following (Figure 1).
the sand cone model’s path. The findings show that The theoretical framework guides the empirical
while the percentage of organizations following the research by identifying the internal organizational
sequence is considerable, the sand cone is not a uni- characteristics that result in differences in the accu-
versal sequence of performance capability develop- mulation of capabilities. The framework allows to
ment. It has been suggested that internal and external test seven hypotheses:
contingencies influence the pattern of competitive There are statistically significant differences in ac-
performance development [1], [13]. cumulation quality, delivery, flexibility, and cost-com-
Contingency research in operations management petitive performance among organizations:
is intended to reveal the contextual conditions under
H1. competing on different strategic priorities;
which operations practices are effective [20]-[21].
H2. characterized by different size;
Contingency research on sand cone sequences would
H3. producing products characterized by
reveal the external conditions and internal factors
different degrees of product complexity;
under which the sand cone sequence is supported.
H4. using different types of the design process;
Empirical evidence was provided for the differenc-
H5. using different types of the manufacturing
es in the patterns of cumulative capabilities between
process;
countries [15], [11], [14], and industries [11]. For ex-
H6. using different batch sizes;
ample, it was revealed that the sequence Q->D->C-
H7. characterized by different levels of capacity
>F, titled “the most comfortable sequence,” is more
utilization.

Sand cone sequence

Flexi-
Quality Delivery Cost
bility

Contingency factors
Competitive priority
Size of organization
Product complexity
Type of product development process
Type of manufacturing process
Batch size
Capacity utilization

Figure 1. Internal factors affecting the sand cone sequenc

International Journal of Industrial Engineering and Management Vol 11 No 4 (2020)

278 Vilkas et al.

In the next section, we describe the research de- were technical managers or production managers at
sign, which allows testing whether competitive prior- manufacturing sites with more than 200 employees,
ities and other internal characteristics of an organiza- and general managers, technical managers, and pro-
tion affect the sequence of capability development. duction managers at manufacturing sites with fewer
than 200 employees. A stratified random sampling
procedure was used. Strata were defined in terms of
3. Research method four regions of the country and four size classes of or-
ganizations (2-19, 20-49, 50-99, 100-249 employees).
3.1. Research design Then 2330 manufacturing sites were contacted. The
effective sample is 500 manufacturing sites, which
We used partial least squares structural equation constitutes a 215% response rate. The sample char-
modeling (PLS-SEM) to test a cumulative capabili- acteristics are presented in Table 1.
ty development model. Furthermore, we employed The effective sample has adequate representation
a multigroup analysis to determine if the cumula- for all firm sizes, regions of the country, and all man-
tive capability development model is contingent on ufacturing subsectors.
strategic priorities and other internal organizational
characteristics. Multigroup analysis enables us to test
for differences between identical models estimated 3.3. Measures
for different groups of organizations [22]. The anal- The selection of measures of competitive perfor-
ysis reveals whether there are statistically significant mance was based on an extensive review of the litera-
differences between individual group models [22]. ture. Several approaches to the operationalization of
For example, we used multigroup analysis to com- performance dimensions are available [1], [12]-]13]
pare whether the sand cone sequence of the develop- [25]-[24]. We used the operationalization proposed
ment of performance capabilities is different among by Schroeder and his coauthors [1] for the measure-
organizations, choosing quality, delivery, flexibility, ment of quality, delivery, flexibility, and cost perfor-
and cost as their primary competitive priorities. The mance (Table 2). The respondents were asked to
sample is split into four subsamples or groups based indicate how well their factory performed compared
on dominating competitive priority. The path coeffi- to its competitors within their industry along the dif-
cients among the capabilities prescribed by the sand ferent performance dimensions. Competitive priori-
cone model are estimated in each group. Finally, af- ties were measured by asking respondents to rank six
ter bootstrapping 5000 samples (Hair et al., 2016), competitive priorities (product price, product quality,
statistically significant differences between sand cone customization to customers' demands, delivering on
sequences in each group were revealed. SmartPLS schedule/ short delivery times, innovative products
3 software [23] was used for path analysis and multi- and services) in order of importance, anchored with
group analysis. 1 as “most important” and 6 as “least important.”
The assignment of equal importance was not al-
3.2. Data collection and description of lowed. The organizations that use strategic priorities
a sample of services [17] and innovative products [18] were not
included in the analysis due to low frequencies. The
Data were collected as part of the European man- scales of performance capabilities and internal orga-
ufacturing survey (EMS) in 2018. EMS is an inter- nizational factors are provided in Annex 1.
national network of research institutions collecting
data in their respective countries. The standardized
questionnaire was used for data collection. The data 3.4. Measurement model
were collected on individual manufacturing sites be- Partial squares structural equation modeling was
cause each manufacturing site in a business unit may employed for the confirmatory factor analysis (CFA)
exhibit different performance capabilities [1]. A tele- of the measurement model. The characteristics of
phone survey was used to collect data. The data from the measurement model are presented in Table 2.
one country were used for this research to minimize The reliability of the employed scales was mea-
the country difference effects. The sampling frame sured by Cronbach’s alpha and the composite reli-
consists of 6122 manufacturing sites covering all sub- ability score. Although reliability estimates are above
sectors of manufacturing and covers the country's 0,95 thresholds, the content validity of the original
population of manufacturing sites. The respondents scales is maintained. EFA suggested four factors, as

