MASTER'S THESIS
Protecting Sensitive Knowledge through Knowledge of Sensitivity
Kristian Ericsson
Master of Science in Engineering Technology Industrial and Management Engineering
Luleå University of Technology
Department of Business Administration, Technology and Social Science
2011‐02‐10
Protecting sensitive knowledge through knowledge of sensitivity
Kristian Ericsson
Abstract English
A key issue in inter‐organizational collaboration, such as joint R&D projects, is how to enable
effective knowledge transfer while simultaneously avoid knowledge leakage. In this paper we seek to mitigate this conflict by investigating the nature of knowledge leakage. We address three critical questions where prior literature has offered insufficient answers. (1) What knowledge should a company protect? (2) How does knowledge leakage occur? And (3) which factors affect the likelihood of knowledge leakage? Given the state of prior theory we approached these questions through an inductive multiple‐case study of collaborative R&D projects involving large manufacturing firms and their partners. Five cases of such projects were studied and twenty‐four interviews conducted. Our findings indicate that companies should protect knowledge based on the potential negative
consequences that it might cause rather than what type of knowledge is present. We also propose a general five factor model for leakage processes, and conclude that leakage can be avoided by manipulating one or several of these factors.
Keywords: Alliances, Collaboration, Knowledge leakage, Knowledge transfer, R&D projects
Abstract – Svenska
En nyckelfråga när det gäller samarbeten mellan företag – som exempelvis gemensamma FoU‐
projekt – är att möjliggöra produktiv kunskapsöverföring men samtidigt undvika kunskapsläckage. I den här artikeln försöker vi närma oss en lösning av det här dilemmat genom att undersöka konceptet kunskapsläckage. Vi ställer tre frågor som tidigare litteratur inte besvarat. (1) Vilken kunskap bör att företag skydda? (2) Hur sker kunskapsläckage? Och (3) vilka faktorer påverkar sannolikheten att kunskap skall läcka ut? Givet att så lite skrivits i ämnet så närmade vi oss problemområdet genom induktiva fallstudier av gemensamma FoU‐projekt mellan storskaliga tillverkande företag och deras partners. Fem sådana projekt studerades och tjugofyra intervjuer genomfördes. Våra slutsatser visar att företag bör skydda kunskap baserat på vilka potentiella konsekvenser den kan få om den läcker ut, snarare än vilken typ av kunskap det handlar om. Vi föreslår också en generell femfaktormodell för läckageprocesser och menar att läckage kan förebyggas genom att påverka en eller flera av dessa faktorer.
Contents
1. Introduction... 4
2. Theoretical background... 7
2.1. Knowledge leakage... 7
2.2. Literature on Sensitive Knowledge... 8
2.3. Literature on the leakage process ... 9
2.4. Literature on factors that ease and prohibit leakage... 10
3. Research Methods... 12
3.1. Data sources ... 12
3.2. Data analysis... 15
3.3. Assessing validity and reliability ... 15
4. Results RQ1 ... 16
5. Results RQ2 ... 21
6. Results RQ3 ... 23
7. Discussion ... 25
7.1. Theoretical Implications ... 26
7.2. Managerial implications ... 26
7.3. Limitations and Outlook ... 27
8. References... 27
1. Introduction
During our studies of knowledge leakage we have been surprised to find how people without particular insights in organizational knowledge can grasp the key issue in our research almost immediately. It seems there’s something intuitively appealing with the notion of sensitive
information falling into the wrong hands which makes the concept of knowledge leakage relatively easier to understand and esteem, compared to for example knowledge transfer, absorptive capacity, or causal ambiguity. Perhaps this is a result of fiction, like the contemporary film “Inception” where agencies specialize in extracting intelligence from corporate leaders through their dreams. Perhaps it is a result from media, where the debate about the whistleblower site Wikileaks has dominated the newsflow in recent months. Whatever the cause, leakage seems to be part of the public
consciousness, and we find it strange how such an established phenomenon have avoided a thorough academic assessment until now.
