The impact of blockchain technology on business models – a taxonomy and archetypal patterns

Blockchain technology enables new ways of organizing economic activities, reduces costs and time associated with intermediaries, and strengthens the trust in an ecosystem of actors. The impact of this seminal technology is reflected by an upcoming research stream and various firms that examine the potential uses of blockchain technology. While there are promising use cases of this new technology, research and practice are still in their infancy about altering existing and creating new business models. We develop a taxonomy of blockchain business models based on 99 blockchain ventures to explore the impact of blockchain technology on business models. As a result, we identify five archetypal patterns, which enhance our understanding of how blockchain technology affects existing and creates new business models. We propose to use these results to discover further patterns fueled by blockchain technology and illustrate how firms can use blockchain technology to innovate their business models.


Introduction
Blockchain is a contemporary technology with the potential to build a foundation for creating unprecedented business models (Iansiti and Lakhani 2017). Of particular interest are business models that remove intermediaries in an ecosystem of actors and those that foster security over performance (Lacity 2018). This change in a ventures' business models adversely impacts several industries (Kshetri 2018;Wang and Kogan 2018). Existing blockchain solutions in the financial industry, for example, eliminate the need for reconciliation and intermediation and enable direct transactions between trading partners (Short 2018). In addition, blockchain technology enables its participants to continuously trace their assets and settle transactions autonomously while providing a secure model that is fault-tolerant, resilient, and permanently available. Upcoming blockchain solutions, such as VeriPart from Moog in the manufacturing industry, facilitate enhanced security and resilience features to increase trust in 3D-printed parts. Inspired by those firms and the expectation for added business value, an increasing number of organizations are actively considering blockchain technology to be a foundational technology (Lacity 2018;Iansiti and Lakhani 2017). In sum, blockchain technology helps authenticate traded goods, facilitate disintermediation, and improve operational efficiency, thereby influencing existing and creating new business models (Nowiński and Kozma 2017).
Whereas the impact of blockchain technology on business models is important (Lacity 2018), current research predominantly focuses on technological aspects (Nakamoto 2008;Wang and Kogan 2018;Eyal and Sirer 2018) and its application in practice (Kshetri 2018;Jun and Vasarhelyi 2017;Radanović and Likić 2018). Examples illustrate the transformation process (Ying et al. 2018;Y. Chen 2018) and indicate how blockchain technology can potentially alter processes and service provision within different industries. However, to the best of our knowledge, no empirical-based research exists regarding how blockchain technology can both change existing and build the foundation for new business models.
In addition to a lack of theoretical focus on the influence of blockchain technology on business models, blockchain companies still fail to deliver the promised business value. Firms lack an understanding of how blockchain technology can create business value for their respective business model (Lacity 2018). Furthermore, it remains unclear what business model patterns have proved to be already successful for this new foundation technology. Building on those shortcomings, this paper focuses on the following research questions: Research Question 1: How can blockchain business models be classified in a taxonomy? Research Question 2: What are archetypal business model patterns for blockchain technology?
To address these questions, we use design science research to develop a taxonomy for blockchain business models and to extract archetypal patterns (Gregor and Hevner 2013). First, we establish a theoretical understanding by conducting a literature review (Webster and Watson 2002) of business models and blockchain technology. Next, we sample a broad database of 99 firms that utilize blockchain technology. We iteratively combine the empirical data with the results of the literature review to develop a business model taxonomy (Nickerson et al. 2013). The blockchain business model taxonomy follows the conceptual representation of business models (Massa et al. 2017). We end by evaluating the ending conditions of taxonomy development (Nickerson et al. 2013). Second, we perform a cluster analysis (Kaufman and Rousseeuw 2009;Struyf et al. 1997) of the blockchain business model taxonomy based on prior theoretical contributions and the database to derive archetypal patterns as reoccurring successful traits (Weking et al. 2018a). We conclude the cluster analysis by evaluating the resulting five patterns using Fisher's exact test (Fisher and Bennett 1990) to confirm that the patterns significantly differ from one another.

Business models
The concept of business models has gained increasing interest among scholars in recent years (Spieth et al. 2014;Frankenberger et al. 2013;Foss and Saebi 2017;Hermes et al. 2019). A consensus is evolving to conceptualize business models as an overall description and architecture of how a firm creates, delivers, and captures value (Osterwalder et al. 2005;Shafer et al. 2005;Teece 2010). In this work, we build on the business model elements proposed by Wirtz et al. (2016): value proposition, value creation and delivery, and value capture. These elements prove to be reasonable for our research as they are used in similar ways by other taxonomies (Täuscher and Laudien 2017;Remané et al. 2017) and business model frameworks (Bocken et al. 2014;Wirtz et al. 2016).
Value proposition refers to the bundle of products and services that a firm offers, whereas value delivery relates to the identification of the target customer. Value creation describes the processes and activities, resources and capabilities, and their orchestration in the firm. Value capture explains how the firm makes money by illustrating revenue and cost structures (Frankenberger et al. 2014;Gassmann et al. 2017). Teece (2010) provided multiple examples of how business models differ based on their respective industry. While traditional firms are mainly concerned with producing physical products and optimizing supply chains, Internet-based firms operate digital services and strive to find the optimal price for information.