International Journal of Industrial Engineering and Management Vol 11 No 4 (2020)

Vilkas et al. 279

Table 1. Sample characteristics

Demographic characteristics N=500

Industry Number %
Engineering 125 25.0
Food 64 12.8
Textiles 70 14.0
Wood and paper 156 31.2
Chemicals and chemistry 11 2.2
Other 74 14.8
Number of employees Number %
< 20 300 60.0
20-49 108 21.6
50-99 44 8.8
100-249 34 6.8
> 250 14 2.8
Competitive priority Number %
Product quality 248 49.6
Customization for customer 104 20.8
Fast/on-time delivery 60 12
Product price 48 9.6
Innovative products 28 5.6
Services 12 2,4
Respondent position Number %
Head/director 334 66.8
Technical manager, director, head of production 162 32.4
Branch head 4 0.8

Table 2. The characteristics of the measurement model

Construct Item Code Factor p-values Cronbach CR AVE

loadings alpha
Product overall quality Q1 0,962 0,000 0,914 0,959 0,921
performance
Quality
Conformance to Q2 0,958 0,000
specification
On-time delivery performance D1 0,955 0,000 0,911 0,957 0,918
Delivery Delivery speed D2 0,961 0,000
(from order until delivery)
Ability to adjust production F1 0,925 0,000 0,92 0,949 0,861
volumes
Ability to produce a range of F2 0,925 0,000
Flexibility
products
Speed on new product F3 0,934 0,000
introduction into the plant
Unit cost C1 0,952 0,000 0,938 0,96 0,89
Cost Manufacturing overhead cost C2 0,965 0,000
Inventory turnover C3 0,913 0,000

International Journal of Industrial Engineering and Management Vol 11 No 4 (2020)

