INTERNATIONAL GROWTH AND DEVELOPMENT OF SMALL FIRMS: PATTERNS OF START-UP AND INTERNATIONALISATION

 

 

 

Marian V. Jones and Stephen K. Tagg

Dr Marian V. Jones
University of Strathclyde
Strathclyde International Business Unit
173 Cathedral Street

Glasgow, G4 ORQ
Scotland UK

Tel.: +141 548 3241
Fax +141 552 2802

e-mail: marianj@market.strath.ac.uk

 

Dr Stephen Tagg
University of Strathclyde
Department of Martketing

173 Cathedral Street
Glasgow, G4 ORQ
Scotland UK

Tel.: +141 548 2409
Fax +141 552 2802

e-mail: s.k.tagg@strath.ac.uk

 



INTERNATIONAL GROWTH AND DEVELOPMENT OF SMALL FIRMS:
PATTERNS OF START-UP AND INTERNATIONALISATION

 

 

Abstract

Empirical evidence on the commencement and development of cross-border activity of very small firms suggests that internationalisation, particularly in high technology sectors is occurring more rapidly than it did in the past, with more variety and scope in the types of cross-border activity undertaken. The purpose of this paper is to determine the first cross-border activities undertaken by a sample of 196, UK based small, high technology firms, over time, from their foundation date. Additionally, the purpose is to describe common patterns of internationalisation start-up and internationalisation process experienced by the firms, and indicate whether emergent patterns are associated with firm, industry, product/market characteristics, and with indicators of firm performance. Implications are that firms can be clustered on the basis behaviour patterns surrounding their internationalisation start-up and process. The characteristics and factors influencing internationalisation behaviour patterns were found to differ amongst firm clusters and specific groups of firms were identifiable by their international performance.

INTERNATIONAL GROWTH AND DEVELOPMENT OF SMALL FIRMS:
PATTERNS OF START-UP AND INTERNATIONALISATION

Introduction

Recent studies of small firm internationalisation suggest that a holistic, non-linear conceptual approach might be appropriate for entrepreneurial firms and those at early stages in their development [Bell et al 1998; Bell and Young 1998; Jones 1998]. Based on a qualitative, inductive, examination of chronological data on the cross-border activities of a sample of small, high technology firms Jones [1998] suggested that internationalisation, at the early stages of a firm’s life, is less of a functional activity than a process of growth and development. A process that may encompass any, or all aspects of the firm’s business. Taking this holistic perspective of international expansion, it is argued that small firms and in particular, small high technology firms, do not follow a prescribed set of steps or stages of export development, but rather, are likely to follow highly individual routes and methods of international expansion associated with their core business activities, over a time frame related to the individual firm’s rate of development and growth.

From that perspective, the research focus is expanded from an examination of entry modes, to include both in-coming and out-going cross-border activities, such as for example, importing, exporting, licensing, contract manufacture etc. undertaken by small firms (see key to Table 2). Internationalisation is considered here to represent organisational behaviour consisting of business decisions and international growth and development processes over time. Patterns of internationalisation may therefore be indicated by; rates of establishment of cross-border business activities, the rate of establishment of those activities, scope of activity in relation to the value chain, the inward or outward direction of cross-border business activities (links) formed, and the intensity, or number of distinct link types formed.

The purpose of this paper firstly, is to determine whether small firms follow specific patterns of internationalisation behaviour. To that end, the aforementioned indicators of internationalisation, are quantified and applied as operational constructs in an inductive, quantitative analysis of small firm internationalisation. Secondly, the purpose is to determine whether the internationalisation behaviour patterns identified through clustering of firms on the fore-mentioned indicators, are associated with firm characteristics, product and industry or market factors, and indicators of firm performance.

Conceptual Development of the Research Approach

Amongst theoretical explanations of the internationalisation of the firm, there are four approaches which are useful in offering some explanation of the process as it applies to small, young and entrepreneurial firms. These are: 1. The internalisation / transaction cost approach [Buckley and Casson 1976; Anderson and Gatignon 1986; Dunning 1988; Erramilli and Rao 1993; Caves 1996]. 2. Resource-based approaches [Teece 1977; Davidson and McFetridge 1985; Kogut and Zander 1993]. 3. Internationalisation and export development approaches [Johanson and Wiedersheim-Paul 1975; Bilkey and Tesar 1977; Cavusgil 1984; Johanson and Vahlne 1977, 1990]. 4. Network approaches [Mattsson 1985; Johanson and Mattson 1987, 1988].

The former two approaches generally offer explanation of specific, and often one-off or static internationalisation decisions, based respectively on firm assets / market imperfections, and resource based capabilities and constraints. The latter two approaches attempt to describe and explain the evolutionary process of internationalisation over time, based in part on incremental, often cost based decisions, and the development of relationships. None of these approaches alone provides an entirely satisfactory explanation of internationalisation as a growth and development process relevant to small firms. Each does provide a slightly different perspective or insight into how firms commence and develop their international business activity.

Examination of empirical literature specific to these four approaches suggests that rather than taking internationalisation as a process of firm growth, the focus tends to be more narrow and emphasises, in the same order as before, the internationalisation of production, technology transfer, exporting, and sometimes other forms of foreign market entry and from a similar marketing and distribution perspective, the expansion of, or entry into networks overseas.

Examination of the four approaches for points of common understanding suggested that, common to each, is the implicit understanding that the firm is a discrete entity with definite boundaries, and that growth is achieved by conducting business ultimately with external markets. In each of the approaches, business activity takes place through a variety of types of arrangements or links with the outside world. These links can for the most part, be divided into two types; external, i.e. transactional or contractual links, and internal, i.e. where part of the external market or value chain is internalised. Each approach treats links in different ways. Internalisation / transaction cost approaches tend to treat links as extensions of the production or value chain. Links are discretely, either external with contractual and transactional arrangements with organisations outside the firm, or internal and emphasise growth through integration, usually on the basis of extension of ownership [Buckley and Casson 1976; Dunning 1988; Buckley et al. 1990; Casson 1992]. Generally however, explanation extends to static decisions associated with growth objectives rather than a growth process over time.