Definition and Technical Foundation
Because the development of blockchain is only in the initial stages, definitions are still emerging and no consistent definition has yet been adopted (Swan 2015). However, several authors agreed that decentralized ledger technology is key to the blockchain concept (Gomber et al. 2018;Du et al. 2019;Swan 2015). Swan (2015) described that the main idea of the blockchain is "that the decentralized transaction ledger functionality of the blockchain could be used to register, confirm, and transfer all manner of contracts and property." Similarly, Gomber et al. (2018) stated that "the central technical innovation associated with blockchain is digital ledger technology, which is defined as the use of decentralized digital trust verification through encrypted digital signatures." Abstracting from a general definition, Christidis and Devetsikiotis (2016) proposed four key advantages of the blockchain: tolerance to node failure; single view of events; transparent, verifiable, predictable, and audible activities; and data ownership without a central authority. In general, blockchains record transactions by creating a chain of data blocks. Du et al. (2019) emphasized five IT artifacts of blockchain technology that support these transactions, which are outlined in Table 1. To provide a brief overview of the functionality of the blockchain, we rely on its first use case: transacting bitcoins. Bitcoin is a peer-to-peer electronic cash system in which transactions are executed without the orchestration of an intermediary (Nakamoto 2008). The blockchain serves as a linear register of all past transactions within the bitcoin system. Previous transactions are documented in chained blocks and new transactions are added in new blocks. The blockchain functions as a general ledger in an accounting system. However, information is not stored centrally but is kept redundantly and decentralized across all nodes within the blockchain (Franco 2014). The members of the network administer the decentralized bitcoin system by corresponding protocols and active bitcoin transactions. Every member is eligible to review the locations to which the bitcoins have been transferred while remaining anonymous as an individual entity.

Application of the Blockchain
The blockchain has the potential to transform multiple industries and to significantly alter the fields of its application. Current research is predominately investigating four domains. First, the blockchain is strongly affecting financial services (Underwood 2016), especially the realms of accounting, auditing, and bank transfer (Wang and Kogan 2018). By eliminating a trusted third party, firms can significantly reduce transaction costs (Nowiński and Kozma 2017). Moreover, blockchain enables cross-border transactions in a short amount of time without having to account for currency exchange fees. Those characteristics disrupt traditional business models in the financial industry (Beinke et al. 2018).
Second, blockchain research has focused on the supply chain as the "most promising non-finance application of blockchain," which is believed "to deliver real Return on Investment at an early stage of blockchain development" (Reyna et al. 2018). The formal register of the blockchain enables every member of the system to identify and track the possession of a circulating item throughout the supply chain Grewal et al. 2018). Incorporating such a transparent, verifiable, and shared database reduces current redundancy of every stakeholder operating and updating their database (Christidis and Devetsikiotis 2016). Another advantage of the blockchain is the ability to use connected objects installed in vehicles or storage refrigerators that track temperature to ensure that a product meets health standards along the supply chain (Pilkington 2016).
Third, blockchain can be related to the concept of twosided markets (Glaser 2017). Transaction parties usually need a trusted third party that facilitates the co-creation of value in the market (Hein et al. 2019b). These include platform intermediaries, such as Google, Uber, or Amazon, to provide a safe and reliable environment for transactions (Hein et al. 2016;Hein et al. 2018). However, blockchain eliminates the need for an intermediary by using a crowd of nodes on the system (Ying et al. 2018). Such technological change drives disintermediation ) and decentralization of transactions b e t w e e n m e m b e r s o f t h e s y s t e m ( S w a n 2 0 1 5 ) . Simultaneously, it eliminates centralized risk, low efficiency, and high transaction costs (Ying et al. 2018). Blockchain technology can replace platform providers with a network of nodes (Subramanian 2018). For instance, the blockchain can potentially transform the reviewing and publishing market (Janze 2017) or even make musicians' careers more sustainable (O'Dair and Beaven 2017).
Fourth, we identify a wide range of research addressing the benefits of blockchain in the domain of social welfare (Li et al. 2018;Jiao et al. 2018). Blockchains solutions with their decentralized approach can leverage smart contracts and enable members of the system to contract service outcomes and automate contract conclusion (Cong and He 2019). A new Adapted from Du et al. (2019) The impact of blockchain technology on business models -a taxonomy and archetypal patterns member can signal her genuineness and participate in market transactions without incurring information asymmetry. For voting systems, blockchain technology can digitalize it, decrease voter tampering, and possibly improve voter participation (Kshetri and Voas 2018). Furthermore, blockchain solutions offer numerous opportunities in the healthcare industry (Mettler 2016), such as sharing patient data among clinics and research institutes (Smith and Dhillon 2017). Blockchain technology can address current concerns regarding security by leveraging cryptography, decentralization, and consensus mechanisms (L. Chen et al. 2017). With an universal exchangeable format, healthcare professionals and institutions can easily access sensitive data without putting it at risk (Swan 2015). Table 2 provides an overview of the identified fields of application and their respective benefits and assets.
In conclusion, we observe that current research investigating blockchain is mainly addressing technical aspects or elaborating possible applications and advantages for suitable industries. However, linking the blockchain technology to business models remains scarce.