280 Vilkas et al.

expected, with high loadings and low cross-loadings. exist, where the accumulation of capabilities may be
Loadings obtained by CFA are in Table 2. Con- higher in economies that differ in their development
vergent validity is demonstrated on the item level stage [14]. On the other hand, other authors [7], [5]
through statistically significant loadings exceeding the have found very similar relationships among the con-
threshold of 0,708 and on the construct level through structs of the model (Q->D, 0,42*; D->F, 0,4*, and
the average variance extracted (AVE) being not less F->C, 0,26*). In general, the cumulative capability
than 0,5 for every construct (Hair et al., 2016). Dis- model's testing provided results consistent with the
criminant validity was established by the absence of results of other studies. In the next section, we pro-
high cross-loadings, matching the Fornell-Larcker ceed with the contingency analysis of the model.
criterion, and passing the heterotrait-monotrait ratio
(HTMT) test (Annex 2). 4.2 The contingency analysis of the sand
cone sequence
4. Results Multigroup analysis [26] was performed to test
whether the sand cone model is contingent on inter-
4.1 The manifestation of the sand cone nal organizational characteristics such as competitive
sequence in the sample priorities, the size of an organization, product com-
plexity, the type of product development process, the
The analysis of the results is presented in the type of manufacturing process, the size of the batch,
following way. First, we verify the sand cone model and the extent of capacity utilization. The analysis re-
using the sample data. The accumulation of quality, veals that the sand cone capabilities development se-
delivery, flexibility, and cost capabilities is confirmed quence is dependent on competitive priorities. There
if the path coefficients are positive and statistically sig- are no statistically significant differences in the accu-
nificant. Then, we apply multigroup analysis to test mulation of capabilities in groups based on other in-
for significant differences among path coefficients be- ternal characteristics. The sand cone models' charac-
tween two groups based on the contingency variables teristics that are based on contingency variables and
identified in the literature review. the extent and statistical significance of differences
The path model was developed according to the among models are provided in Annex 3.
classic cumulative capability model proposed by Fer- The analysis of the sand cone capability devel-
dows and de Meyer [2]. In the model, quality com- opment sequence of organizations that prioritize
petitive performance is the antecedent of delivery contrasting competitive priorities reveals statistically
performance, competitive delivery performance is significant differences among the accumulation of
the antecedent of flexibility performance, and com- capabilities. The sand cone model is observed; the
petitive flexibility performance is the antecedent of path coefficients between pairs of capabilities are
cost performance. substantial and statistically significant only among the
The path analysis revealed the relationships organizations that prioritize quality as a strategic pri-
among the latent variables of the sand cone model ority (Figure 3). Within organizations that compete
(Figure 2). The latent variables are bounded by the for delivery, the accumulation of quality and delivery,
medium relations in the model. Schroeder et al. delivery, and flexibility is observed. If an organization
(2011) found a stronger association between perfor- chooses competitive priorities of flexibility or cost,
mance capabilities (Q->D, 0,61**; D->F, 0,7**, and then only quality and delivery capabilities are posi-
F->C, 0,64**). However, country differences may

F2
Q1 Q2 F1 F3
,487** ,321**
Quality Delivery Flexibility Cost
,358**
D1 D2 C1 C3
C2

Figure 2. The cumulative capability development model (N=397)

International Journal of Industrial Engineering and Management Vol 11 No 4 (2020)

Vilkas et al. 281

tively related at a statistically significant level. product development process, type of manufacturing
The Q->D relationship is statistically and signifi- process, batch size, or extent of capacity utilization.
cantly different at 0,15* (p=0,028) among groups of However, the sand cone accumulation of capabilities
organizations that emphasize quality to cost, indicat- is not manifested in all groups of organizations. The
ed as difference D1 in Figure 2. The D->F relation- sand cone sequence is observed in groups of organi-
ship is statistically and significantly different at 0,313* zations with the following characteristics: fewer than
(p=0,021) among groups of organizations that prior- 20 and more than 20 employees; produce in a single
itize quality to flexibility (D2). Estimates of the F->C unit; small or large batches; and operate over 90% of
relationship are statistically and significantly different capacity and less than 90% capacity. However, the
at 0.400* (p=0,03) among organizations that compete sand cone sequence does not hold consistently when
on quality versus delivery (D3). Finally, the D->F re- other organizational characteristics are considered.
lationship is statistically and significantly different For example, the sand cone pattern is observed in
at 0,559* (p=0,034) among groups of organizations organizations whose products are characterized by
that prioritize delivery to cost (D4). In summary, the medium complexity but not in companies producing
findings reveal that the sand cone capabilities devel- simple (Q->D at 0,275, p=0,068) or complex prod-
opment sequence tends to manifest if organizations ucts (D->F at 0,252, p=0,055). The accumulation of
choose to compete on quality. Organizations that capabilities remains in organizations that develop
choose to compete on other strategic priorities tend products according to customer specifications and
to follow different paths of capabilities development. as a standard program from which the customer can
The Q->D relationship is universal, despite strategic choose. However, it does not hold in organizations
priority choices regarding quality, delivery, flexibility, that develop products according to a standardized ba-
or cost. sic program incorporating customer-specific options
The results show no statistically significant differ- (F->C at 0,279, p=0,082). We confirm the sand cone
ences in the accumulation of capabilities in groups capabilities development sequence in organizations
based on other internal characteristics, such as the that manufacture upon receipt of a customer order,
size of an organization, product complexity, type of i.e., made-to-order products. However, we do not

,468**,D1 ,420**,D2 ,522**,D3

Flexi-
Quality Delivery Cost
bility
Model 1: The sand cone sequence when an organization prioritizes the strategic priority of quality

