Measuring the Efficiency of Introducing Businesses’ Digitalization Elements over Time in Relation to Their Performance ^†

Abstract

The introduction of digitalization elements into the life of companies is significant in terms of achieving better economic results. The aim of the research was to determine the technical efficiency, as well as the change in efficiency and the technological change in the digital transformation of companies in EU countries in relation to their performance. The Malmquist index was used to measure these parameters over time. The results of the research indicate the significance of the dynamic measurement of the efficiency of digital transformation. Interesting results also point to the importance of evaluating the efficiency of the use of already established elements, as well as evaluating the introduction of new technological changes.

1. Introduction

In recent times, digitalization and digital transformation have increasingly come to the fore and are now becoming part of strategic European documents. The key document adopted in 2022 was the 2030 Digital Compass: the European Way for the Digital Decade, which sets up concrete goals for enhancing Europe’s digital transformation until 2030 [1]. For EU member states, but also on a broader level, various digitalization indicators and indexes are monitored. Probably the most well-known one is DESI (Digital Economy and Society Index) which monitored digital progress of EU member states between 2014 and 2022. From 2023, DESI was integrated into the Report on the State of the Digital Decade [2]. Another interesting index is the EIBIS Corporate Digitalization Index, which summarizes digitalization indicators and companies’ evaluation of digital infrastructure and investment, not only in the European Union, but also in the United States. It was based on the European Investment Bank Investment Survey (EIBIS), results of which show that the European Union has been narrowing the gap in digital adoption compared with the United States in the last four years. In the period from 2021 to 2022, the share of EU businesses adopting advanced digital technologies increased, reaching 69% in 2022, compared with 71% in the United States. Medium and large companies achieved better results than micro and small ones. According to the EIBIS survey, in 2022 only 30% of micro-enterprises in the European Union had taken measures to improve digitalization, compared with 62% of large firms. Digital transformation in the EU requires the introduction of more advanced digital technologies like artificial intelligence, big data, 3-D printing, and advance robotics. There have been significant differences among the EU countries in the introduction of these technologies [3].

When measuring the efficiency of digital transformation in companies doing business in EU countries, most studies [4,5,6] have performed DEA (Data Envelopment Analysis). However, the Malmquist index, which is also based on the DEA methodology offers some advantages compared with DEA. The most important advantage is that it is able to offer results related to change in efficiency and technological change. Moreover, when applying the Malmquist index, we are able to compare efficiency results in two periods. We can say that the DEA-based Malmquist index (MI DEA) is able to expand the results available by applying DEA methodology. When searching in literature, just one application of the MI DEA in a related field was found [7]. Therefore, this study aimed to fill this research gap by applying the MI DEA to measure and compare the efficiency of the digitalization process in companies doing business within EU member states. When applying the MI DEA, we followed our previous research into the efficiency of digital transformation [8,9]. However, in this study, different outputs were used to express business performance.

2. Literature Review

The concept of organizational performance has gained much attention during recent decades [10]. Literature offers various definitions of this term. Organizational performance is defined as the ability to achieve an organization’s goals [11,12], the ability to transform production factors into final products and services [13], the ability to satisfy expectations of stakeholders [14,15], the determinant of the value of the organization [16], the ability to create value for customers [17], etc. According to several authors [18,19], performance is closely related to the efficiency and effectiveness of the organization. Sedláček et al. [20] argue that this term is linked to the realized output of the organization, while performance can be measured and then further analyzed. This approach is further discussed by Richard et al. [21] who argued that organizational performance includes three areas of business output: financial performance (measured by profit, return on assets, etc.), product market performance (measured by market share, sales, etc.), and shareholder return (measured by total shareholder return, economic value added, etc.).

Howell [22] describes performance measurement as a metric that serves to obtain more detailed information about the business process. One of the innovative ways of measuring performance of organizations is business-process management, which has come to the fore in recent decades. It is based on the idea that a business can make better use of its current resources by increasing the efficiency of its business processes [23]. