On what resources do management strategists need to focus, if they want to leverage Artificial Intelligence and create Competitive Advantage? To provide an answer from business academia, a framework for leveraging Artificial Intelligence in a business context is presented in this 5-part article series. For Part 1, first some basic concepts from strategic management literature are introduced.

Fundamentals and Overview

In the following, a solid theoretical foundation from strategic management is developed. It is mainly based on the Resource-Based-View (RBV), additionally extended by the Knowledge-Based View (KBV) and Dynamic Capabilities View (DCV), if appropriate. Afterwards, a short overview of the elements of the framework is provided.

Resource-Based View

The RBV was coined by Barney in 1991. Since then, the RBV has been widely-cited in strategic management literature and further developed on a continuous basis. It contrasts the hitherto dominant paradigm of Porter’s Five Forces (Porter, 1979). This paradigm takes a firm-external market-based view. That means it emphasizes the position of the firm within a market containing the five competitive forces:

  1. bargaining power of the buyers,
  2. bargaining power of suppliers,
  3. rivalry among existing firms,
  4. threat of new market entrants and
  5. threat of substitute products of services.

In contrast, the RBV takes a firm-internal perspective instead. It focuses on (firm) resources, which is defines as “all assets, capabilities, organizational processes, firm attributes, knowledge, etc. controlled by a firm” (Barney, 1991). These resources can be classified into three categories, physical capital resources (tangible assets, e.g. IT infrastructure), human capital resources (human labor and their attributes, e.g. managerial or technical skills) and organizational capital resources (intangible, organizational capabilities and knowledge, e.g. company culture). Barney postulates that sources of Competitive Advantage (CA), or even Sustained Competitive Advantage (SCA), are resources fulfilling four distinct criteria. A resource must first and foremost be

  • valuable (i.e. it exploits opportunities or neutralizes threats) and
  • rare (among a firm’s current and potential competition)

order to obtain CA. Additionally, it needs to be

  • imperfectly imitable and
  • non-substitutable

to sustain this CA. This set of four criteria for SCA are commonly known as VRIN criteria.

Knowledge-Based View

In 1996, Spender and Grant introduced the KBV as an extension of the RBV, to emphasize the importance of the shift towards knowledge work and the emergence of the Information Age (Spender & Grant, 1996). Actually, knowledge-based resources might even be the most critical ones for gaining SCA (DeNisi et al.). Consequently, Grant (1996) states the primary role of the firm as “integrating the specialist knowledge resident in individuals into goods and services”, while the primary task of management simply entails coordinating that knowledge integration.

Spender (1996) differentiates knowledge along two dimensions, explicit vs. implicit and individual vs. social, resulting in four types of knowledge. On an individual level, while explicit knowledge is conscious (knowledge about something, e.g. facts), implicit or tacit knowledge is automatic, meaning it is subconscious and hard to transfer (knowledge of how to do something, which is associated with experience). On a social or organizational level, explicit knowledge is objectified (often written down and publicly available, e.g. patents), whereas implicit or tacit knowledge is collective (often unwritten and embedded in routines, norms and culture).

In a knowledge management context, Alavi and Leidner (2001) state four elements of the knowledge process: first the creation (also referred to as construction), second the storage and retrieval, third the transfer and forth the application of knowledge. While classical IT-infrastructure can help with storage and transfer of knowledge, AI and ML can help with the creation and application of it. On the one side, ML, applied in the context of DM, aims at knowledge creation. On the other side, rule-based expert systems, as a form of AI, can apply knowledge by automating decision-making.

Dynamic Capabilities View

In an environment of rapid change, especially of rapid technological change, the RBV was found to be insufficient in explaining how certain firms (e.g. IBM, Texas Instruments, Philips) achieved CA on a regular basis (Teece et al., 1997). The RBV aims at unique resource configurations to leverage long-term CA. Thus, the nature of economic rent creation is Ricardian, i.e. rent comes from scarce firm-specific assets. The RBV is criticized to be only suitable in static environments and to “[overemphasize] the strategic logic of leverage” (Eisenhardt & Martin, 2000).

Therefore, Teece et al. (1997) introduced dynamic capabilities, defined as “the firm’s ability to integrate, build, and reconfigure internal and external competences to address rapidly changing environments […] to achieve new and innovative forms of competitive advantage”. As such, they use the term “dynamic” to emphasize the capacity to renew competences and the term “capabilities” to refer to the key role of strategic management. From an economical perspective, the nature of rent in the DCV is Schumpeterian, which means CA is innovation-based and gained through “creative destruction” of existing competences. Hence, in dynamic markets, “the strategic logic is opportunity” (Eisenhardt & Martin, 2000). Helfat and Winter (2011) particularly differentiate ordinary, operational capabilities – which enable a firm to maintain the status quo – from dynamic capabilities – which enable a firm “to alter how it currently makes its living”. However, they also argue that the transition between those two types of capabilities is smooth, making it impossible to draw a bright line between them.

Overall, dynamic capabilities are best conceptualized as “tools that manipulate resource configurations” (Eisenhardt & Martin, 2000). Even if they are idiosyncratic and firm-specific in detail, for particular dynamic capabilities, there are common ways to execute them effectively, e.g. there are "best practices”. However, dynamic capabilities themselves cannot be the source of SCA, since they do not fulfill the VRIN criterion of having a persistent, heterogeneous distribution across firms. Eisenhardt and Martin argue, that the potential for SCA rather lies in using dynamic capabilities “sooner, more astutely, or more fortuitously than the competition” in order to create resource configurations with SCA. Consequently, dynamic capabilities are necessary but not sufficient conditions for SCA.

Overview of AI Resources

Having introduced a strategically managerial foundation composed of the RBV, KBV and DCV, the business applications of AI can now be structured as a conceptual framework. In total, the framework contains four firm resources:

  1. (Big) Data as a primary raw material,
  2. IT Infrastructure to deal with the data,
  3. Skilled Labor like Data Scientists and
  4. AI Knowledge, e.g. in the form of Open Source Software.

It is inspired by Bharadwaj’s classification of IT-based resources (2000), but it specifies the key elements in applying AI on a team or individual level. Note that the framework itself is independent of the concrete business need addressed by AI (process automation, supporting decision-making, product or service development). However, when the four stated resources are examined in detail during the following subsections, the emphasis lies on the application of ML for BDA to support data-driven decision-making.

In the second part of this 5-part series, the relevance of data, or even Big Data, is elaborated. Stay tuned!

Show Sources
Unser Newsletter

Abonniere und verpasse keine Neuigkeiten

Vielen Dank! Jetzt bleibst du immer auf dem neusten Stand!