Market survey and
blockchain adoption model

Image Block 1

“80% of blockchain technology is related to change in business processes and 20% to implementation of the technology”

Blockchain applications require fundamental changes to strategies, business models, operations and users’ mindsets. Such changes bring unprecedented levels of technological, regulatory and collaborative complexity.

Blockchain applications appear to be on the agendas of leading corporations, as well as small and medium businesses, and start-ups. Researchers, industry reports and specialized media continue to publish and circulate views on the blockchain-enabled business transformation. In the wake of the initial hype, however, it is becoming more apparent that this transformation will take much longer than many experts claim.

600+

industry articles

60+

academic papers

125+

executives

With the TAM parameters shown on the y-axis, the following definitions apply for technology acceptance. “Level of knowledge” (LK) refers to how aware users are of the technology in question.

Some degree of knowledge about a technological system can be understood as a prerequisite to acceptance. “Perceived usefulness” (PU) is a factor of motivation to accept and use a technology. If users do not perceive a technology as useful, they are unlikely to accept and use it. “Perceived risk” (PR) is the level of subjective uncertainty among users when deciding whether to accept a technology. PR represents the possibility that the technology will not live up to user expectations. “Perceived ease of use” (PEU) is also a crucial factor influencing sustainable acceptance and use of a technology. Any system must be perceived as easy-to-use, or it is less likely to be accepted by users. Whereas PU deals with a technology’s usefulness or relevance, PEU is a measure of how operable or easy a technology is perceived to be. “Intention to use” (IU) refers to the degree of concrete future plans to use a technology. This indicates a clear intention and a relatively advanced stage of acceptance. “Active system use” (ASU) refers to the technology or aspect of the technology that is currently in use. This is an indicator of the actual level of acceptance. If a system is active in use, it has already been accepted.

We decided to extend the classic TAM model to include the parameters LK, and PR as shown in Table 2. In doing so, we can provide a more complete picture of acceptance, taking early stages of awareness (LK) into consideration along with the subjective uncertainty (PR) surrounding the technology in question. 

“Level of knowledge” (LK) has been added to the classic TAM model as a parameter for assessing acceptance because this is an important signifier of how much users already know about the technology and, in turn, how much facilitating activity is needed to bring awareness of the technology’s value. 

“Perceived risk” (PR) is a risk value based on uncertainty of the outcomes of deciding for or against the technology in question. Business leaders and users have several goals or expectations when accepting a novel technology, and the possibility of these goals and expectations not being met is represented by a degree of perceived risk. Several types of perceived risk can be involved in technology acceptance, for instance: financial risk, performance risk, physical risk, psychological risk, social risk, or convenience risk (Li & Huang, 2009). Due to the general nature of PR, it is difficult to gain data to measure this risk value by means of a survey. However, we have applied a statistical formula to this integrated model, which allows for aggregating values of perceived risk across industries. PR is a significant parameter for risk management, as technology designers and all agents of transformation design can predict levels of readiness or resistance among technology users. This statistical formula and method for analysis is further discussed in Chapter 6. 

The CMM shows seven levels of maturity along the x-axis. As each subsequent level is reached, the acceptance factors will be influenced. The individual levels can be defined as follows. 

Awareness is an indicator of having heard about the technology. The level of awareness increases when a specific technology becomes more widely diffused. Knowledge of a technology goes beyond awareness and indicates that a potential user has invested in understanding and learning about the system in question. As knowledge of a system is gained, risks can be more precisely defined, and values of PU, PEU and PR become more precise and relevant. This precision is needed to reach the next level of maturity. Initiation refers to projects that start the transformation process through introducing the novel technology into existent business operations. 

Implementation is the process of making something active or effective. At this level of maturity, real transformation begins by changing operations, working processes and the habits and behaviors of all parties involved. Once a project has been implemented, these changes and new operations can be tested, and new standards can be defined. 

Standardization is the next maturity level, at which the once novel technology moves beyond the project stages to become a new standard. The highest level of maturity with this model is optimization. At this point, a technology adoption is matured and processes of improvement can be carried out to increase functionality or effectiveness.

We also extended the CMM model to include the levels of “awareness” and “knowledge” as the two initial phases, as seen in Table 3. The intention is to create space for processes that take place prior to initiating a novel technology project within business operations. For example, risks parameters should be identified prior to initiation.

The design and customization of this adoption model is based on principles of transformation design for a decentralized economy and the role that risk plays in this process. We advance that this customized adoption model could have unique value by 1) aiding individual companies with risk management across the entire process of business transformation, and 2) estimating risk levels across industries at the broader market level. To test this model as a possible solution to sustainably meet these objectives, we applied the primary survey and statistical methods discussed in Chapter 6. 