The concept of knowledge leakage is perhaps most conveniently understood using the premises of the knowledge Based View (KBV). According to the KBV, knowledge serves as a basis for firm competitive advantage and therefore plays a major role in all corporations (Cohen and Levinthal 1990; Dyer and Nobeoka 2000; Easterby‐Smith, Lyles, and Tsang 2008, Grant 1996a; Kogut and Zander 1992; Spender 1996; Teece, Pisano, and Shuen 1997). Hence the loss of important knowledge is intimately associated with a decrease in long‐run business performance and, ultimately, a decrease in financial performance (Day 1995; Liebeskind 1996; Norman 2002).
Using assumptions from the knowledge‐based view, several papers have discussed the concept of skills and capabilities leaking to other organizations (Becerra, Lunnan, and Huemer 2008; Easterby‐
Smith, Lyles, and Tsang 2008; Norman 2002; Oxley and Sampson 2004). Albeit these scholars have made important contributions to the understanding of knowledge leakage, the papers propose no clear description of the concept. We see a need for a definition and therefore define knowledge leakage as a process where sensitive knowledge is transferred to an external organization despite the intentions of the originator firm. That is, knowledge leakage is the unintended outflow of a specific, perilous set of knowledge from the focal firm.
If leakage was the only concern of a corporation it could be eliminated by simply prohibiting all kinds of knowledge transfer out of the company. But the KBV offers another concept that strongly advices against such restrictions: knowledge transfer. Knowledge transfer is “an event through which one organization learns from the experience of another” (Easterby‐Smith, Lyles, and Tsang 2008, p.677).
In contrast to knowledge leakage, knowledge transfer is believed to have a substantial positive effect on the participating companies and is even considered a prerequisite for future firm competitiveness (Dyer and Nobeoka 2000; Levinson and Asahi, 1996; March and Simon, 1958 p. 188).
This conflict between creating and maintaining a suitable level of knowledge transfer while at the
same time avoiding knowledge leakage is a well‐known issue in the literature (Easterby‐Smith, Lyles,
and Tsang 2008; Kale, Singh, and Perlmutter 2000; Norman 2002; Oxley & Sampson 2004; Simonin
2004). Quintas, Lefrere and Jones 1997 address this as the “Boundary paradox”, but we prefer the
simpler and perhaps more descriptive The knowledge transfer/leakage dilemma. The aim of this
paper is to further investigate the nature of knowledge leakage and thereby enhance our understanding of the knowledge transfer/leakage dilemma.
Prior literature has presented a multitude of methods to mitigate or overcome leakage issues. For example Baughn, Stevens, Denekamp, and Osborn (1997) give several hints through a structured assessment of the whole lifeline of a collaborative endeavour. Kale, Singh, and Perlmutter (2000) views the creation of relational capital as a mean for lowering the risk of leakage. Dyer and Nobeoka (2000) promotes the initiation of a strong knowledge‐sharing network where all participants present sensitive information to one another; creating a sort of leakage preventing terror balance. Oxley and Sampson (2004) propose limiting the scope of an alliance as another way to avoid sharing sensitive information.
Although these studies may allow firms to achieve enhanced knowledge sharing/protecting performance, they lack a thorough understanding of what really constitutes knowledge leakage.
Specifically, our appraisal of the literature has revealed three issues that ought to be studied before the knowledge transfer/leakage dilemma can be fully apprehended and subsequently dealt with.
First, prior literature has not thoroughly defined what knowledge companies should protect. Rather, the literature often refers to a basic understanding of what is believed to be sensitive – such as “core proprietary information or know‐how” (Kale, Singh, and Perlmutter 2000 p.232) or “highly tacit and core” (Norman,2002) – but rarely discuss what should be included in that definition. We argue that it’s vital to assess what risks are associated with sharing various types of knowledge. Without a proper definition companies might unconsciously share sensitive knowledge, or be unnecessarily overprotective and thereby weaken the positive consequences from knowledge transfer. In addition, it’s difficult to initiate meaningful intellectual property protection activities without knowledge of what must be protected.
Second, previous literature has not yet presented a comprehensive model for how knowledge leakage actually occurs. For example, several authors discuss leakage in network (Dyer and Nobeoka, 2000) or alliance (Norman, 2002; Oxley & Sampson 2004; Baughn et. al., 1997) situations involving competitors, but few mention the leakage occurring in other kinds of settings involving other actors.
We argue that to be able to understand where and how leakage prohibitive activities should be employed there is a need to grasp the full variety of potential leakage processes. Also, an
understanding of the flow of knowledge might render efficient ways of limiting the consequences of leakage other than simply try and hinder knowledge from reaching an external organization.