Resource based approaches, currently gaining popularity in explanations of small firm internationalisation also treat links as part of the more widely defined value chain or innovation chain [Teece 1977; Davidson and McFetridge 1985; Kogut and Zander 1993]. Links with other individuals and organisations are identified as inward, outward and collaborative with the primary purpose of knowledge or technology sharing or transfer. Growth is seen as occurring as a result of the accumulation, exploitation and diffusion of knowledge. Internationalisation approaches, concerned with the process of market entry and development, are rooted for the most part in marketing and distribution studies, emphasise trading links, especially export, but a large and fragmented literature extends to other market entry modes including licensing and foreign investment. These studies tend to fall into two groups, step/stage export development [Bilkey and Tesar 1977; Cavusgil 1984], and internationalisation process [Johanson and Wiedersheim-Paul 1975; Johanson and Vahlne 1977, 1990]. Links are predominantly seen as outward, external and emphasise transactional exchange. Internal links in these approaches are associated with integration in distribution channels, and as with internalisation approaches, with ownership and control.

Network approaches, concerned again with the process, or behaviour of firms in relation to their international development, examine links, in the form of relationships or bonds, often categorised as economic, financial, technical, knowledge, and social bonds [Easton 1992]. Emphasis is again, predominantly on external links, but these may be inward, outward, reciprocal, cooperative, informal or formal contractual [Axelsson and Easton 1992]. Growth takes place through the extension of the firm’s network through investment in network positions and the development of network relationships [Mattsson 1985; Johanson and Mattsson 1987, 1988].

Despite their different perspectives, each of the theoretical approaches discussed, implicitly indicate external links as the main vehicle of, or means to growth. Examination of the types of cross-border business links made by firms is therefore an appropriate way to determine the extent and nature of a firm’s international activity. Taking a chronological dimension into account, as suggested by Andersen [1993], the firm’s international expansion activity over time i.e. its internationalisation behaviour, can be tracked and compared with that of similar firms. Together, the types of cross-border links formed, and the dates on which they were formed provide an empirical foundation on the basis of which both the process of internationalisation, and internationalisation decisions can be further examined.

There are a number of reasons for examining the internationalisation process of very small, very young, or entrepreneurial firms in this way. Firstly, small firms (independent firms of 200 or less employees), despite common constraints and problems in relation to growth [Buckley 1979], are heterogeneous and vary a great deal in relation to sector, industry role and position, the activities and services they perform, and in the way they are resourced and managed [Storey 1994]. Even amongst small firms which are classed as manufacturers, the balance of their business from activities such as R&D contracts and services, technology, production and marketing or distribution services and consultancy may vary considerably. Looking at the development of a small firm from start-up, it has been recognised that some technology based firms undergo a soft-start wherein they begin life providing a service based activity such as R&D and eventually harden into manufacturing firms. Similarly, manufacturers, over time, may concentrate on services or software as their competencies develop and competitive advantage or market conditions change. Taking this heterogeneity into account, it is likely that firms may commence international business by exploiting an activity based on competence, resource, or competitive considerations, a perspective more often associated with strategy driven multinational development than small firm internationalisation. Cross-border links formed are likely to reflect a range of business activities, goals, aspirations and imperatives which are as likely to be multi-purpose as they are to be connected to a single functional activity such as export sales.

Secondly, very small, young, and entrepreneurial firms tend to be holistic entities rather than collections of functions or parts. Generally, there is an individual or small team of owner/managers well informed about the whole organisation. These people are likely to report a more holistic perspective of the firm’s international activities than, for example, export managers in larger, more structured organisations.

Thirdly, the starting point of international activity for small firms in not well known. It is broadly assumed that international activity will commence with export, or at least with a low cost, low risk, mode of market entry. Also, research has established a range of reactive and proactive triggers or motivators, usually in relation to first export sales, and unsolicited orders and managerial impulse are frequently recognised as responsible for initial steps. Explicit categorisation of the type of cross-border business activity undertaken by firms, either at commencement, or during the process of internationalisation is scarce in the literature. Extending the latter argument, different types of business activity are not necessarily mutually exclusive, production and sales arrangements may be associated or converge, products may be sold through different channels or intermediaries simultaneously, and import and export deals may be reciprocal. With a conceptual focus on firm growth and development, compound as well as discrete business arrangements need to be examined.

Fourthly, the globalisation of business in general, and of technology in particular, has exposed small firms to approaches from overseas based organisations and to the activities of foreign firms in their home markets. Business arrangements made with overseas firms extending their activities into the UK market are therefore important as components of the overall internationalisation process of UK based small firms.

Fifthly, the categorisation of small firms on the basis of international activity has predominantly emphasised the extent of their export activity based on steps and stages of export development [Crick 1995]. Recent exploratory research has indicated that while small firms may go through stages of development [Luostarinen and Hellman 1994; Luostarinen et al 1994], these do not necessarily follow the conventionally accepted export development route but may differ in organisational factors such as the speed of market entry, the types of business activities involved and the combinations and sequences of cross-border business activities undertaken over time [Oviatt and McDougall 1994; Jones 1998].

From the perspective of theoretical development, small firm studies in general have been criticised for a widespread failure to utilise or develop theory. Development of appropriate theory of small firm internationalisation requires attention to be paid to the characteristics of small firms other than scale [Wynarczyk et al 1993], establishment of an appropriate level for the research focus within a specific context [Toyne and Nigh 1998], and the delineation of assumptions (e.g. value, scope and time) [Andersen 1993].

Objectives and Assumptions

The objectives of this study are:

  1. To determine the viability of this research approach for survey research on small firm internationalisation through
  1. To determine the extent to which firm, market, and industry characteristics are specific to patterns of internationalisation behaviour.

The objectives are based on a number of assumptions drawn from the conceptual argument presented above, these are: 1. that internationalisation is part of the overall growth and development process of the firm, 2. that internationalisation consists of both processes and decisions which, over time, will form patterns of activity, and 3. that patterns of activity can be tracked through chronological examination of a firm’s establishment of international links, 4. That firms may be grouped according to patterns of international activity (internationalisation behaviour).

Methodology

The main data collection device was a structured mail questionnaire. The questionnaire was tested for construct and content validity through examination by, academic experts in the fields identified in the literature , experts in research methods and industry members and government officials involved in small high technology firms and sectors. A pilot run of 50 firms was conducted and the results analysed. The questionnaire was modified on the basis of those opinions and analysis of the pilot run results, and mailed to a further 1001 small firms (excluding the 50 targeted in the pilot phase), selected by SIC code from four high technology sectors: plastics and composites, biotechnology, advanced surgical instruments, and advanced electronic instruments for industry. Questionnaires were addressed to individuals, by name, identified as the owner or chief executive of the firm. Reminders were sent out after two weeks. A total of 117 usable questionnaires were received from the first mailing, and 96 from the second. There were no significant differences between responses to key questions across the two mailings.