Survey framework

The online survey consists of 21 predefined, closed-ended questions, 8 Likert scale-type closed-ended questions, and 6 open-ended questions: a total of 35 questions.

Interviewees

The 125 interviewees represent a sampling from enterprises, small and medium businesses, startups, 6 levels of executive positions and 20 industries. None of the companies represented are direct blockchain startups, as this would represent a different perspective of acceptance and maturity with the technology. 

Table 4 shows the breakdown of the 125 interviewees according to six levels of management positions, shown according to corporate hierarchy.

Additionally, Table 5 shows the shares of survey participants according to three categories of company size.

Industries

The survey provides a sampling of 125 different companies across 20 industries, excluding the blockchain startup industry. These industries include blockchain use in case-relevant fields such as retail, logistics, consulting, healthcare, e-commerce, finance, advertising, marketing, media, fashion and travel and transportation. Table 6 shows the shares of interviewees according to the 20 industries represented in the survey.

Descriptive analysis

The survey results show a demand for measures that increase solutions for methods that assist in the blockchain adoption process. Furthermore, it appears that there is potential for improvement and new design of adoption models that support companies at each phase of acceptance and level of maturity.

In summary, the topic of blockchain technology in the context of business transformation seems to be relevant. This can be qualified by a high awareness rate, as 91.2% of interviewees have heard about blockchain technology (LK), and 76.8% of interviewees claim to know what blockchain technology is and how it functions (LK).

However, a strong decline in TAM parameters (level of knowledge, perceived usefulness, perceived ease of use, intention to use, active system use) across the CMM maturity levels, regardless of industry sector or company size, indicates that there is a substantial gap in managing the blockchain adoption process in a sustainable way. Only 5% of interviewees are using a comprehensive methodology to apply blockchain for business transformation. Furthermore, only 1 out of 125 interviewees responded positively across all 7 maturity levels (CMM) and TAM parameters.

Empirical analysis

The empirical analysis of certain companies within an industry is conducted via the hypotheses H1-H4 and a Fisher test for proportions. The empirical analysis regarding risk measure at the industry scale, as explored with H5 is conducted via a Binomial test. As described above, we conducted a survey of 125 interviewees from the upper management levels – so-called executives – and created questions that relate to the integrated adoption model (TAM and CMM). The goal was to measure the TAM parameters (see Table 8) at each of the CMM maturity levels (see Table 9) throughout the transformation process of blockchain adoption.

The main assumption is that if executives possess a certain level of knowledge about blockchain technology, it can be concluded that there is a real tendency to accept this new technology, based, in part, upon TAM parameters PU and PEU. 

We expected a significant correlation between the survey answers related to acceptance of blockchain technology, on the one hand, and to applying blockchain across maturity levels on the other hand. In summary, we expected a correlation in responses to questions of TAM and CMM, respectively. 

With the empirical analysis, we tested 5 hypotheses advanced in relation to our research questions. For each of the following sets of research questions and hypotheses, certain survey questions apply. 

RQ1: Which blockchain acceptance and maturity levels do executive business leaders have across different industries and company sizes, in terms of awareness, knowledge and created value? 

H1: Executives are knowledgeable of blockchain technology and its value in terms of perceived usefulness (PU). 

RQ2: Do executive business leaders use a methodology that enables them to apply blockchain technology in an easy way across different capability maturity levels? 

H2: Executive business leaders are knowledgeable of blockchain technology and its value in terms of perceived ease of use (PEU). 

RQ3: Do executive business leaders use a comprehensive methodology, which enables them to initiate and implement blockchain technology in an easy way?

H3: Executive business leaders are eager to apply a comprehensive, easy-to-use methodology to initiate blockchain-enabled services for their business. 

H4: Executive business leaders are eager to apply a comprehensive, easy-to-use methodology to implement blockchain-enabled services for their business. 

RQ4: What is the risk measure across industries that arises by not using an integrated adoption model with respect to TAM and CMM when applying blockchain technology?

H5: The risk measure across industries that arises by not using an integrated adoption model with respect to TAM and CMM when applying blockchain technology is higher than 50%.

To test the hypotheses 1-4, we applied the Fisher’s exact test, due to the small sample sizes. To test hypothesis 5, we applied a Binomial test, as described below. 

With the Fisher’s exact test, one can test whether proportions from two variables are independent of one another. For two-by-two contingency tables (with sample proportions of a, b, c and d in the first, second, third and fourth cells, respectively) one can calculate the exact probability for obtaining a set of proportions by using the following formula:

Equation 1: Fisher test equation

P = \frac{(a+b)!(c+d)!(a+c)!(b+d)!}{a!b!c!d!n!}

Hypothesis 1: Executives are knowledgeable of blockchain technology and its value in terms of perceived usefulness (PU). 