Third, prior literature lacks a general description of what factors in the leakage process affects the
likelihood of leakage. As was presented above, several papers present various kinds of leakage
preventing methods, but it seems their lack of insight into what knowledge should be protected and
how leakage occurs make them dispersed and uncoordinated; the protective methods presented are
mere samples from a yet unknown framework of methods actually available. We argue that a more
complete understanding of what methods are at hand could greatly enhance the repertoire and help
companies tailor their protective activities to their specific needs. Another benefit is the ability to
combine these methods with the insights of what knowledge is most vital, to make sure protective
resources are used where they have the greatest impact.
To sum up, prior literature has been unsatisfactory in showing (1) what knowledge companies should protect, (2) how leakage actually occurs, and (3) what factors in the leakage process affect the likelihood of leakage. In accordance with this we seek to further investigate the nature of knowledge leakage using the following research questions:
1. What characterizes knowledge within a specific firm in need of protection from external appropriation?
2. Which steps and activities constitute the knowledge leakage processes?
3. Which factors in the leakage process affect the likelihood of leakage?
Given the state of prior theory, these research questions were addressed using an inductive multiple‐
case study. Data was primarily gathered through deep interviews with participants in collaborative R&D projects surrounding the mining and metal industries. The data analysis was based on the generation and refinement of tentative relationships between constructs, as described in seminal work on building theory from case study research (Eisenhardt 1989). Our results contribute to all literature considering collaboration between companies, including knowledge transfer, networks and alliances.
Our findings can be divided in three sections. First, we propose that the notion of Core knowledge has less relevance for knowledge protection activities than previously assumed. Instead we suggest that an understanding of the negative consequences that certain knowledge can impose should serve as means for assessing what knowledge to protect. We also show how appropriation of pieces of knowledge can trigger actions from an external organization with as severe consequences as the loss a complete system of complex knowledge.
Second, we propose a model with five factors and two cases which encompasses all major types of possible knowledge leakage processes. This model extends our understanding of leakage beyond the alliance and network settings previously discussed in the literature. Another interesting finding is that, under some circumstances, the receiver can make use of leaked knowledge without anyone in the organization being able to understand it.
Third, we suggest that all knowledge protection activities can be linked to the manipulation of the five factors constituting the leakage process. This opens up for a new understanding and
development of leakage preventive activities. Our results also indicate that knowledge protection is more about temporary delaying than completely avoiding knowledge leakage.
Apart from the specific results above we have executed a qualitative study in a field previously dominated by quantitative assessment, and thereby enhanced the rich, contextual understanding of what happens during collaborative R&D projects. We have also specifically defined the concept of knowledge leakage to make future studies of it more cogent and to ease the ability for industry professionals to make use of it in daily operations.
2. Theoretical background
The main assumptions for grasping the importance of knowledge leakage and knowledge transfer can all be found in the knowledge Based View. In this paper we adopt a broad definition of knowledge proposed by Liebeskind (1996): “knowledge is information whose validity has been established through tests of proof. Knowledge can therefore be distinguished from opinion, speculation, beliefs, or other types of unproven information” (p.94)
The KBV depicts firms as repositories of knowledge and competencies (Grant 1996b; Kogut and Zander, 1996; Spender, 1996). It assumes knowledge to be a driving force in a firm’s development activities, because knowledge opens up new opportunities but also enhances the firm’s abilities in exploiting these opportunities (Cohen and Levinthal, 1990; Kogut & Zander, 1992). In particular, organizational knowledge allows organizational members to carry out their work tasks based on previously evolved collective understandings (Tsoukas and Vladimirou, 2001), which underscores the path‐dependent nature of organizational knowledge creation and dissemination.
With firms seen as repositories of knowledge and competencies, the exchange of knowledge – knowledge transfer ‐ is easily perceived as an important activity with positive competitive implications. Knowledge transfer comprises both the exchange of knowledge for the sake of increasing the skill levels of the participating organizations as well as the exchanges during, for example, a collaborative development of a new product.