As this study is of internationalisation in its broader sense, the choice of sampling frame was considered to be of some importance. The sample was selected from a data base constructed from directories of firms other than export directories to avoid a bias towards export based internationalisation. The criteria for selection of sample firms were that they should belong to one of the listed sectors based on their SIC classification, and should have less than 50 employees at the time of listing. Assuming that firms were likely to have grown after registering with the data base, the ultimate cut-off point was 200 employees. Most interest was in small and young firms but it was considered important to allow a reasonable margin for growth. Not including those firms would eliminate fast growth firms from the sample and bias the results towards slower growth firms. Completed returned questionnaires were included in the analysis if: they had less than 200 employees (if expansion to that size took 10 years or less), were a manufacturer, or a firm under 5 years old, and could be classed as high technology based. In the latter case, firms were included if they had high technology products or processes, or some involvement in either R&D and/or innovation.

The incidence of international involvement was not used as a screening criterion and therefore the sample included both firms which had, and which had not established international activity. A number of firms had gone out of business which reduced the sample size to 860. The resulting usable response was 213 firms representing a 25% response rate. Of these, 196 were able to provide a full chronological account of their international linkage activities. The credibility of respondents in terms of corporate memory in the sample would seem to be high as almost 70% indicated that they were one of the founders of the firm, and over 85% indicated that they were the chief executive. In very small firms, as in this sample, management tends to be centralised and vested in a few individuals who are likely to be knowledgeable about all aspects of their firm’s development. Examination of the questionnaires indicated that respondents were familiar with the historical development of the firm and where they were not, indicated that they were unable to supply the information. The sample profile is included as an endnote.1

Operational Constructs and Analytical Procedures

Cross-border links in the form of activities which take place across national borders, or alternatively, between partners located in different countries, were selected from extant studies on internationalisation, on the growth and development of small firms, and from technology and innovation studies which focus specifically on firm growth and development. The resulting list of distinct cross-border link types (see key to Table 2) includes business activities associated with each of the three main value chain activities, R&D, production and marketing/distribution. Cross-border links formed through business coming in from abroad (i.e. import and contract work for overseas based firms), and going out (i.e. export, contracted out work, and investment overseas) are included in the list. In addition the link types distinguish between external/transactional links and internal, integrated or investment links.

Information elicited from the respondents resulted from questions asking the specific date on which the firm made its first cross-border link in each business activity listed. This data, together with the foundation date of each firm provided ratio scale measures, for each firm, of the time taken to commence internationalisation, the rate at which new types of cross-border business activity were established, and the concentration or intensity of link forming activity at any specific point in time. The data was sorted chronologically on a firm by firm basis, and systematic qualitative analysis determined that patterns in internationalisation behaviour were based on time-scales unique to individual firms. Firm internationalisation patterns consisted of combinations or sequences of cross-border business formed over time [Jones, 1998]. The patterns established by the qualitative analysis required conversion to numeric format to allow content based patterns to be established.

Table 1 Internationalisation: Clustering Variables

Rate

Time to first event1

Average time between events2

Scope

Marketing, R&D and Production combinations3

  • One VC type only (R or M or P)
  • All of M and P and R&D
  • Combinations of two value chain types

Direction

Inward and outward4

Inward only

Outward only

Intensity

Average number of links at the first event5 (start-up)

Total number of links over 4 events6 (internationalisation process)

  1. Length of time in years between the foundation date of the firm and the establishment of the first cross-border link.
  2. Length of time in years (gaps) between specific years (events) during which cross-border links are established.
  3. Any cross-border link or combination specific to one of the three main value chain activities, R&D, production, and marketing / distribution.
  4. The direction of the cross-border link determined by the extension of the firm’s value adding activity overseas, or its retention in the domestic market.
  5. The number of distinct cross border link types formed at the first year in which they were established.
  6. The average number of distinct cross border link types formed over the four or more events examined in the study.

 

The clustering variables (Table 1) were determined by: 1. The rate at which firms commence international business activity and diversify through various forms of international business activity, 2. The scope of international business activity in terms of its value chain coverage, 3. The direction of cross-border activity, and 4. The intensity of cross-border business activity measured by the number of distinct cross-border link types at the first occasion (event) at which firms establish links, and again over time. The time element was variable by firm. The study accommodated the variable time periods but restricted the analysis to the first four occasions (events) at which each firm established a new type of links.

The clustering scheme followed was an adaptation of that in Hair et al (1998). This method was used for both the start-up and the process (four-event) data view. The variables chosen to cluster (Table 1) were all standardised to lie in the range 0-1. This gave each variable equal influence on the outcome of the clustering. Ward's method of hierarchical clustering analysis was used to generate solutions between two and twenty clusters. These solutions were used as starting points for Hair et al’s [1998] split-half k-means re-allocation clustering. This involved analysing one half with reallocation, using this solution as starting points for the second half analysed with and without reallocation: these two analyses were then compared with Kappa coefficients. In both pattern sets a five-cluster solution was among those that gave the highest values of kappa. It was an acceptably high number of clusters to facilitate interpretation. A final k-means clustering was then done for the complete sample. The cluster solutions suggested that firms could be appropriately clustered based on:

  1. Internationalisation Start-Up Patterns, i.e. the time taken to establish first cross-border links together with their scope, direction and intensity.
  2. Internationalisation Process Patterns, i.e. the rate of link formation over time together with the scope, direction and intensity of that activity over time.

The two resulting sets of clusters (Table 2 and Table 5) were then subjected to bi-variate analysis with firm, market, industry and effect (performance) variables to establish a preliminary picture of relationships between patterns of internationalisation behaviour and internationalisation performance, and to identify any influencing factors from firm characteristics and/or market/industry conditions (Tables 3 and 4, and Tables 6 and 7).