Blockchain technology and its implications for business models
Blockchain technology offers multiple opportunities to innovate business models. Simultaneously, it imposes certain limitations on the design of new business models. On the advantageous side, blockchains provide various incentives to convert customers to blockchain-based business models (Nowiński and Kozma 2017;Subramanian 2018;Wörner et al. 2016). Depending on its implementation, these can include significant cost reductions emerging from disintermediation (Ying et al. 2018;Xu et al. 2017), faster transaction times (Underwood 2016), reduced record-keeping for customers resulting from the distributed ledger technology, and enhanced data traceability and verification.
The blockchain also offers an alternative approach for authenticating assets, thus setting it apart from centralized transaction systems that rely on an individual organization (Nowiński and Kozma 2017). Blockchains substitute for the trust between entities, which is usually provided by central transaction systems. More specifically, blockchain technology enables small, distributed stakeholders to exercise control over transactions and to hide their identity (Subramanian 2018). A layer of encryption shields all transactions. In combination with decentralization and complex validation mechanisms, blockchains ensure security while promoting trust among members of the system (Zhu and Zhou 2016;Underwood 2016).
Besides cost reduction, traceability, and security improvements, blockchains support the business model and organizational concept of a distributed autonomous organizations (DAO) (Adams et al. 2017;Chohan 2017;Shermin 2017;Elsden et al. 2018;Jun and Vasarhelyi 2017). DAOs exist without central governance, are monitored by shareholders, and are coordinated through smart contracts (Diallo et al. 2018;Chohan 2017). In this way, organizations can replace intermediaries enabling them to services at a lower price.
The implications on the business model and business practices are also related to the underlying assets of the blockchain (Jun and Vasarhelyi 2017). Assets circulating through the blockchain can be physical, virtual, monetary, or userspecific (Smith and Dhillon 2017; Y. Chen 2018; Jun and Vasarhelyi 2017). Implementing the blockchain technology for different assets provides various opportunities for changing and improving underlying business models and firm practices with customers, competitors, and suppliers.
Finally, engaging in a blockchain-based business model enables the use of cryptography and tokenization. Cryptography can substantially change a business model's value proposition as it ensures authenticity behind all interactions in the network (Christidis and Devetsikiotis 2016). Tokenization generally refers to substituting a confidential data component by a non-confidential data component (Liu 2016;Panarello et al. 2018). The business model can enhance its value by rewarding stakeholders with tokens or by accepting third-party tokens (Subramanian 2018). Tokens in the blockchain ledger can also be utilized as certificates to verify the ownership of assets among the firm and its stakeholders (Jun and Vasarhelyi 2017). Drawbacks of blockchain technology stem from diverse requirements, such as platform openness, integration of multiple features, such as identity and privacy (L. Chen et al. 2017), and interoperability as well as performance, scale, and stability (Underwood 2016).
In sum, blockchain solutions offer various opportunities to alter existing business models and create new ones. However, research lacks empirical studies on how blockchain technology impacts business models.

Research method
This paper employs design science research to develop a business model taxonomy and business model patterns for blockchain applications. By following this approach, we ensure practical relevance and scientific rigor (Hevner 2007). We iteratively build on the three cycles of design science research: the rigor cycle, relevance cycle, and design cycle (Hevner 2007). The rigor cycle incorporates the existing knowledge base and ensures that state-of-the-art research will be incorporated in the taxonomy. The relevance cycle connects design activities with real-world problems and strengthens the practical relevance of the taxonomy. The design cycle iteratively develops and evaluates the taxonomy. Figure 1 summarizes the three cycles and their relationships.
Conducting two iterations, we developed the blockchain business model taxonomy and blockchain business model patterns. Both iterations include rigor, relevance, and design cycles with a concluding evaluation. In the first iteration, we developed and evaluated the taxonomy by combining design science research with the taxonomy development method according to Nickerson et al. (2013). The second iteration builds and evaluates the archetypal business model patterns for blockchain technology. Figure 2 summarizes the iterations. In the following, we describe both iterations in more detail.