,501** ,536*,D4 ,122

Flexi-
Quality Delivery Cost
bility
Model 2: The sand cone sequence when an organization prioritizes the strategic priority of delivery

,551** ,229,D2 ,209,D3

Flexi-
Quality Delivery Cost
bility
Model 3. The sand cone sequence when an organization prioritizes the strategic priority of flexibility

,363*,D1 -,023,D4 ,072

Flexi-
Quality Delivery Cost
bility
Model 4. The sand cone sequence when an organization prioritizes the strategic priority of cost

Figure 3. The sand cone model of in groups of organizations that prioritize contrasting competitive priorities

International Journal of Industrial Engineering and Management Vol 11 No 4 (2020)

282 Vilkas et al.

confirm it in organizations that manufacture to stock ity, type of design, manufacturing process, batch size,
(prior to customer orders) (Q->D at 0,289, p=0,112). or extent of capacity utilization result in statistically
In the next section, we discuss the findings. significant differences in the sand cone sequence. A
summary of our findings is provided in Table 3.
We confirm hypothesis H1 out of the seven hy-
5. Discussion potheses, proposing statistically significant differenc-
es in accumulation quality, delivery, flexibility, and
In this article, we sought to reveal whether the
cost-competitive performance among organizations
sand cone capabilities development sequence is de-
that compete on different types of strategic priorities
pendent on competitive priorities or other internal
and differ in other internal organizational character-
organizational characteristics, such as the size of an
istics.
organization, product complexity, type of design,
The findings show that the sand cone pattern is
manufacturing process, batch size, and extent of ca-
observed despite the size of the organization, size of
pacity utilization. First, we confirmed that the sand
the batch, and capacity utilization. The sand cone
cone pattern is manifested in our sample of 500 orga-
pattern is not observed across all types of competitive
nizations. Second, we performed a multigroup anal-
priorities, product development processes, manufac-
ysis to reveal the sand cone model's differences in
turing processes, or product complexity. For exam-
groups of organizations characterized by the selected
ple, sand cone patterns are observed in organizations
contingency variables. Our efforts contribute to stud-
that produce medium complexity products, but they
ies of cumulative capability models and the opera-
are not observed in organizations producing simple
tions strategy field in several ways.
or complex products. These findings constitute an
First, our empirical results confirm that the choice
important result of the study, and possible explana-
of strategic priority is an important contingency factor
tions of the reasons for the sand cone pattern's insta-
predicting whether the sand cone pattern of capability
bility are warranted.
development will be followed. In addition, our find-
Second, our findings show that the sand cone pat-
ings do not show that other internal characteristics,
tern of capability development is evident among orga-
such as the size of an organization, product complex-
nizations that choose quality as their competitive pri-
Table 3. Summary of the findings of the contingency analysis of the san cone model

The contingency factor Hypotheses Does the sand cone pattern hold in groups characterized by Are diffe ences
testing contingency variables? statistically significant
Quality Yes
Delivery No
Competitive priority Confirme Yes
Flexibility No
Cost No
Less than 20 employees Yes
Company size Rejected No
20 and more employees Yes
Simple products No
Product complexity Rejected Medium complexity products Yes No
Complex products No
According to customer specificatio Yes
Standardized basic program incorporating cus- No
Type of product develop-
Rejected tomer-specific option No
ment process
a standard program from which the customer can Yes
choose
Type of manufacturing Make-to-stock No
Rejected No
process Make-to-order Yes
Single unit production Yes
Batch size Rejected Small or medium batch Yes No
Large batch Yes
Operating at production frontier Yes
Capacity utilization Rejected No
Operating below production frontier Yes

International Journal of Industrial Engineering and Management Vol 11 No 4 (2020)