Results

Internationalisation Start-Up Patterns

The five clusters of firms based on internationalisation start-up patterns (Table 2) were named as follows:

OIS firms (Table 2) ranked second in their speed to commence internationalisation with an average period of just under 4 years before cross-border activities began. This group contained 29 firms all of which had commenced internationalisation activity with more than one value chain link type. Seventy six percent of OIS firms were found to have established cross-border links in marketing and production, and a further six percent had combinations of marketing and research and research and production. Almost a fifth of OIS firms had links in all three value chain types. All OIS firms had established both inward and outward links, not necessarily of the same value chain link type. None of the OIS firms had either unidirectional links or links restricted to one type of value chain activity. The composition of activity at commencement for OIS firms was not only complex but also relatively intense, with an average of 4 distinct types of cross-border activity per firm at commencement.

Table 2: Firm Clusters: Internationalisation Start-Up Patterns

 

Clusters

Clustering Variables

OIS

RIS

TES

DFS

EIS

Years to commence internationalisation

4

6

8

5

2

Rank speed to commence internationalisation

2nd

4th

5th

3rd

1st

Number of value chain activity types (M,R,P)1

2-3

1~

1~

2-3

1~

Direction of Link (I/O)2

I+O~

I~

O

I or O

I+O~

Intensity (number of link types)3

4

1

1~

3

3

Rank intensity

1st

4th

4th

3rd

2nd

Number of firms

29

80

47

16

20

Key

 

1 M = Marketing,

R = R&D,

P = Production

2 I = Inward Cross-border link

O = Outward cross-border Link

~ = Universal amongst cluster firms

3 Distinct Link Types

  • Import from overseas supplier
  • Import plus distribution in the UK
  • Marketing or management service/ consultancy for overseas firm in the UK
  • License-in technology from overseas
  • R&D performed under contract in UK for overseas based firm
  • Manufacturing performed under contract in UK for overseas based firm
  • Technical service/consultancy performed in the UK for overseas based firm
  • UK based export
  • Overseas agent/distributor export
  • Integrated export (own overseas reps. or branch)
  • Management/marketing consultancy performed overseas
  • License out technology to overseas licensee
  • R&D contracted out to overseas based firm
  • Manufacturing contracted out to overseas based firm
  • Technical service/consultancy performed overseas
  • Investment in overseas production

RIS firms (Table 2) were relatively slow to commence internationalisation and ranked fourth amongst the five internationalisation start-up clusters identified, taking around 6 years to establish their first cross-border link. Links formed by RIS firms were universally of one value chain type and, again universally, links formed were found to be in the inward direction only. The intensity of linkage formation for RIS firms was low with an average of only one type of cross-border activity in the first year of international activity.

TES firms (Table 2) were the slowest to commence internationalisation with an average time of 8 years lapsing before the first cross-border link was formed. Similar to RIS firms, this group consisted of firms which formed only one type of value chain at commencement and again the direction of internationalisation was universally unidirectional but this time, in an outward direction. The intensity of link formation for TES firms was low with an average of one distinct link type in the first event.

DFS firms (Table 2) ranked third in their speed to commence internationalisation, taking on average 5 years to do so. In sharp contrast to RIS and TES firms, this group tended to form links in more than one value chain activity with almost 50% having combinations of marketing and production, 20% with marketing and research and 20% with all three. The direction of activities was interesting in that no firms had activities in both directions, and other than those which specifically stipulated having collaborative arrangements, activity seemed to be unidirectional but in either direction. The intensity of links formed by DFS firms indicated an average of just under three link types at commencement, and ranked third across the clusters.

EIS firms (Table 2) were the fastest to commence international activity and on average took less than two years to do so. All EIS firms commenced international activity with one value chain link type only, but interestingly, all had formed links in both inward and outward directions. The EIS group of firms ranked second in the intensity of link formation and on average established three distinct link types in the first event. OIS,RIS and TES firms showed a great deal more variability, within clusters, in the time taken by member firms to form their first cross-border links than did DFS and EIS firms, based on measures of standard deviation. The intensity of link formation on the other hand showed more variability within OIS and EIS clusters than in the RIS, TES, and DFS clusters.

Internationalisation Start-Up Patterns: Firm Characteristics, Product Market Factors and Performance

Differences in the variables affecting the composition of firms in the internationalisation start-up clusters was indicated by means of bi-variate analysis consisting of Chi-square tests for categorical variables and F tests for continuous variables. The results are listed in Tables 3 and 4.

Table 3 Internationalisation Start-Up Patterns: Firm Characteristics

Cluster

OIS

RIS

TES

DFS

EIS

 

%

%

%

%

%

Chisq

Sig.

Firm Characteristics

NTBF

48

34

45

44

35

2.86

.582

UKIPRs

69

63

62

43

57

1.43

.838

Overseas IPRs§

54

52

38

29

57

2.41

.661

New Start

29

71

83

50

50

27.45

.000***

Independence§

66

81

81

69

70

4.46

.348

R&D department§

59

36

40

63

50

7.21

.125

Export department§

4

17

15

29

42

12.68

.013**

Cooperative R&D§

52

35

26

56

35

8.07

.089

Contacts &Links

Foreign nationals

11

10

9

43

5

14.44

.006**

Foreign languages

32

23

22

50

35

6.13

.189

Overseas education

21

20

17

44

22

5.18

.269

Worked Overseas

61

48

54

56

60

1.91

.752

1. New technology based firm, i.e. one established specifically to exploit a scientific or technological innovation.

2. Intellectual property rights, based on the firm’s application for protection.

3. Firm with no previous corporate existence.

4. Existence of founders with this attribute at firm start-up.

§ Dichotomous variable

*** Sig. at £.001; ** Sig. at £.01; * Sig. at £.05

Amongst firm characteristics, four were found to differ significantly between clusters, these were; whether or not the firm was a new start rather than one which evolved from a previous organisation, whether the firm had been involved in cross-border collaborative R&D, the existence of foreign nationals amongst its founders, and whether or not it had an export department (Table 3). The results are interesting. The TES cluster was dominated by firms which classified themselves as "new starts", i.e. they had no previous corporate existence. This group was slow to commence internationalisation and although internationalisation was outward only, it was normally confined to one value chain link only at the start. By contrast, OIS firms, with less than a third of its member firms claiming to be new starts, commenced internationalisation more quickly with a more complex and more intense bi-directional first year, consisting of more than one value chain activity. This suggests that the start-up pattern of international activity is associated with the firm’s evolution from another organisation. The most likely explanations, from the literature, are experience, and pre-established links and contacts at the firms’ foundation.