Iteration 1: Develop taxonomy
We started the first iteration with the rigor cycle and conducted a structured literature review, according to Webster and Watson (2002). We searched in the databases EBSCO, Scopus, and ScienceDirect following the search string "Blockchain*" AND ("Business Model" OR "Business Value" OR "Application" OR "Cryptography" OR "Smart Contracts"). To ensure that we include only high-quality, peer-reviewed journals, we used the VHB-JOURQAUL3 ranking. 1 To further increase topicality and completeness of our review, we included conference papers of the AIS Library matching the keyword "Blockchain." Next, we screened the abstracts and eliminated irrelevant papers and duplicates, resulting in 45 remaining articles. The full-text screening excluded an additional 15 papers. The remaining 30 documents were the basis for the forward and backward search, which yielded 17 additional papers. In sum, the literature review generated 47 relevant articles. Table 6 in the appendix shows the concept matrix of the review. The related work section builds on these results.
In the first iteration of the design cycle for the interim result, we built on the results of the literature review. The design cycle is the heart of any design science research, considering the results of the rigor and relevance cycles iteratively as input for the construction and evaluation of the underlying artifact (Hevner 2007). As the artifact is aimed to be a taxonomy, we used the taxonomy development method according to Nickerson et al. (2013) to guide the design cycle. Hence, we first defined meta-characteristics (step 1 of Nickerson et al. (2013)). We chose the business model elements value proposition, value creation and delivery, and value capture as commonly used in other business model taxonomies (Täuscher and Laudien 2018;Remané et al. 2017) and in business model frameworks (Bocken et al. 2014;Wirtz et al. 2016). Second, we defined ending conditions for the iterative method (step 2 of Nickerson et al. (2013)), where we followed the conditions according to the authors. After setting the foundations of taxonomy development, we conducted the first design cycle.
From the literature review, we followed a conceptual-toempirical approach (Nickerson et al. 2013) to derive the dimensions and characteristics of the blockchain business model taxonomy. For business model dimensions, we refined the metacharacteristics from above with the dimensions service provision, key channel, key resources, key partner, target segment, revenue stream, and cost structure (Wirtz et al. 2016). For blockchain-specific aspects, we included four additional dimensions based on the literature review: Incentives  (Panarello et al. 2018). This led to an initial, conceptual taxonomy. For the relevance cycle, we compiled a database of firms that use blockchain technology as an integral part of their business model. We used CrunchBase as the world's largest database for new ventures (Marra et al. 2015). Because blockchain is a novel technology, the focus on new ventures enables us to analyze a breeding ground of emerging business models. We included all firms in the CrunchBase category "Blockchain" as of January 2018. We discovered 1237 firms as a possible sample. To ensure that our sample only contains successful and relevant firms, we used the following four criteria. First, we included only startups that already received funding to ensure data quality and potential success of firms (Krishna et al. 2016). Second, we excluded firms that went bankrupt or did not have an English homepage. Third, we analyzed the business model of the firms and included only firms that use blockchain technology as an integral part of their business models. For example, we excluded firms such as wallet providers or cryptocurrency exchange markets because they do not directly use the technology. Fourth, we analyzed the required information about the firms and excluded ones that did not provide sufficient information about their business model. After considering all factors, the final set of relevant blockchain firms covered 99 cases.
In the next design cycle, we continued the iterative taxonomy development with an empirical-to-conceptual approach (step 3 according to Nickerson et al. (2013)) building on the results of the previous rigor and relevance cycles. Using case survey guidelines (Larsson 1993) and taxonomy development steps (Nickerson et al. 2013), we classified all firms of our sample (step 4e) with the current taxonomy, identified new characteristics (step 5e), grouped these characteristics into dimensions, and revised the taxonomy (step 6e). In this coding process, we included the firms' webpages, existing technical or white papers, and information from CrunchBase to ensure data triangulation. This procedure led to the following new dimensions intermediation form, user diversification, customizability, value chain position, consensus mechanism, additional technology, and currency acceptance. We distinguished three dimensions, blockchain sourcing, blockchain type, and underlying blockchain as key resources and separated target segment into user and customer.
In the last step of the first iteration, we evaluated the taxonomy according to the ending conditions (step 7) of Nickerson et al. (2013). We were able to distinctly classify all firms of our sample without changing dimensions or characteristics. All objective and subjective ending conditions were fulfilled (Nickerson et al. 2013). To ensure statistical independence among dimensions, we analyzed their correlations. No dimension could be explained with another dimension or combinations of others. Hence, every dimension is important for explaining blockchain business models. Accordingly, we can show that taxonomy provides value, validity, and applicability (Hevner 2007).

Iteration 2: Develop patterns
In the second iteration, we derived the blockchain business model patterns. Again, we started with a rigor cycle incorporating the results of the literature review. The rigor cycle resulted in an initial set of patterns focused on application areas, such as financial services (Underwood 2016), supply chain (Reyna et al. 2018), two-sided markets (Glaser 2017), and social welfare (Li et al. 2018;Jiao et al. 2018).
In the subsequent relevance and design cycles, we performed a cluster analysis (Punj and Stewart 1983) on the sample of 99 firms and the underlying business model taxonomy to refine the initial business model patterns. We transformed the dataset of 99 firms and 22 dimensions of the taxonomy to dichotomous dummy variables for each characteristic of each dimension to measure distances. This process resulted in a vector with 84 binary entries for each firm. We performed hierarchical, agglomerative clustering using the Ward method (Kaufman and Rousseeuw 2009;Struyf et al. 1997). We used several approaches to determine the appropriate number of clusters, as the a priori definition of the number of clusters is a well-known issue in cluster analyses (Anderberg 2014). First, we used the point biserial correlation (Milligan and Cooper 1985) and the C-index (Hubert and Levin 1976), which both indicate five clusters as the best solution. Second, we qualitatively analyzed the five clusters to ensure that clusters are separable (inter-heterogeneity) and that single clusters share common characteristics (intra-homogeneity).  (Fisher and Bennett 1990) to confirm that the five patterns significantly differ in each dimension. We tested if the five clusters significantly vary in each of the 22 dimensions of the taxonomy. All five clusters differ in almost every dimension with a p value lower than 5%. Two dimensions show a p value lower than 10%. Table 7 in the appendix shows the exact p values. Moreover, we already qualitatively evaluated the patterns by confirming the number of clusters. The results indicate that the number of clusters and the clusters themselves are valid and applicable.