Among the positive consequences proposed are (1) Enhanced knowledge and Innovative capabilities (Dyer and Nobeoka 2000; Easterby‐Smith, Lyles, and Tsang 2008; Gratton, Voigt, and Erickson 2007;
Oxley and Sampson 2004), (2) Enhanced corporate efficiency (Dyer, 1996; Dyer and Nobeoka 2000), (3) Enhanced alliance performance (Crossan and Inkpen 1995; Hutt et al. 2000; Norman 2002), and (4) Increased pace of innovation (Easterby‐Smith, Lyles, and Tsang 2008; Norman 2002; Rycroft and Kash 1999)
A good example of positive consequences comes from recent literature on open innovation, which conceptualize innovation processes as characterized by active outward and inward technology transfer. Another example is innovation processes with multiple actors involving active inward and outward technology transfer, where the transfer of organizational knowledge between partners is critical to end results (Lichtenthaler, 2010).
2.1. Knowledge leakage
But despite these substantial positive effects knowledge transfer has a potential dark side – knowledge leakage. The concept of knowledge leakage has been discussed within areas such as theory of corporate networks (Dyer and Nobeoka 2000), alliances – specifically R&D alliances (Kale, Singh, and Perlmutter 2000; Norman 2002; Oxley & Sampson 2004), and transaction cost economics Pisano 1989; Oxley 1997; Sampson 2004.
Despite the interest from a variety of different theories there has been little effort to specifically define what is meant by knowledge leakage. Various synonyms have been put forth, like “knowledge loss” (Norman 2002 p.178), “Unregulated, unmonitored and unbridled information exchange”
Yoshino and Rangan 1995, “knowledge Diffusion” (Dyer and Nobeoka 2000), “Loss of control of
technological assets” (Oxley & Sampson 2004), and “unilaterally or disproportionately losing one’s
own core capability or skill” (Kale, Singh, and Perlmutter 2000), but what is actually meant by these concepts are contextually rather than specifically defined.
This is troublesome since the common use of the word “leakage” has meanings that are inconsistent with the loss of knowledge. The daily use often indicates something of value escaping a boundary;
like water escaping a pipe or radioactive material escaping the inside of a power plant. This is unsuitable since knowledge is not lost due to leakage, only multiplied. Another difference is that leakage in daily use often indicates that the leaked substance has reached a location where it is never welcome, while knowledge reaching an outside firm is many times considered a positive
phenomenon. This creates another type of confusion, namely what is the significant difference between knowledge transfer considered leakage and not considered leakage. It could be the fact that the exchange was in control/out of control by the sender, known/unknown to the sender,
wanted/unwanted by the sender, or simply positive/negative to the sender. Most interesting about this is that different authors seem to use different definitions.
To overcome this confusion we define knowledge leakage as a process where sensitive knowledge is transferred to an external organization despite the intentions of the originator firm. A few things need to be noted about this definition. First, we assume the viewpoint of the focal firm and thereby regard leakage as a negative concept, in contrast to a societal point of view where “spillorvers” or
“diffusion” often has a pleasant ring. Second, the definition is built on a premise that Sensitive knowledge – the knowledge that needs to be protected from external operators – is a subset, not all, of firm total knowledge. Leakage thus only occurs when there is sensitive knowledge transferred.
Third, the intentions of the originator firm is of importance since companies might choose to share sensitive information with external organizations to achieve some of their objectives, and such deliberate sharing can hardly be addressed as leakage. Fourth and last, according to our definition knowledge leakage is a type of knowledge transfer, not an independent term or an opposite.
The definition does not include relative loss of knowledge resources compared to other firms. A common example from the industry is when a firm has outsourced some of their operations and later suffer consequences because of the knowledge advantage exercised by the external provider. This is not knowledge leakage since the issue is not a result from the transfer of sensitive knowledge.
According to the literature there are substantial negative potential consequences from knowledge leakage (Oxley and Sampson 2004). The most common is the fear of appropriation or imitation of valuable resources by an alliance partner, following the logic of Norman (2002): “a firm’s resources and capabilities are valuable only if they are distinctive and not possessed by competitors” (p.180).
She also states that with the external appropriation of some of its knowledge resources, a firm’s attraction as an alliance partner decreases which further damages its ability to compete.
Grounding in a specific definition of knowledge leakage we continue by discussing how prior literature has dealt with the research questions.