The association between positive answers of question 11 (Q11) and question 13 (Q13) is considered to be statistically significant. This statement is based on P-value = 8.16E-09. Hence, we can assume with sufficient statistical security that the answers from both questions are dependent. Thus, executives who know about blockchain technology may indeed know how blockchain could be used in their company, as well.  At this point, it can be highlighted that we consider CMM level 2: the level of knowledge.

Hypothesis 2: Executive business leaders are knowledgeable of blockchain technology and its value in terms of perceived ease of use (PEU). 

The association between positive answers of question 11 (Q11) and question 14 (Q14) is considered to be statistically non-significant. This statement is based on P-value = 0.199. Hence, we can assume with sufficient statistical security that the answers from both questions are independent. Therefore, the statement of executives’ knowledge of how blockchain functions may not be supported by an appropriate methodology for applying blockchain technology in their company. 

Hypothesis 3: Executive business leaders are eager to apply a comprehensive, easy-to-use methodology to initiate blockchain-enabled services for their business. 

The association between positive answers of question 16 (Q16) and question 17 (Q17) is considered to be statistically significant. This statement is based on P-value = 6.04e-09. Hence, we can assume with sufficient statistical security that the answers from both questions are dependent.  Thus, the companies that have blockchain projects running currently may also use some comprehensive methodology when conducting blockchain projects.  

Hypothesis 4: Executive business leaders are eager to apply a comprehensive, easy-to-use methodology to implement blockchain-enabled services for their business. 

The association between positive answers of question 20 (Q20) and question 21 (Q21) is considered to be statistically significant. This statement is based on P-value = 8.95E-08. Hence, we can assume with sufficient statistical security that the answers from both questions are dependent. In other words, the companies that have implemented some blockchain projects in their companies may also use some comprehensive methodology at this implementation level.  

Due to the complexity, the assessment of risk was not included within the scope of this survey. Therefore, we applied another method for research question 4: What is the risk measure across industries, which arises by not using an integrated adoption model with respect to TAM and CMM when applying blockchain technology?

The related hypothesis can be formulated as follows:

Hypothesis 5: The risk measure, as proposed in Equation 2, is higher than 50%.

For this purpose, the Binomial test has been applied to test if risk measures are higher than 50%. We can use this test to establish whether the proportion of ‘Yes’ and ‘No’ answers within a group is higher than 50%, which yields our hypothesis 5. The statistic of the Binomial test is distributed as follows:

Equation 2: Binomial test equation

P (X = i) = B (i|p_0,n) = \binom{n}{i} p_0^i (1 - p_0)^{n-i}

Here n is the number of answers, and p0 is the assumed proportion of ‘Yes’ answers, which are coded with 1. In our case, we test whether p0 is higher than 50%: 0.5 < p0. The binomial distribution is discrete, and one must compare this empirical value from Equation 2 with the critical values on the particular significance level. 

The Binomial test is an exact test, so that it can be used in the case of small data sets. Actually, there is no alternative to the Binomial test when analyzing small data samples. (Conover, 1971, pp. 97–104). This is the case for our data.

We would like to point out that the risk measure we introduce in Equation 2 is based upon ‘No’ answers. Hence, testing whether this risk measure is statistically significantly higher than 50% is equivalent to testing whether the number of ‘No’ answers is higher than 50%. With the calculation, we only consider industry sectors with two or more responses. 

In doing this, we differentiate between individual industry sectors. The following table shows the results for hypothesis 5 across different industry sectors. All missing responses are interpreted as ‘No’ answers.

Table 10 presents the data produced by applying the Binomial test. The results are organized by the following columns, from left to right: number of the industry sectors (according to alphabetical order), name of the industry sectors represented in the survey, number of answers submitted for each industry sector, and the number of ‘Yes’ answers submitted for each industry sector, the risk measure R and the P-values for both perceived usefulness (PU) and perceived ease of use (PEU), respectively.

The P-value indicates the level of significance at which the null hypothesis can be rejected. If the P-value is lower than 0.10, 0.05 or 0.01, the null hypothesis can accordingly be rejected at a significance level of 10%, 5% or 1%. In this case, it can be assumed that the empirically estimated measure of risk is higher than 50%. 

Image Block 2

With the P-values for PU, the statistical security of the risk confirmation is applicable for the industry sectors Marketing (10%) and Manufacturing (5%). With PEU’s P-values, the statistical security of the risk confirmation is applicable for the industry sectors Advertising, Consulting, Education, Fashion, IT, Law, Manufacturing, Marketing, Media, Pharma, Retail, Software and Travel.