Business model taxonomy
The resulting taxonomy contains 22 dimensions, each with two to five distinct characteristics. The taxonomy is visualized as a morphological box wherein a specific combination of characteristics describes the business model of a firm. The taxonomy contains dimensions in which the business models differ. Any further dimension would unnecessarily extend the taxonomy without providing additional information. Table 3  shows the complete taxonomy and Tables 8, 9, 10, and 11 in the appendix define each dimension and related characteristics.

Business Model Patterns
Five archetypal patterns emerged as salient and similar configurations of the business model taxonomy. The five patterns cover every business model of the sample firms, ranging from 10 firms in pattern 3 to 37 firms in pattern 2. Each pattern has different centers along the dimensions and characteristics in the blockchain business model taxonomy, demonstrating intra-homogeneity and inter-heterogeneity. Table 4 shows all five business model patterns and their definitions. Table 5 provides an overview of the specific characteristics of each business model pattern. The table shows the differentiating characteristics of each pattern. These characteristics are based on a relative value of the frequency of a particular characteristic within one pattern to the overall frequency. In this manner, we can indicate characteristics that make a pattern unique and different from others. Consequently, not all firms of one pattern have precisely the same combination of characteristics. One pattern can cover more than one characteristic in some dimensions. In the following, we explain every pattern in detail by focusing on the most differentiating dimensions and characteristics for each pattern.

Pattern 1: Blockchain for Business Integration
The first pattern represents business models that provide the integration of blockchain solutions into existing value networks. Providers sell, for example, blockchain solutions for improving the data interoperability of firms in a supply chain from the Internet of things (IoT) devices. Providers increase data transparency for every participant in the blockchain ecosystem. Contractual time-stamped handshakes of IoT devices enable tracing and verifying data for all parties and, hence, prevent cyber-physical attacks. Customers and users are legal persons. The underlying asset of the blockchain is typically physical. The provider offers a system where IoT devices can generate data and communicate with the blockchain. Providers usually do not offer one standardized product but adjust it to the individual business needs of their customers. Participants store all relevant asset information in the blockchain. Hence, every member can continuously track the current state of the physical assets. Such distributed databases enable smart contracts and can provide additional business value for customers. The challenge is to integrate every member's system, typically ERP systems, in the blockchain solution to leverage its full potential.
Contrary to other patterns, providers offer individual blockchain solutions to customers with a specific business need instead of using the blockchain themselves. Hence, their value chain position is blockchain mediator. They typically use several underlying blockchain technologies and modified consensus mechanisms for a consortium of users. For revenue streams, operators typically do not provide a currency or token in the blockchain but charge their customers with periodic fees.
Modum 2 is a typical example of this pattern. They offer services for supply chain monitoring using IoT sensor devices to generate data (Modum 2018). They store data in a distributed database, granting access to every member on their blockchain. A characteristic of this pattern is that Modum provides its services for different industries, including pharmaceuticals and the supply chain in general.

Pattern 2: Blockchain as Multi-Sided Platform
The second pattern comprises providers that use blockchain solutions as a medium to operate a platform or multi-sided market. Customers and users are typically end users. Platform providers integrate firms to offer complementary products or services. Hence, they rely on industry partners as key partners. Some providers additionally enable users to offer or sell new assets using the blockchain and enable their customers to become complementors. The underlying asset is typically a virtual or a user-specific asset. The former primarily covers enhancing online-gaming experiences (e.g., DMarket 3 ), whereas the latter mainly targets the distribution and selling of data (e.g., BitClave 4 ) or labor (e.g., Bitjob 5 ). Complementors offer their services through websites and do not offer customization. Providers typically use the blockchain framework Ethereum and existing consensus mechanisms because their business does not require specific blockchain modifications.
Contrary to other patterns, the operators typically renounce relying on additional technology. Customers are mainly charged with transactional fees for propagating transactions into the blockchain. Some providers even offer two different tokens in their blockchain, for example for separating currency and assets (e.g., Decentraland 6 ). An additional example is Storm, 7 which provides a marketplace where different consumers can provide and fulfill microtasks. Storm earns rewards in the form of cryptocurrency. Similarly, Power Ledger 8 offers its complementors the ability to sell self-generated electricity, thereby replacing traditional energy suppliers. A more disruptive example is Rega, 9 which offers peer-to-peer insurance ("Crowdsurance"). It enables users to join forces in different insurance platforms, protect each other, and reduce costs by eliminating insurance firms as intermediaries.