Vilkas et al. 283

ority. The companies that choose other competitive sand cone sequence do not provide the frequencies
priorities, such as delivery, flexibility, or cost, follow of organizations competing on different strategic pri-
other capability development sequences. The sand orities in their samples. In our case, the part of the
cone pattern in the complete sample is manifested sample competing on strategic priority (i.e., 49,5%)
with decreasing direct effects (Q->D, 0,487**; D->F, is similar to the overall measured prevalence of the
0,358**, and F->C, 0,321**). The direct effects of a sand cone sequence in other studies (approx. 50%)
pattern among organizations that compete on quality [1], [5], [13]. The numbers may coincide accidental-
become more uniform and tend to increase instead ly; however, our contingency analysis reveals that the
of decrease (Q->D, 0,468**; D->F, 0,42**, and F->C, choice of strategic priority of quality is a strong pre-
0,522**). The extent of accumulation considering di- dictor of whether the sand cone sequence of capabil-
rect and indirect effects is substantial (Table 4). ities development will be followed.
The extent of the indirect effects of quality on
flexibility and cost, and the indirect effects of delivery
on cost is especially intriguing. The results reveal that 6. Conclusions
an increase in quality results in a direct increase in
The article reports the results of contingency anal-
delivery competitive performance (0,468**). It also
ysis performed to reveal whether the sand cone capa-
increases competitive performance in terms of flex-
bilities development sequence is dependent on com-
ibility (0,197**) and cost (0,103**). Such indirect
petitive priorities and other internal organizational
effects reveal the importance of quality competitive
characteristics, such as the size of an organization,
performance on other sand cone competitive per-
product complexity, type of design, manufacturing
formance capabilities. The accumulation of com-
process, batch size, and extent of capacity utilization.
petitive performance through direct and indirect
We find that the choice of strategic priority results
effects predicted by the sand cone model [2], [10]
in statistically and significantly different development
is manifested among the organizations that compete
paths. The choice of strategic priority of quality re-
on quality. The sand cone sequence proposes that
sults in the sand cone sequence of cumulation of
cost performance results from perfected quality, de-
capabilities. The choice of other strategic priorities,
livery, and flexibility capabilities [10], [5]. Our results
such as delivery, flexibility, and cost, results in alter-
show that cost performance benefits from increased
nate sequences of development of capabilities. The
quality (0,103**), delivery (0,219**), and flexibility
previous results show that the sand cone sequence
(0,522**) performances. This finding allows quantifi-
is followed by approximately 50% of organizations.
cation of the extent of the accumulation of sand cone
Our results suggest that this percentage may be solely
performance capabilities among organizations char-
explained by one contingency factor: the choice of
acterized by quality and competitive performance.
strategic priority of quality.
Finally, previous research showed that 53% [1],
Several limitations characterize the research.
53% and 49% (in two samples) [13], and 52% [5] of
The reliability values of the quality, delivery, cost,
companies follow the sand cone pattern of improve-
and flexibility performance dimensions are over 0,9.
ment in capabilities. However, previous research has
High values of reliability measure indicate that items
provided conflicting insights into which contingency
are measuring the same dimension of the construct
variables may account for these organizations. Our
instead of measuring different dimensions of the con-
results suggest that the choice of the competitive ca-
struct. The groups of companies that are character-
pability of quality may be responsible for the mani-
ized by low cost (N=48) and delivery (N=60) compet-
festation of the sand cone sequence. A total of 49,5%
itive priorities are smaller comparing with a group of
of companies in our sample report competing on the
organizations characterized by quality (N=248) and
strategic priority of quality (Table 1). Unfortunately,
flexibility (N=104) competitive priorities. However,
the other studies that observe the prevalence of the

Table 4. The direct and indirect effects of sand cone sequences among o ganizations that are characterized by strategic
priority of quality

Quality Delivery Flexibility Cost

Quality - 0,468** 0,197** 0,103**
Delivery - - 0,42** 0,219**
Flexibility - - - 0,522**
Cost - - - -

International Journal of Industrial Engineering and Management Vol 11 No 4 (2020)