Surprisingly, the two clusters in which firms are most likely, and least likely to have been involved in cross-border R&D cooperation, were relatively slow to commence internationalisation. The TES cluster tended to form outward links only, and these limited to one value chain link and with low intensity. The DFS cluster, on the other hand, was more likely to have been involved in cross-border R&D cooperation, experienced a slow start to internationalisation but with more complex value chain combinations, and a greater intensity of link types. Amongst DFS firms, it would appear that cooperative R&D has been the trigger for internationalisation but took place, on average, a few years after the foundation of the firm. In terms of the existence of a dedicated export department, the EIS cluster contained the highest proportion of firms with export departments and this group were fastest to commence internationalisation. Interestingly, OIS firms, the second fastest group, contained fewer firms with export departments. What is not clear here, and also in the case of cross-border cooperative R&D discussed above, is whether the Export department was established at start-up, or at a later date.

Table 4 Internationalisation Start-Up Patterns: Product/Market Factors and Performance.

Cluster

OIS

RIS

TES

DFS

EIS

F

Sig.

Product & Market / Industry Factors

Specialisation R&D1

20

11

14

28

16

2.67

.034*

Specialisation Production. 1

48

50

50

39

43

.78

.542

Specialisation Mktg. 1

15

19

17

17

19

.47

.757

Technology2

3.74

3.51

3.41

3.94

3.68

.82

.512

Innovation2

3.46

3.41

3.73

3.63

3.53

.44

.777

Software2

2.29

2.29

2.21

2.07

2.50

.21

.933

Product complexity2

3.25

3.16

3.13

3.56

2.90

.90

.465

Standardisation2

3.18

3.16

3.33

2.94

3.84

1.71

.149

Customisation2

4.21

3.70

3.52

4.33

3.42

2.29

.061

Industry Applications2

3.04

2.68

2.84

3.13

3.00

.49

.741

UK Substitutes2

2.82

2.82

2.48

2.68

1.90

2.17

.074

R&D intensity3 (Turn)

12

14

14

25

13

1.11

.352

R&D Intensity4 (employees)

2.39

2.49

2.45

2.79

2.50

.49

.743

Effect Variables

Age at survey date

12

17

20

12

13

1.65

.165

Size (employees)

35

30

31

22

26

.66

.619

Size (turnover)

2.45

2.24

2.04

1.81

2.50

2.62

.036*

Turnover Growth5

43

20

36

19

26

.823

.512

Export Countries

18

10

17

12

20

1.87

.118

Export Value

703

544

487

649

627

.287

.886

Export Growth6

67

21

30

16

40

1.84

.125

International Ratio7

42

26

29

41

42

2.92

.022*

1. Ratio of staff employed in this function to total employment.

2. Five point interval scale from 1 = low, to 5 = high

3. Ratio of annual investment in R&D to total turnover.

4. Ratio of annual employment in R&D to total employment.

5. Average change in turnover over the 3 years prior to the survey.

6. Average change in exports over the 3 years prior to the survey.

7. Ratio of all income generated from overseas, to total turnover.

*** Sig. at £.001; ** Sig. at £.01; * Sig. at £.05

 

In the second group of variables, which includes product and industry factors, and the firms’ current age, and level of investment in R&D, only three variables were found to differ significantly between firm clusters (Table 4). These were, proportion of staff employed in R&D, and approaching significance at 95% were, the extent of customisation of products to customer specifications, and the availability of substitute products in the UK market. Firms with the highest proportion of staff in R&D were found amongst DFS firms, which in terms of their initial international activity, were relatively slow to start, and concentrated on outward activity across a couple of value chain activities. It is interesting to note that R&D intensity, whether measured by investment or by employees, was not a significant factor in relation to any of the clusters.

Performance measures produced more interesting results (Table 4). Firms with high turnover tended to found within the EIS group, the cluster with the fastest rate of commencement, with internationalisation activity in both directions, but concentrated on one value chain type. By contrast, the lowest turnover firms were found in the DFS group, with a less rapid commencement speed and a more directionally focused, but functionally generalised approach to their cross-border activity. This would suggest that a fast, focused entry into international business has higher subsequent returns in terms of turnover. The size of the firm, measured by employees, did not show any significant difference between clusters. The international ratio was highest in the OIS cluster where firms experienced a fast, intense start to international activity, with links in both directions and across several value chain activities. It was lowest in the RIS cluster, where firms experienced a slow, inward, unidirectional start, which due to the universally inward direction of the links in that cluster, would be expected.

Although several firm characteristics, product/market factors and performance factors showed significant results across clusters, no one cluster stood out as unique as regards a number of combination of factors. Further examination through multivariate methods might determine composite and interaction effects.

Internationalisation Process Patterns

Firms were reclustered based on their patterns of internationalisation over variable periods of four or more time-based events following firm foundation. Most firms in the sample indicated, on average, four events. The first three events were determined discretely, while the fourth, where relevant includes the fourth and any subsequent events. The five clusters of firms based on internationalisation process patterns (Table 5) were named as follows:

Table 5: Firm Clusters: Internationalisation Process Patterns Over Time

 

Clusters

Clustering Variables

INV

IBS

ISH

IEN

ISSI

Years between internationalisation Events (average)

4

3.8

9

3.7

7.7

Rank speed of internationalisation

3rd

2nd

5th

1st

4th

Number of value chain activity types (M,R,P)1

3~

2*~

1,2 or 3

1-2

1~

Direction of Link (I/O)2

I+O~

I+O

I

I+O~

O~

Average Intensity (number of link types per event)3

1.87

1.23

.58

.98

.38

Rank intensity

1st

2nd

4th

3rd

5th

Number of firms

56

49

38

32

18

Key as Table 2

* combinations of M & P

 

The INV group, clustered by internationalisation process pattern, consisted of 56 firms (Table 5). The average rate of internationalisation for INV firms was just under four years measured as the average period of time between events. The INV group exhibited the third fastest rate of internationalisation across the five clusters, although there was very little difference between the rates of the INV, IBS and IEN clusters. Firms in the INV group ultimately formed links across all three value chain activities and links were formed in both inward and outward directions. This group exhibited the highest intensity of link formation with on average 1.87 links per event.