Pattern 3: Blockchain for Security
The third pattern builds on cryptography and solutions that provide ownership clarification using cryptography. The solution can be applied to non-physical goods such as data or intellectual property. The asymmetric cryptography allows only the corresponding owner to alter either the data or its ownership. The distributed blockchain informs every participant regarding an ownership change.
Various firms address those features of blockchain technology to provide security of distributed data. The combination of trusted hardware and blockchain technology enable users to authenticate themselves securely. Providers offer these services to legal persons (customers) but target the needs of natural persons (users). The underlying assets are typically users' data. Providers often join forces with technology partners to offer security features. Those services often include additional technologies, such as cloud. Providers sell these without a specific channel. Most of our sample firms do not offer customizable solutions. However, they let firms use their solutions for various application areas. The value chain position is blockchain enabler. Many firms of these patterns use an existing, modified blockchain, typically Bitcoin for a private network with own consensus mechanism. Providers generate revenues with transaction fees and by generating their tokens.
An example is Bluezelle, 10 who provide decentralized storage. They build on the enhancement of privacy, reliability, and immutability of blockchain solutions in addition to reduced costs compared with single system databases. Similarly, NuCypher 11 provides a privacy infrastructure for the decentralized web. Rivetz 12 offers security features for existing assets that slowly become digital.

Pattern 4: Blockchain Technology as Offering
The fourth pattern offers blockchain-APIs to developers. Providers offer a blockchain infrastructure without any further specification of assets. Most customers are legal persons, whereas users also include natural persons. Compared with other patterns, they do not specify the underlying asset or provide a specific channel as they often distribute their API as an open-source. This isolated business model does not intermediate existing value chains and does not rely on strong partnerships. To separate from existing blockchains, providers develop their blockchain solution with modifications, such as modifying the underlying consensus mechanism. These modifications target more specific business needs. Therefore, we attribute those business models as DAO enabler with the value chain position blockchain enabler because they offer various implementations independent of the application area. Although they do not charge their customers, providers profit from the distribution of their tokens; they typically keep a portion of their tokens. A subsequent increase in demand leads to a higher value of those tokens and generates indirect income. Typically, they additionally accept other cryptocurrencies.

Pattern 5: Blockchain for Monetary Value Transfer
The fifth pattern covers firms related to cryptocurrencies. Bitcoin, the first publicly introduced blockchain, developed an electronic cash system, which can replace intermediaries such as financial institutions and saves expenses. Currently, many firms discovered the various advantages of cryptocurrency and implemented their platforms for value transfer. These firms typically follow the Blockchain for Monetary Value Transfer business model pattern. Providers target both legal and natural persons as their customers and users. The intermediation typically occurs within groups (intragroup), enabling the seamless transaction between different peers. Providers focus on money as the underlying asset. They do not heavily rely on partners and offer their services based on mobile applications without customizability as they aim for convenient and cheap transactions. Their value chain position is blockchain user. Providers typically use an external blockchain infrastructure and its underlying consensus mechanism. They do not introduce additional technology. Providers generate revenue by charging fees for every transaction propagated into the network. Furthermore, they distribute their tokens. To increase the value of their tokens, they allow for the transfer of only their tokens within their offerings.
An example is MakerDao, 16 which aims to mitigate the lack of price-stability of cryptocurrencies. They minimize volatility by linking their tokens to an existing asset, such as fiat currency. ETHLend 17 shifts their attention to the loan-aspect of currency. They reduce costs for creditor and debtor by removing crossborder transaction costs.

Discussion
The blockchain is a contemporary technology that has the potential to build the foundation for new business models (Iansiti and Lakhani 2017). However, there is a vast gap between the promised business value and actual value. In addition, it remains unclear how blockchain technology can influence the emergence of new business models.
To investigate the business value of blockchain, we follow a design science research approach (Hevner 2007) and develop a taxonomy of blockchain business models. The taxonomy is based on blockchain and business model literature, and 99 firms that use blockchain as a fundamental technology of their business model. Building on the taxonomy, we conducted a cluster analysis of the 99 firms to derive five archetypal business model patterns that illustrate how blockchain changes existing and triggers new business models.
The first pattern is Blockchain for Business Integration, which illustrates how companies provide a standardized shared database that improves interoperability among actors of a value chain. Second, Blockchain as Multi-Sided Platform, uses direct peer-topeer transactions to overcome restrictions on what can be offered by whom, hence, enabling new business models (Hein et al. 2019a). Those peer-to-peer transactions replace the need for centralized intermediaries (Hein et al. 2019c). Third, Blockchain for Security uses cryptography for security improvements and enables services that previously required physical authentication. Fourth, Blockchain Technology as Offering provide the technical infrastructure to enable blockchain-based business models. Fifth, Blockchain for Monetary Value Transfer uses cryptocurrencies and reduces transaction costs by eliminating the need for a trusted third party. The five patterns reveal that some, blockchain-based firms provide a new value proposition, while others constitute entirely new business models.

Contribution to theory
These findings contribute to the blockchain technology literature and the business model literature, including business model innovation.