284 Vilkas et al.

this weakness is mitigated in part by bootstrapping [11] B. B. Flynn and E. J. Flynn, “An exploratory study of the
nature of cumulative capabilities,” Journal of Operations
procedures.
Management, vol. 22, no. 5, pp. 439-457, 2004.
Further contingency research on cumulative capa- [12] A. Grossler and A. Grubner, “An empirical model of the
bility models could also consider the following things. relationships between manufacturing capabilities”,
This paper reveals that sand cone sequence of devel- International Journal of Operations and Production
Management, vol. 26, no. 5, pp. 458-485, 2006.
opment of capabilities is characteristic to companies [13] R. Narasimhan and T. Schoenherr, “Revisiting the
that compete on quality. Further research could re- progression of competitive capabilities: results from a
veal what sequences are characteristic to companies repeated cross-sectional investigation,” International Journal
of Production Research, vol. 51, no. 22, pp. 6631-6650,
that compete on delivery, cost, or flexibility compet- 2013.
itive priorities. Second, there is a gap of knowledge [14] K. Amoako-Gyampah and M. Acquaah, “Manufacturing
on the mechanisms of accumulation of competitive strategy, competitive strategy and firm performance:
performance capabilities. Finally, other empirical An empirical study in a developing economy environment”,
International journal of production economics, vol. 111,
methods that could support the sequentially of de- no. 2, pp. 575-592, 2008.
velopment of capabilities would be welcomed in the [15] L. M. Corbett and G. S. Claridge, “Key manufacturing
future research of cumulative capability models. capability elements and business performance,” International
Journal of Production Research, vol. 40, no. 1, pp.109-131,
2002.
Funding [16] W. Skinner, “Manufacturing strategy on the "S" curve,”
Production and Operations Management, vol. 5, no. 1,
pp. 3–13, 1996.
This work was supported by the Research Coun- [17]
U. Marjanovic, B. Lalic, N. Medic, J. Prester, and I.
cil of Lithuania (grant number S-MIP-17-128). Palcic, “Servitization in manufacturing: role of antecedents
and firm characteristics,” Int. J. Ind. Eng. Manag., vol. 10,
no. 2, pp. 133–144, 2020.
References [18] B. Lalic, N. Medic, M. Delic, N. Tasic, and U. Marjanovic,
“Open innovation in developing regions: An empirical
[1] R.G. Schroeder, R. Shah, and D. Xiaosong Peng, “The analysis across manufacturing companies,” Int. J. Ind. Eng.
cumulative capability ‘sand cone’ model revisited: A new Manag., vol. 8, no. 3, pp. 111–120, 2017.
perspective for manufacturing strategy,” International [19] C. Corbett, and L. Van Wassenhove, “Trade-offs? What
Journal of Production Research, vol. 49, no.16, pp. trade-offs? Competence and competitiveness in
4879-4901, 2011. manufacturing strategy,” California management review,
[2] K. Ferdows and A. de Meyer, “Lasting improvements in vol. 35, no. 4, pp. 107-122, 1993.
manufacturing performance: In search of a new theory,” [20] M. Ketokivi and R. Schroeder, “Manufacturing practices,
Journal of Operations Management, vol. 9, no. 2, pp. strategic fit and performance: a routine-based view,”
168-184, 1990. International Journal of Operations & Production
[3] E.D. Rosenzweig and G.S Easton, “Tradeoffs in Management, vol. 24, no. 2, pp. 171-191, 2004.
manufacturing? A meta-analysis and critique of the [21] R., Sousa and C. A. Voss, “Contingency research in
literature,” Production and Operations Management,’ operations management practices,” Journal of Operations
vol.19, no.2, pp. 127-141, 2010. Management, vol. 26, no. 6, pp. 697-713, 2008.
[4] K. K. Boyer, M., Swink, and E. D. Rosenzweig, “Operations [22] J. F. Hair Jr, G. T. M. Hult, C. Ringle and M. Sarstedt,
strategy research in the POMS journal,” Production and “A primer on partial least squares structural equation
Operations Management, vol. 14, no. 4, pp. 442-449, 2005. modeling (PLS-SEM),” Sage Publications, 2016.
[5] S. Boon-Itt, and C. Y. Wong, “Empirical investigation of [23] C.M. Ringle, S. Wende, and J.-M. Becker, “SmartPLS 3,”
alternate cumulative capability models: a multi-method Boenningtedt: SmartPLS GmbH, http://www.smartpls.com,
approach,” Production Planning & Control, vol. 27, no. 4, 2015.
pp. 299-311, 2016. [24] R. Narasimhan, M Swink, and S. W. Kim, “Disentangling
[6] S. Gold, R. Schodl, and G. Reiner, “Cumulative leanness and agility: an empirical investigation,” Journal of
manufacturing capabilities in Europe: Integrating operations management, vol 24, no 5, pp. 440-457, 2006.
sustainability into the sand cone model,” Journal of Cleaner [25] P. J. Singh, F. Wiengarten, A. A. Nand, and T. Betts,
Production, vol. 166, pp. 232-241, 2017. “Beyond the trade-off and cumulative capabilities models:
[7] M. Hallgren, J. Olhager and R.G. Schroeder, “A hybrid alternative models of operations strategy,” International
model of competitive capabilities. International Journal of Journal of Production Research, vol 53, no 13, pp.
Operations & Production Management,” vol. 31, no. 5, 4001-4020, 2015.
pp. 511-526, 2011. [26] M. Sarstedt, J. Henseler, and C. M. Ringle, “Multigroup
[8] M.A. Noble, “Manufacturing strategy: testing the cumulative analysis in partial least squares (PLS) path modeling:
model in a multiple country context,” Decision Sciences, Alternative methods and empirical results. In Measurement
vol. 26, no. 5, pp. 693-721, 1995. and research methods in international marketing”, Emerald
[9] M.A. Noble, “Manufacturing competitive priorities and Group Publishing Limited, 2011.
productivity: an empirical study,” International Journal of
Operations & Production Management,’ vol. 17, no.1/2, pp.
85-99, 1997.
[10] E.D. Rosenzweig and A.V. Roth, “Towards a theory of
competitive progression: evidence from high-tech
manufacturing,” Production and Operations Management,
vol.13, no. 4, pp. 354-368, 2004.
International Journal of Industrial Engineering and Management Vol 11 No 4 (2020)
Vilkas et al. 285