2.2. Literature on Sensitive Knowledge
In our definition of knowledge leakage we expressed the idea that it’s only a certain subset of the
knowledge ‐ Sensitive knowledge ‐ within a company that needs to be protected from external
appropriation. Previous literature shares this idea but use different ways of defining which
knowledge belongs to this delicate set. Examples include Firm‐specific knowledge (Norman 2002
p.178), Competitively important knowledge Kumar and Seth 1998, Technological capabilities (Oxley &
Sampson 2004 p. 745), Critical information, know‐how, or capabilities, and Core proprietary capabilities (Kale, Singh, and Perlmutter 2000 p. 217).
Out of these descriptions, the fourth is perhaps the most specific since it utilizes the word “core”, and several authors consider core knowledge and sensitive knowledge as tightly associated. Norman (2002), for example, investigates the “closeness to the core and the tacitness of the contributed resources and capabilities” of the knowledge companies share while in R&D alliances, since “these two resource dimensions indicate the competitive value of the knowledge to the focal firm” (p. 180).
The last statement is in turn based on previous literature regarding the importance of tacit knowledge Amit and Shoemaker 1993 and core capabilities Leonard‐Barton 1995, respectively.
Norman (2002) also states that firms benefit from sharing core in joint development alliances: “when all partners contribute their core assets to such alliances, the potential payoffs are greatly increased”
(p.181).
These statements about core knowledge are indeed sensible, but to equate Core knowledge and Sensitive knowledge strikes us as going too far in the line of argument. As previously mentioned we regard Sensitive knowledge as knowledge in need of protection. Some core knowledge is indeed in need of protection, but we question whether the external appropriation of any Core knowledge by definition causes harm, and whether no other knowledge has any of the same potential.
Therefore we choose to disregard core in our research setting. Instead we investigate what constitute Sensitive knowledge based on an understanding of what consequences various types of knowledge can impose if obtained by an external organization. The research questions is based on a thought experiment: imagine an external organization temporarily obtaining full access and the ability to immediately absorb any knowledge within a company, what parts of this knowledge could be used by this organization to cause harm to the originator firm? Note that this question
deliberately ignores the presence of a leakage process, which will be discussed in the coming two research questions.
2.3. Literature on the leakage process
Since the literature on leakage usually identifies the concept as merely a factor in some particular field of study there has been no thorough assessment of leakage as such. This has led to a one sided and perhaps simplistic view of how leakage does occur. As an example, previous literature has not directed much interest to the possibility of knowledge leaking through intermediaries like
consultants, equipment suppliers or accountants, although this is a well known issue among industry professionals.
The knowledge transfer process, on the other hand, is well studied and understood, and since we
define knowledge leakage as a sub‐phenomenon to knowledge transfer there might be suggestions
for us to use this literature as a starting point. But there are several important differences between
the transfer studied and the leakage we wish to understand. Most importantly, transfer is often
presented as a conscious action from both parts while leakage can occur without knowledge by the
sender. Our study of the knowledge leakage process is therefore entirely theory building/inductive.
2.4. Literature on factors that ease and prohibit leakage
As indicated in the introduction several papers have presented various kinds of leakage preventing methods and also discussed factors that affect leakage in general. For example Norman (2002) does an assessment of what partners in alliances perceive as indicators of what amount of protective efforts are needed: “The level of knowledge protection in an alliance depends on characteristics of the resources and capabilities contributed by the focal firm to the alliance as well as the desire and ability of the partner to use any knowledge acquired.” (p.180). Studying other literature reveals that these are just samples of factors that managers are recommended to assess while designing their protective measures. Among these are:
Cooperating structure – Many authors discuss the use of various organizational forms of
collaboration that has an impact on knowledge transformed – in particular alliances, networks and various forms of EJV:s Dodgson 1993; Gulati 1995; Mowery, Oxley and Silverman 1996; Oxley 1999.
“Prior research in transaction cost economics suggests that choosing an appropriate governance structure or organizational form is one mechanism that firms use to promote knowledge sharing and protection in an alliance” (Oxley & Sampson 2004 p.723).