The IBS cluster (Table 5) rated second fastest in the firms’ rate of internationalisation with just under 4 years between events. Interestingly, all IBS firms had formed links in marketing and production, none had specialised in a single value chain activity, and none had formed R&D links of any type. Almost all (98%) firms in the IBS group had formed links in both inward and outward directions and 2% had formed links in an outward direction only. IBS firms exhibited the second highest intensity of link formation with an average of 1.23 links per event. In total there were 49 firms in this group.

The ISH cluster (Table 5) consisted of firms which were relatively slow to internationalise and in fact this was the slowest cluster with a mean period of 9.15 years between each year of link establishment. In terms of value chain activity, there was no clear pattern and ISH firms represented a hybrid assortment of activity types although just over half of the group reported having only one value chain activity type in cross-border links. All other value chain types are included in the ISH group with marketing and production combinations accounting for 24% of the group. Internationalisation activity amongst ISH firms was almost totally inward (95%) and the intensity of linkage formation was the second lowest across the clusters with an average of .58 links per event. The group accounted for 38 firms.

The IEN cluster (Table 5) containing 32 firms, exhibited the highest rate of internationalisation across the four events with an average of 3.74 years between each year of link establishment. IEN firms appeared to specialise in one type of value chain activity only in relation to cross-border links formed. Just under 70% of the sample formed one type of value chain link only. Of some interest is that almost 30% of IEN firms formed combinations of marketing and research based links. Firms in the IEN group universally formed links in both inward and outward directions with an average intensity of .98 links per year of link establishment.

The ISSI cluster (Table 5), which contained only 18 firms, exhibited a slow rate of internationalisation with 7.69 years on average between years of link establishment. Universally, the ISSI group formed value chain links of only one type, and again universally, these were in an outward direction. The intensity of linkage formation was very low with only .38 links per event. Firm clusters ISH and ISSI showed a great deal more variability in the time taken by member firms to form their first cross-border links than did firm clusters INV, IBS and IEN, based on measures of standard deviation. The intensity of link formation for the ISH an ISSI clusters, on the other hand, showed very little variability with low standard deviation scores for each cluster on that measure.

Internationalisation Process Patterns: Firm Characteristics, Product Market Factors and Performance

Bi-variate analysis of internationalisation process pattern clusters proved to be more informative and interesting than for internationalisation start-up clusters. Significant results were found between clusters in three firm characteristic variables (Table 6). These were; the number of firms which were classified as new starts, those with an R&D department, and those which had undertaken cross-border, cooperative R&D projects. Approaching significance, were those which had made application for overseas intellectual property rights (IPRs). Firms were most likely to be new starts in the ISSI group, and less likely the INV and IBS groups. Firms with an R&D department, and with involvement in cross-border cooperative R&D were predominant amongst INV firms.

Table 6 Internationalisation Process Patterns: Firm Characteristics

Cluster

INV

IBS

ISH

IEN

ISSI

%

%

%

%

%

Chisq

Sig.

Firm Characteristics

NTBF

43

43

21

50

39

7.39

.117

UK IPRs

78

48

50

56

71

5.35

.253

Overseas IPRs§

70

29

38

50

29

8.87

.064

New Starts

56

53

68

72

89

9.89

.042*

Independence§

70

71

74

91

89

7.34

.119

R&D department§

66

41

26

38

33

17.47

.002**

Export department§

20

20

18

19

6

1.99

.738

Cooperative R&D§

57

31

32

28

17

15.39

.004**

Contacts &Links

Foreign nationals

16

11

19

3

0

7.79

.100

Foreign languages

42

21

25

22

22

7.10

.131

Overseas education

25

28

19

19

6

3.94

.414

Worked overseas

69

49

46

47

47

7.48

.113

Key as Table 3

*** Sig. at £.001; ** Sig. at £.01; * Sig. at £.05

 

Variables relating to product/market factors which showed significant results across the clusters were, level of specialisation in R&D, level of customisation of products, and the availability of substitutes in the UK market (Table 7). Approaching significance was the level of technological innovation input by the firm. Significant performance or effect variables were, the size of the firm measured by turnover, the number of export countries supplied by the firm, the value of exports, and the international turnover ratio. None of the explicit firm growth measures showed any significance across clusters, neither did firm age, nor size measured by the number of employees. The INV cluster was particularly interesting in that while it was not the fastest group of firms to internationalise, this group had significantly higher export turnover, international ratio, number of export countries, was more likely to customise products, to have made a technological innovation, and in comparison to all other clusters except one, to have a relatively high number of substitute products in the UK

Table 7 Internationalisation Process Patterns: Product/Market Factors and Performance

Cluster

INV

IBS

ISH

IEN

ISSI

F

Sig.

Product & Market / Industry Factors

Specialisation R&D1

24

10

13

13

11

3.43

.010*

Specialisation Production 1

47

50

50

45

51

.29

.885

Specialisation Mktg.1

16

17

21

20

16

.76

.553

Technology2

3.79

3.63

3.26

3.50

3.56

1.20

.314

Innovation2

3.67

3.47

3.06

3.63

4.13

2.24

.067

Software2

2.23

2.20

2.03

2.82

2.13

1.39

.239

Product complexity2

3.29

3.22

3.17

2.90

3.21

.66

.619

Standardisation2

3.38

3.16

3.23

3.27

3.19

.24

.918

Customisation2

4.00

3.84

3.91

3.00

3.87

3.29

.012*

Industry Applications2

2.82

2.96

2.82

2.53

3.06

.44

.779

UK Substitutes2

2.56

2.51

3.4

2.24

2.31

4.08

.003**

R&D intensity (Turnover)3

18

11

16

13

12

.80

.524

R&D Intensity (employment) 4

2.68

2.26

2.39

2.62

2.54

1.66

.162

Effect Variables

Age at survey date

17

17

16

13

19

.412

.800

Size (employees)

37

30

25

27

20

1.71

.149

Size (turnover)

2.30

2.31

2.05

2.39

1.72

2.63

.036*

Turnover Growth5

20

29

23

42

39

.64

.633

Export Countries

20

14

5

13

14

3.45

.010*

Export Value

821

672

286

455

212

2.77

.029*

Export Growth6

35

35

13

45

11

1.01

.407

International Ratio7

42

36

17

32

23

5.09

.001**

Key as Table 4

*** Sig. at £.001; ** Sig. at £.01; * Sig. at £.05

 

The INV group was also more likely to have been involved in cross-border cooperation, to have an R&D department, to have applied for intellectual property rights overseas, and significantly, were less likely to have been "new-start" firms. Of interest is that firm size, age, and R&D intensity and overseas connections and contacts were of no significance across clusters. Summarising the pattern of internationalisation exhibited by INV firms, this group exhibited relatively fast internationalisation, which included all or most of the value chain link types, in both inward and outward directions and with high intensity in the number of distinct link types formed. The INV group were the most successful in terms of the value of exports, international ratio, and geographical spread of exports.

Discussion and Conclusions

The primary objective of this paper was to empirically examine the internationalisation process of small firms, over time from the date of foundation to the survey date. The establishment of cross-border links relating to a range of business activities associated with the key value chain activities was the means by which the internationalisation process was tracked. Cross-border links established by the firms were considered to represent specific internationalisation decisions, and in that respect the direction of the activity, inward, or outward, was recorded, as was the form of the link, i.e. external/transactional or internal/integrated. The pattern formed by the intensity of activity, the combination of links and the time period over which it took place, was taken to represent individual firms’ internationalisation behaviour, i.e. organisational behaviour as regards the decisions, processes and patterns of internationalisation. Cluster analysis suggested that there were two dimensions to internationalisation behaviour and that these should be treated separately. The first was the pattern of activity at the start-up of internationalisation, and the second was the pattern of activity which emerged over time during the process of internationalisation.

Internationalisation start-up clusters were found to distinguish clearly between firms on the basis of clustering variables and separated firms with inward links only (RIS, Table 2) and firms with predominantly outward links only (TES, Table 2). In addition, those two groups tended to specialise in only one type of value chain activity in relation to their international business. The speed of commencement of international activity appears to be associated with intense link forming activity at start-up as firm clusters OIS and EIS (Table 2) suggest early commencement, coupled with a range of distinct business activities. The main difference between Clusters OIS and EIS is that firms in the EIS group tend to specialise in one value chain activity while those in the OIS group involve more aspects of their business.

Clusters based on Internationalisation Process Patterns over time (Table 5) again distinguished two clusters in which firms were involved in either inward, or outward cross-border business activity, but not both (ISH and ISSI respectively). There was no clear association between the speed of internationalisation and the intensity of links formed over time. Looking across the clusters on both Internationalisation Start-Up Patterns, and Internationalisation Process Patterns, suggest firms are differentiated in their internationalisation behaviour on the basis of speed, functional diversity and concentration of activity. In addition, firms which confine themselves to either inward, or outward internationalisation are identified and grouped through the clustering method.

Turning to the second objective, variables which were not indicated as significant across clusters were of some interest. Firm age and size (employees) were not significant and overseas contacts and experience, i.e. overseas work experience, overseas education and language ability were of no significance either at commencement or during the process of internationalisation, although the incidence of foreign founders was significant for commencement. Turning to the significant results, of some importance was the finding that the internationalisation process clusters included one group (INV, Tables 6 & 7), which appeared to be more successful than the firms in the other clusters based on several performance measures, had experienced a relatively rapid internationalisation process utilising a range of distinct business activities and accommodating all aspects of the value chain. In addition, firms in the INV group, universally, were involved in both inward and outward cross-border business activity. The similarity of INV firms to global start-ups or international new ventures identified by Oviatt and McDougall [1994], and international entrepreneurs [Jones 1998] as firms venturing into international business, with apparent significant involvement and commitment at early stages in their development, is striking.

Also of importance is that some of the variables identified as significant across clusters included, involvement in cooperative R&D, level of customisation of products, level of innovation, specialisation in R&D, and application for overseas IPRs suggest specific strategy in relation to firms’ cross-border activity.

Overall, the findings indicate that internationalisation, at least in the early stages of a small firm’s development, tends to be a holistic process extending across a range of cross-border business and value chain activities. Contrary to receive theory, findings indicate that internationalisation may begin with any, or a combination of business arrangement encompassing any or several of the main value chain activities, R&D, Production and marketing/distribution. In support of the view in both internalisation and export development approaches, the process does tend to begin with lower cost, lower risk approaches, at least to the extent that investment modes occur later in the process. The results also reflect the internationalisation and network view of international expansion as an evolving process although evidence here suggests that the evolutionary process is triggered by and interspersed with decisions concerning cross-border links, based, in the early stages, on contractual arrangements with overseas contacts. The internationalisation processes identified here suggest periods of exponential as well as incremental development, together with periods with little or no additional international activity. The sometimes complex combination and bi-directional nature of links established by some firms at specific points in time would tend to support the resource-based view that firm growth, and hence internationalisation is based on accumulation as well as exploitation of firms assets, competencies, capabilities and knowledge.

Finally, the heterogeneity of the small firm sector into account, and accepting that small firms’ competitiveness is based on bundles of competencies, capabilities and resources, assembled in unique ways by entrepreneurs or management teams, it makes sense to examine internationalisation in a similarly holistic way in order to take these issues into account. Establishment of the link between composite internationalisation behaviour and the development, growth and performance of small firms has important implications for managerial and government policy where the concern is with the survival, growth and development of the firms.

The conceptual and methodological approach developed here from retrospective data represents an appropriate research design for a truly longitudinal study of small firms beginning at their inception and tracking their international development as it takes place. This study focused on high technology based firms of less than 200 employees, in specific high technology sectors, and recorded and examined the first link of each type of cross-border business activity. Future research might usefully expand these parameters to other industry sectors and countries. In addition examination of the continuity and value of cross-border links, and multivariate rather than bi-variate analysis would be worthy developments of the research design. The contribution made by this research is in its attempt to establish an appropriate conceptual framework and empirical basis from which to examine the internationalisation process of firms, from start-up, over time.

 

Notes

1. Sample Profile

Size of Firm (employees) n = 213

Less than 10

11 to 20

21 to 50

51 to 200

f

56

52

78

27

%

26

25

37

13

Age of Firm n = 212

New (5 years or less)

Young (6 to 10 years)

Adolescent (11 to 25 years)

Mature (26 years and over)

42

53

80

37

20

25

38

18

Mode of Foundation n = 212

New start up

Evolution

135

77

64

36

R&D Intensity. (% of turnover 1993) n=181

Zero

1% to 5%

6% to 20%

21% to 100%

20

84

49

28

11

47

27

16

R&D Ints. (% of employees 1993) n=182

Zero

1% to 10%

11% to 20%

21% to 100%

51

62

32

37

28

34

18

20

Industry Sector n = 212

1. Plastics and Composites

2. Biotechnology/Pharmaceuticals

3. Adv. medical instruments/appliances

4. Electronic equip/instruments for industry

5. Other

22

34

48

54

54

11

16

23

25

25

 

 

References

Andersen, Otto. 1993. On the internationalisation process of firms: a critical analysis. In Journal Of International Business Studies, 24(2): 209-231.