Annex A. Scales of measurement of Manufacturing

performance capabilities and
- Upon receipt of customer’s order, i.e. made-to-or-
contingency factors
der
Indicate well your factory performed compared to its - Final assembly of the product is carried out upon
competition within your industry along the different receipt of customer’s order, i.e. assembly-to-order
performance dimensions (anchors: poor or low end - To stock (before customer's order)
of industry; average; superior) - Does not exist in this factory
Product overall quality performance
Product conformance Batch or lot size
On-time delivery performance
Delivery speed (from the order until delivery) - Single unit production
Ability to adjust production volumes - Small or medium batch/lot
Ability to produce a range of products - Large batch/lot
Speed on new product introduction into the plant
Unit cost Capacity utilization
Manufacturing overhead cost
Inventory turnover Degree of manufacturing capacity utilization (average
in 2017)
Competitive priorities
Annex 2 The measurement model
Please rank the following competitive factors in order
of significance to distinguish your factory positively Cross-Loadings
from competitors. Please rank from 1 to 6, 1 indicat-
ing "the most important". Please do not assign equal Quality Delivery Cost Flexibility
importance to any factors. Q068_1 0,962 0,479 0,268 0,373
Product price Q068_4 0,958 0,455 0,264 0,399
Product quality Q069_2 0,458 0,955 0,337 0,319
Customization to customers' demands Q069_5 0,475 0,961 0,366 0,365
Delivering on schedule/short delivery times
Q070_1 0,237 0,335 0,952 0,317
Innovative products
Q070_2 0,272 0,319 0,965 0,302
Services
Q070_3 0,277 0,388 0,913 0,287
Q071_1 0,367 0,335 0,335 0,925
Which of the following characteristics best describes
Q071_4 0,385 0,302 0,262 0,925
your main product or line of products?
Q071_5 0,369 0,355 0,29 0,934

Product complexity
Fornell-Larcker Criterion
- Simple products
- Products with medium complexity Cost Delivery Flexibility Quality
- Complex products Cost 0,943
Delivery 0,367 0,958
Product development Flexibility 0,321 0,358 0,928
Quality 0,277 0,487 0,402 0,96
- According to customers’ specification
- As a standardized basic program incorporating cus-
tomer-specific options Heterotrait-Monotrait Ratio (HTMT)
- For a standard program from which the customer
can choose options Cost Delivery Flexibility Quality
- Does not exist in this factory Cost
Delivery 0,397
Flexibility 0,343 0,389
Quality 0,3 0,533 0,439

International Journal of Industrial Engineering and Management Vol 11 No 4 (2020)