Cooperating scope – Oxley and Sampson (2004) define cooperating scope as “decisions such as whether to restrict joint activity to pre‐competitive R&D only or to extend it to include
manufacturing and/or marketing. Related to this is the question of whether a development project can be effectively ‘modularized’ and conducted in relative isolation by the partner firms, only to be brought together at the final stages” (p.724). They also perceive a strong connection between scope and protectiveness that sets the debate of cooperating structure in perspective. “The elevated leakage concerns associated with knowledge sharing in particular competitive contexts, and that restrictions to alliance scope may substitute for protective governance structures in such circumstances” (Oxley and Sampson 2004 p.745)
Focal firm’s knowledge flow management – Another suggested way of dealing with knowledge protection is by managing the flows of knowledge in and out of the firm. For example Baughn et al (1997) discusses “managing knowledge flows and communication”, Levinson & Asahi 1995 adresses
“structure of information flows”, and Kumar & Seth 1998 deals with the issue in total as “limitations on the amount of information sharing and the frequency of communication used for protection”.
Learning intent by external firm – Several papers discuss the importance of the intentions of the external organization (Hamel 1991; Lyles & Salk 1996). “The success rate of both unintentional and intentional learning is related to the motivation of the firm and its individual employees to exploit learning opportunities, the resources allocated to learning, and the organizational mechanisms in place to absorb knowledge. Thus, firms without an explicit learning intent will be much less likely to learn” (Norman p. 182). This also relates to the discussion of “Learning races” brought up by Khanna et al. (1998).
Tacitness and explicitness of the knowledge – The type of knowledge present is regarded as crucial
to whether it can leak and with what ease. “The tacitness of knowledge is the most important
determinant of the ease with which the knowledge can be transferred to another party” (Norman
2002 p.181). Kale, Singh and Perlmutter (2000), presents some requirements for the transfer of tacit
knowledge; “Learning, especially the acquisition of, difficult‐to‐codify competencies, is best achieved
through wide‐ranging, continuous and intense contact between individual members of the alliance partners” (p.232)
Regulations and legal contracts – Both informal, socially restricting, like “clear rules for participation in the network's knowledge‐sharing activities.” (Dyer & Nobeoka 2000 p. 364) and formal, like
“contractual safeguards and investments” Parkhe 1993, are discussed.
Resource overlap between firms – In the knowledge transfer literature it is a well‐established fact that corporations with similar knowledge bases find sharing easier than firms with divergent capabilities (Mowery, Oxley and Silverman 1996; Oxley and Sampson 2004 p.724). “A firm has the greatest potential to internalize and use knowledge gained from a partner when that partner has similar skills, resources, and capabilities” (Norman 2002 p. 183). This is also known by firms, and they are therefore more protective if the opposing firm lies closer to their own field of specialty; “Greater resource overlap leads firms to more restrictive communication and information sharing practices”
(Norman 2002 p. 192).
Trust/relational capital between firms – Much has been said about the impact of trust on successful knowledge transfer. “Strong ties produce the trust (social capital) necessary to facilitate the transfer of tacit knowledge” (Dyer & N 2000 p.365), and “with higher levels of trust, firms are more willing to share information and open up communication channels, actions that have been linked to higher learning” (Norman 2002 p.193). Trust seems to lower the perceived risk of opportunism by both sides; “trust is a substitute for formal control mechanisms” (Norman 2002 p. 192).
Although these statements about leakiness all seem probable and the amount of them giving the impression of a thoroughly investigated topic, there are a couple of issues questioning their
relevance. Firstly, none of them is based on a thorough understanding of (1) what knowledge should be protected and (2) how the leakage process occurs (we know this since, as stated in the paragraphs above, both of these topics have been given little attention in prior literature). This despite the fact that the knowledge transfer literature emphasizes assessing both knowledge characteristics and process factors as vital for being able to determine the success of any transfer (Easterby‐Smith et. al., 2008), and we see no reason as to why this should not be the case in leakage.
Secondly, and perhaps originating from the confusion regarding the process above, the various ways of prohibiting leakage are directed at several unrelated factors. An initial distinction is that some of them aim at prohibiting leakage as such (cooperating scope, knowledge flow management, learning intents, tacit‐ and explicitness) while others aim rather at affecting the consequences from leakage (cooperating structure, regulations and legal contracts, resource overlap, relational capital). Going closer, they also seem directed at manipulating different target variables; like knowledge type, cooperative situation, choice of partner, degree of opportunism by the counterpart etc. This of course raises an interest in knowing the total amount of factors present – not just samples of them ‐ and how each of these can be manipulated.