Contribution to business model literature
The findings contribute in two ways to the business model literature. First, current literature acknowledges the potential of blockchain to change existing models and trigger entirely new business models in various industries (Iansiti and Lakhani 2017;Lacity 2018;Kshetri 2018;Wang and Kogan 2018) without empirically addressing how this change occurs. This study empirically investigates this phenomenon. The taxonomy enhances the understanding of how blockchain impacts business models. It can serve as a vocabulary that fosters a systematic description of blockchain business models. The taxonomy further illustrates opportunities for business model innovations without oversimplifying their complexity.
Moreover, the five archetypal business model patterns structure the impact of blockchain on business models and further support an understanding of it. The patterns demonstrate possible options for innovating a business model to utilize blockchain technology. We follow the call for analyzing business model patterns in a changing field (Remané et al. 2017). Further, we investigate the understudied topic of how new technologies, i.e., blockchain technology, influence traditional business models (Johnson et al. 2008;Bock and Wiener 2017). Hence, with a business model taxonomy and five archetypal business model patterns, we establish an understanding of how blockchain technology can impact existing and new business models.
Second, our research method illustrates how to derive a technology-specific business model taxonomy and business model patterns utilizing design science, taxonomy development, case survey, and cluster analysis. Design science serves as an overall research strategy with rigor, relevance, and design cycles (Hevner 2007). Guidelines of case surveys, taxonomy development, and cluster analysis further define these cycles. Case surveys provide the empirical base with a generalizable, crosssectional analysis (Larsson 1993). Taxonomy development then adds a systematic approach to integrate empirical and conceptual research (Nickerson et al. 2013). Cluster analysis finally ensures rigor in designing patterns (Kaufman and Rousseeuw 2009;Struyf et al. 1997). We build on all three levels of business models: real-world instances (cases), business model elements (taxonomy), and patterns (Osterwalder et al. 2005). Therefore, we leverage the full potential of the business model concept. Building on these methods, we show how to systematically derive a specific business model taxonomy and patterns that incorporate the existing knowledge base while ensuring practical relevance.
In sum, we first offer a shared language for describing, classifying, visualizing, and analyzing blockchain business models as a basis for future research. Second, our widely applicable research approach shows how to develop a business model taxonomy and how to identify business model patterns for a particular field of application. Hence, we contribute to the business model literature and the growing field of enterprise classifications based on business models (Täuscher and Laudien 2018).

Contribution to Blockchain literature
Literature about blockchain technology predominantly focuses on technological aspects (Nakamoto 2008;Wang and Kogan 2018;Eyal and Sirer 2018) and neglects its business value. Further, current research lacks empirical studies on how blockchain technology changes business models. This study extends blockchain literature by incorporating contemporary research on blockchain technology and its applications with recent developments from practice. By empirically and conceptually developing a blockchain business model taxonomy and extracting five archetypal patterns for blockchain business models, we enhance the understanding of how blockchain technology impacts business models and business value. The taxonomy indicates critical dimensions describing and analyzing businesses operating the blockchain technology. These dimensions include both technical aspects of blockchain solutions as well as elements of business models. The patterns further show tangible instances of how to leverage blockchain technology for business. By investigating blockchain business models, this paper opens up a business perspective on the technology-driven body of literature on blockchain technology.

Implications for practice
For practice, we provide a market overview enhancing the understanding of important aspects of blockchain business models by aggregating from many single entities to archetypal business model patterns. The taxonomy further can serve as a tool for business model innovation (Remané et al. 2016;Weking et al. 2018b). Practitioners can use the taxonomy and patterns to assess opportunities and barriers to integrating blockchain technology in their current business model. The characteristics of the taxonomy and the case examples may inspire practitioners to innovate their business model and allow managers to discover business model innovation opportunities. The five patterns guide the transformation process by specifying the relevant dimensions for a business model innovation. Firms can use patterns and related cases in an ideation phase to identify options for business model innovation toward blockchain and assess their implementation in the firm's context. The taxonomy and its business model patterns can be used as decision support for the evaluation and implementation of business ideas, such as building on an own blockchain solution or external blockchain solution, knowing ways to tokenize assets, or knowing ways to generate revenues.
Moreover, managers can build on the patterns to analyze their current market and identify opportunities and possible market entries for blockchain firms. The results can indicate white spots in the value network as entry points for blockchain provides and support decision making on whether and how to implement blockchain technology. Hence, the business model taxonomy, patterns, and related cases serve as technology-specific support for business model innovation.