286 Vilkas et al.

Annex 3 The results of the multigroup

analysis

Competitive priority

Competitive N Q->D p-value D->F p-value F->C p-value

priority
Quality 248 0,468** 0 0,42** 0 0,522** 0
Delivery 60 0,501** 0,001 0,536** 0,002 0,122 0,578
Flexibility 104 0,551** 0 0,229 0,131 0,209 0,204
Cost 48 0,363* 0,022 -0,023 0,9 0,072 0,722

Quality -Flexi- p-Value Quality -Delivery p-Value Quality -Cost p-Value p-value
bility
Q-> D 0,192 0,115 0,033 0,853 0,105* 0,028 0
D -> F 0,313* 0,021 0,116 0,573 0,443 0,068 0,578
F -> C 0,083 0,766 0,4* 0,03 0,45 0,608 0,204
Delivery -Cost p-Value Flexibility - Cost p-Value Delivery - Flexibility p-Value 0,722
Q-> D 0,138 0,709 0,188 0,309 0,05 0,759
D -> F 0,559* 0,034 0,252 0,767 0,307 0,210
F -> C 0,05 0,848 -0,518 0,352 0,087 0,759

Company size

Company size N Q->D p-Value D->F p-Value F->C p-Value

Less than 20 300 0,459** 0 0,344** 0 0,391** 0
employees
More than 20 200 0,532** 0 0,377** 0 0,23** 0,024
employees

Less than 20- More than 20 p-Value

Q-> D 0,073 0,477
D -> F 0,032 0,787
F -> C 0,161 0,18

Product complexity

Product N Q->D p-Value D->F p-Value F->C p-Value

complexity
Simple 83 0,275 0,068 0,31* 0,027 0,251 0,151
Medium 288 0,479** 0 0,397** 0 0,295** 0,001
Complex 114 0,565** 0 0,252 0,055 0,432** 0

Simple - Me- p-Value Simple - Com- p-Value Medium - p-Value

dium plex Complex
Q-> D 0,203 0,154 0,29 0,089 0,086 0,469
D -> F 0,087 0,586 0,059 0,76 0,146 0,349
F -> C 0,044 0,818 0,181 0,367 0,137 0,341

International Journal of Industrial Engineering and Management Vol 11 No 4 (2020)

Vilkas et al. 287

Type of product development

Product development N Q->D p-Value D->F p-Value F->C p-Value
According customer specifi- 242 0,502** 0 0,331** 0 0,235** 0,001
cation
As standardized basic pro- 88 0,293** 0,007 0,487** 0 0,279 0,082
gram incorporating customer
specific option
For a standard program from 142 0,415** 0 0,236* 0,02 0,287** 0
which the customer can
choose

Customer p-Value Custom spec p-Value Basic - p-Value

spec - Basic - Standard Standard
program
Q-> D 0,209 0,088 0,087 0,379 0,122 0,341
D -> F 0,156 0,256 0,095 0,425 0,251 0,073
F -> C 0,044 0,776 0,052 0,668 0,008 0,963

Type of manufacturing process

N Q->D p-Value D->F p-Value F->C p-Value
Made-to-order 389 0,554** 0 0,379** 0 0,293** 0
Made-to stock 85 0,289 0,112 0,253** 0,001 0,443* 0,044

Make-to order - Make-to-stock p-Value

Q-> D 0,265 0,052
D -> F 0,126 0,439
F -> C 0,15 0,361

Batch size
Batch size N Q->D p-Value D->F p-Value F->C p-Value
Single 119 0,514** 0 0,405** 0 0,422** 0,001
Small 289 0,478** 0 0,334** 00 0,343** 0
large 76 0,492** 0 0,322** 0 0,217** 0

Single - Small p-Value Single - Large p-Value Small - Large p-Value

Q-> D 0,036 0,765 0,023 0,885 0,014 0,93
D -> F 0,07 0,628 0,082 0,689 0,012 0,947
F -> C 0,079 0,563 0,205 0,342 0,126 0,521

Capacity utilization
Capacity N Q->D p-Value D->F p-Value F->C p-Value
utilization
Less than 90% 246 0,449** 0 0,389** 0 0,284** 0
More than 90% 167 0,544** 0 0,346** 0,001 0,336* 0,002

Low - High p-Value

Q-> D 0,096 0,398
D -> F 0,044 0,735
F -> C 0,053 0,689

International Journal of Industrial Engineering and Management Vol 11 No 4 (2020)