We argue that an increased understanding of the leakage process might reveal these factors and
thereby gather the dispersed statements regarding knowledge leakage prevention into a unified
framework. Another benefit is the ability to combine these methods with the insights of what
knowledge is most vital, to make sure protective resources are used where they have the greatest
impact. This could in turn enhance the repertoire at hand and help companies tailor their protective activities to their specific needs.
3. Research Methods
In this study we have used an inductive multiple case‐study framework as described by Eisenhardt (1989). The reasons for this are, firstly, the relative lack of theory concerning especially RQ 1 and RQ2. Secondly, that previous studies have been quantitative and perhaps overlooked the contextual, deep understanding for what has happened during projects. And thirdly, one of the research
questions is trying to map a process, which is preferable to do in a qualitative setting. Thus, given the state of prior theory, the research strategy was selected to assure methodological fit (Edmondson &
McManus, 2007).
Our cases were chosen from the mining and metal industries. The slow‐moving pace of this industry makes leakage an issue since the loss of sensitive information can have competitive consequences for several years. This contrasted to fast‐paced industries where the impact of negative
consequences stemming from knowledge leakage might be limited to a year or even a few months.
The organizations involved in this study are mining companies, metal refinement corporations and their equipment suppliers, consultants and a collaborating university.
This study focuses on collaborative R&D projects. These are a particular case of knowledge transfer since the exchange of knowledge is subordinate to the development of a product – “...we should not just be focusing on knowledge transfer, but also on the transformation and integration of knowledge into commercial innovation.” (Easterby‐Smith, Lyles and Tsang 2008 p.684). Personnel involved in such collaborative projects involving the metal and mining industries must face issues of knowledge leakage on a daily basis.
3.1. Data sources
Our research process consisted of four data gathering cycles with subsequent data analyses – each cycle slightly shifting the process from theory building into investigating preliminary conclusions.
Table I: List of interviews and respondents
# C ycl e D at e C ase T i t le A cad emi c t r aini ng F ir m Y . emp l . D ur . ( mi n)
1 1 2010-11-11 3D M easuring Device Research engineer M Sc Comput er engineering First mining corporat ion 20 70
2 1 2010-11-15 3D M easuring Device Development engineer M Sc Automat ic cont rol engineering St eel corporat ion 24 35
3 1 2010-11-22 3D M easuring device Project manager M Sc Automat ic cont rol engineering Universit y 7 75
4 3 2010-12-01 3D M easuring Device Specialist measurement syst ems PhD Nuclear physics Second mining corporat ion 26 70
5 3 2010-12-02 3D M easuring Device Researcher/lecturer PhD Elect rical engineering Universit y 5 90
6 2 2010-11-29 Drilling rig Development engineer Upper secondary engineer Second mining corporat ion 25 60
7 1 2010-11-15 Feeding device Project manager M Sc Industrial design engineering First mining corporat ion 15 70
8 2 2010-11-30 Feeding device Sit e manager BSc M echanical engineering First consultancy corporat ion 7 70
9 3 2010-12-06 Feeding device Aut omat ion engineer BSc Elect rical engineering First mining corporat ion 22 35
10 2 2010-11-30 Filtering cloth Research engineer M Sc M ining engineering M et allurgy corporat ion 20 70
11 3 2010-12-01 M aint enance system M ng Process dev., below ground Upper secondary engineer Second mining corporat ion 17 70
12 3 2010-12-01 Blender prototype M ng Process dev. PhD Process met allurgy M et allurgy corporat ion 10 70
13 1 2010-11-12 St eering device Research engineer PhD Fluid mechanics First mining corporat ion 3 60
14 2 2010-11-29 St eering device Senior research engineer M etallurgist First mining corporat ion 31 80
15 3 2010-12-01 St eering device Consult ant Prof essor fluid mechanics Second consult ancy corporatio 30 70
16 3 2010-12-02 St eering device Project M anager PhD Chemical engineering Third consult ancy corporation 4 70
17 1 2010-11-01 None (Explorat ory) Senior M ng Process technology Upper secondary engineer First mining corporat ion 20 120
18 1 2010-11-08 None (Explorat ory) Specialist mineralogy M ining engineer First mining corporat ion 29 70
19 1 2010-11-08 None (Explorat ory) M ng Process dev., above ground M Sc Chemist ry First mining corporat ion 6 45
20 1 2010-11-11 None (Explorat ory) Global supply coordinat or BSc Development engineering M et allurgy corporat ion 14 50
21 1 2010-11-12 None (Explorat ory) M ng Process dev., below ground M ining engineer First mining corporat ion 21 50
22 1 2010-11-16 None (Explorat ory) Research engineer PhD M etallurgy M et allurgy corporat ion 8 45
23 1 2010-11-16 None (Explorat ory) M ng Component and materials dev. M Sc M at erials design and engineering M et allurgy corporat ion 12 75
24 1 2010-11-19 None (Explorat ory) M anager market intelligence M ining engineer First mining corporat ion 25 50
Data was gathered primarily using interviews and documentation. Twenty‐four interviews spanning from 35 minutes to 120 minutes with an average of 70 minutes, were conducted. Respondents were R&D project participants, personnel knowledgeable about protection routines, lawyers, senior managers, and project managers (see table I). Documents investigated were knowledge protection policy documents handed out to project participants in the focal firms.