Anderson, Erin and Hubert Gatignon. 1986. Modes of foreign entry: a transaction cost analysis and proposition. In Journal Of International Business Studies, 17(3): 1-26.

Axelsson, Bjorn. and Geoffrey Easton. 1992. Industrial networks: a new view of reality, Routledge, London.

Bell, James, David Crick. and StephenYoung. 1998. A holistic perspective on small firm growth and internationalisation. In Academy of International Business Conference, proceedings. City University Business School.

Bell, James. and Stephen Young. 1998 Towards an integrative framework of the internationalisation of the firm, in G. Hooley, R. Loveridge and D. Wilson (eds), Internationalisation: Process, Context and Markets, London: Macmillan, pp

Bilkey, Warren.J. and George Tesar. 1977. The export behaviour of smaller-sized Wisconsin manufacturing firms. In Journal Of International Business Studies. 8(1): 93-98.

Buckley, Peter.J. and Mark Casson (1976) The future of the multinational enterprise, Macmillan, London.

Buckley, Peter J. 1979. Foreign investment success for smaller firms. In Multinational Business, 3:12-19.

Buckley, Peter J., C.L. Pass, and Kate Prescott. 1990. Foreign market servicing by multinationals: an integrated approach. In International Marketing Review, 7, (4) pp25-40, as reprinted in Buckley, Peter and Pervez Ghauri , editors. 1993. The internationalisation of the firm: a reader, Academic Press, London.

Casson, Mark. 1992. Internalisation theory and beyond. In , Peter J. Buckley, editor. New directions in international business: research priorities for the 1990s, Edward Elgar, Aldershot: 4-27.

Caves, Richard E. 1996. Multinational enterprise and economic analysis. (2nd ed). Cambridge University Press, Cambridge Mass..

Cavusgil, S.Tamer. 1984. Differences in exporting firms based on their degree of internationalisation. In Journal of Business Research. 12:195-208.

Crick, David 1995. An investigation into the targeting of UK export assistance. In European Journal of Marketing, 29 (8), 76-94.

Davidson, William H. and Donald G. McFetridge. 1985. Key characteristics in the choice of international technology transfer mode. In Journal of International Business Studies. 16(2): 5-21.

Dunning, John.H. 1988. The eclectic paradigm of international production: a restatement and some possible extensions. InJournal Of International Business Studies, Spring: 1-31.

Easton, Geoffrey. 1992. Industrial networks: a review, in Bjorn Axelsson, and Geoffrey Easton, Industrial networks: a new view of reality: 3-26.

Erramilli, M. Krishna & C.P.Rao. 1993. Service firms’ international entry-mode choice: A modified transaction-cost analysis approach. Journal of Marketing. 57(July): 19-38.

Hair, Joseph F., Rolph E. Anderson, Ronald L.Tatham and William C. Black (1998) Multivariate data analysis, 5th Ed., Prentice Hall, New Jersey.

Johanson, Jan. and Lars-Gunnar Mattsson. 1988 Internationalisation in industrial systems-a network approach. In Hood, Neil and Jan-Erik Vahlne, editors. Strategies in Global Competition. Routledge, as reprinted in Buckley, Peter and Pervez Ghauri, editors. 1993 The internationalisation of the firm: a reader. Academic Press, London: 303-321.

Johanson, Jan. and Lars-Gunnar Mattsson. 1987 Interorganisational relations in industrial systems: a network approach compared with a transaction cost approach. In International Studies Of Management And Organisation. 17: 34-48.

Johanson, Jan. and Jan-Erik Vahlne. 1990. The mechanism of internationalisation, International Marketing Review. 7(4): 11-24.

Johanson, Jan and Jan-Erik Vahlne. 1977. The internationalisation process of the firm - a model of knowledge development and increasing market commitments. Journal Of International Business Studies. 8: 23-32.

Johanson, Jan and Finn Wiedersheim-Paul. 1975. The internationalisation of the firm - four swedish case studies. In Journal Of Management Studies. Oct: 305-322. As reprinted in Buckley, Peter and Pervez Ghauri, editors. 1993. The internationalisation of the firm: a reader, Academic Press. London: 16 -31.

Jones, Marian V. 1998. A preliminary analysis of the patterns, processes and decisions that constitute the internationalisation of small UK high technology firms. Strathclyde International Business Unit, Working Paper Series, University of Strathclyde, 98/7.

Kogut, Bruce and Zander, Udo 1993. Knowledge of the firm and the evolutionary theory of the multinational corporation. InJournal of International Business Studies. 24(4): 625-641.

Luostarinen, Reijo and Harri Hellman. 1994. The internationalization processes and strategies of finnish family firms. CIBR Research Papers. Helsinki School of Economics and Business Administration.

Luostarinen, Reijo, Heli Korhonen, Jukka Jokinen & Timo Pelkonen. 1994. Globalisation and SME. Ministry of Trade and Industry. Studies and Reports. Helsinki: 59/1994.

Mattsson, Lars-Gunnar. 1985. An application of a network approach to marketing: defending and changing market positions. In Dholakia, N. and Arndt, J. Research in Marketing Supplement 2, Changing the course of marketing: alternative paradigms for widening marketing theory: 263-288.

Oviatt, Benjamin M. and Patricia McDougall. 1994. Toward a theory of new international ventures. In Journal Of International Business Studies. 23(1): 45-64.

Storey, David J. 1994. Understanding the small business sector. Routledge. London. Chapter 1.

Teece, David J. 1977. Technology transfer by multinational firms: the resource costs of transferring technological know-how. In Economic Journal, 87: 242-261.

Toyne, Brian and Douglas Nigh. 1998. A more expansive view of international business. In Journal of International Business Studies,29,4, pp 863-876.

Wynarczyk, P. R. Watson, D.J. Storey, H.Short & K.Keasey. 1993. The managerial labour market in small and medium sized enterprises. Routledge. London.