Limitations and future research
This paper is subject to three main limitations. First, Nickerson et al. (2013) stated that taxonomies are never perfect and exhaustive. However, the taxonomy and patterns serve as the current state of blockchain business models. As the blockchain application field is evolving rapidly, further research can extend both results with future business models.
Second, we build on the CrunchBase database, the world's largest startup database (Marra et al. 2015). Consequently, we focus on rather new ventures and cannot ensure that all firms that use blockchain technology are part of our sample. For example, we did not find the blockchain platform hyperledger in our cases, since its target group is rather large and incumbent firms. Hyperledger typically comes with licensing costs, whereas startups look for open blockchain solutions. However, the focus on new ventures allows us to better capture the new and still emerging field of blockchain applications. Furthermore, CrunchBase ensures proper data quality, and several other studies have already used it (Block and Sandner 2009;Marra et al. 2015;Werth and Boeert 2013;Yu and Perotti 2015). For data triangulation in the coding process, we consider the firms' webpages, existing technical or white papers, and CrunchBase information. This approach strengthens the validity of our dataset.
Third, we could not evaluate the archetypal business model patterns regarding business performance (cf. Weking et al. 2019). However, while blockchain applications are at an early stage to assess their success, the archetypal patterns indicate a fruitful avenue for blockchain business models and applications.
As blockchain technology is still emerging, and there is a gap between promised and actual business value, this research allows for future research of several aspects. The business potential of blockchain technology is still in its infancy and will evolve further. Future research can build on our taxonomy and archetypal business model patterns as constructs for further empirical studies, qualitative or quantitative. The taxonomy and its patterns serve as an extendable basis for further research providing the main dimensions and corresponding business model patterns. With this business model taxonomy, we want to encourage researchers to study and hypothesize about the relationships among concepts, as proposed by Glass and Vessey (1995). With increased progress of leveraged business value of blockchain technology in practice, research can extend the taxonomy and patterns toward a maturity model for blockchain business models. Moreover, researchers can build on the method integrating case surveys (Larsson 1993), taxonomy development (Nickerson et al. 2013), cluster analysis (Kaufman and Rousseeuw 2009;Struyf et al. 1997), and design science research (Hevner 2007) to develop business model taxonomies and patterns for various fields of application.

Conclusion
Current research emphasizes the technological advantages and the possible application fields of the blockchain technology (Kshetri 2018;Wang and Kogan 2018). Studies have stressed the ability of blockchain solutions to alter and disrupt existing business models and create entirely new business models (Iansiti and Lakhani 2017). However, current research about blockchain solutions focuses on only technological aspects (Nakamoto 2008;Wang and Kogan 2018;Eyal and Sirer 2018) and its application in practice (Kshetri 2018;Jun and Vasarhelyi 2017;Radanović and Likić 2018). Research does not explain a possible influence of blockchain technology on business models and lacks empirical investigations. Consequently, in practice, there is still a gap between possible business value and actual business value delivered.
Therefore, we develop a taxonomy of blockchain business models based on the literature and 99 firms building on blockchain technology. In addition, we identify five archetypal patterns of business models levering blockchain technology, namely Blockchain for Business Integration, Blockchain as Multi-Sided Platform, Blockchain for Security, Blockchain Technology as Offering, and Blockchain for Monetary Value Transfer. We build on design science research (Gregor and Hevner 2013) as a research strategy to guide the research methods case survey (Larsson 1993), taxonomy development (Nickerson et al. 2013), and cluster analysis (Kaufman and Rousseeuw 2009;Struyf et al. 1997).
This paper contributes to business model research by fostering an understanding of how technology, i.e., blockchain technology, influences existing and new business models. The blockchain business model taxonomy is a framework for describing, classifying, visualizing, and analyzing technologyspecific business models, whereas the archetypal patterns show typical instances of it. We also provide a research method to develop field-specific business model taxonomies and patterns by combining the three cycles of design science with case surveys, taxonomy development, and cluster analysis. The results contribute to blockchain literature by introducing the business model concept and combining business and technical aspects for blockchain business models. Findings show how blockchain technology can create business value and enrich the technology-driven blockchain literature. For practice, the taxonomy and patterns identify opportunities for leveraging blockchain technology and help understand important aspects of blockchain business models. Future research can build on our extendable taxonomy and archetypal patterns as constructs for further studies to shed more light on the still rapidly evolving topic of blockchain technology and its business models.

Cluster Analysis
Blockchain Business Model Taxonomy  Physical Asset Underlying asset is physical.

Virtual Asset
Underlying asset is virtual.

User specific Asset
Underlying asset is user-personalized (IP, Labor, Data, …).

Money
Underlying asset represents equivalent of money.

No Asset Specification
Underlying asset is not specified. Stand-Alone Business model exists without partner.

Key Channel
Mobile Application Distribution of offerings through mobile application (and potentially website).

Website
Distribution of offerings through website.

ERP Integration
Integration of business model into existing ERP systems.

Technology Provision without Channel
Provision of business model without channel specification.

External Developer Integration
Non-customer development necessary for integration of blockchain.

Internal Developer Integration
Customer development necessary for integration of blockchain. More than one blockchain solution.

Consensus Mechanism
Self-Created Implementation of own consensus mechanism. Existing Usage of existing consensus mechanism.

Modified
Usage of existing modified consensus mechanism.

Additional Technology
IoT Business Model builds on IoT technology.

Dapps
Business Model builds on distributed apps.

Cloud
Business Model builds on cloud technology.

Big Data Analytics
Business Model builds on big data analytics.

None
No additional integrated technology within Business Model.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.