The four data gathering cycles were conducted as follows. In the first cycle our aim was to get a general overview of how industry professionals perceived knowledge leakage issues. Specifically, we wished to know in what contexts knowledge leakage issues were most prominent to be able to see which R&D projects would be most suitable as research cases. To do this we interviewed several project managers and others possibly familiar with leakage issues in the participating companies.
Aided by this initial assessment we came to conclude that the projects chosen as cases should reach the following criteria:
‐ They should be collaborative endeavours involving a firm and an external organization
‐ They should be research or development projects where either the projects themselves involved intended sharing – or where there was a substantial risk of unintended sharing – of knowledge perceived as sensitive by any of the involved parties.
‐ They should have employed at least two people from each organization (to allow for triangulation of the whereabouts in a particular case)
‐ They should have involved less than 20 people (so all respondents would have been able to maintain an overview of the project)
‐ They should have been completed recently or be close to completion (for the respondents to remember the details and have the full picture of what happened)
This led to the choice of the five projects presented in table II as cases for this study.
The project managers for the projects perceived by us to be most promising (this assessment was made early in the interviews) were asked further questions about their specific project, while those managing projects with less knowledge leakage issues or just generally familiar with such issues (the Table II: List of case projects
# C ase Pr o j ect o b j ect i ve C o mp ani es i nvo l ved D ur at i o n
1 3D M easuring Device An online size measurement syst em of ore product s running on a conveyor. Primarily a collaborat ion wit h First mining corporat ion - t o replace a f ormer of f line sampling met hod - and Universit y - supplying t he t echnology. The ot hers part icipat ing in preparat ion of own, f ut ure implement at ion of t he syst em.
First mining corporat ion, Second mining corporat ion, St eel corporat ion, Universit y
Spring 2003 t o early 2007
2 Blender prot ot ype Full-scale prot ot ype of a blender designed t o mix iron ore part icles wit h a chemical subst ance. Init iat ed by t he met allurgy corporat ion t o enable t hem t o of f er new product s t o t heir cust omers.
M et allurgy corporat ion, Equipment supplier
16 mont hs f rom 2005 t o 2006
3 Feeding device A device f eeding an organic subst ance t o a mix of iron ore and ceramics. First mining corporat ion wished t o experiment wit h various organic subst ances and engaged First consult ancy corporat ion t o assist in f inding and implement ing a proper device f or t hese experiment s.
First mining corporat ion, First consult ancy corporat ion
Early 2010 t o early 2011
4 M aint enance syst em Int egrat ion of t he mine operat ors' working int erf aces wit h t he maint enance depart ment 's f ault report ing and working order syst ems. Second mining corporat ion init ially bought a f inished syst em f or t heir needs, but made requirement s t hat event ually t riggered great development ef f ort s f rom t he t wo ot hers.
Second mining corporat ion, Indust rial aut omat ion corporat ion, Comput er syst ems corporat ion
Early 2008 t o middle 2010
5 St eering device Comput er model of a new rot or t o increase t he mix qualit y and energy-ef f iciency in a silo used t o buf f er a mixt ure of iron ore and wat er. The project was init iat ed by First mining corporat ion t o avoid downst ream product ion problems in t he processes.
First mining corporat ion, Second consult ancy corporat ion, Third consult ancy corporat ion
Early 2010, st ill running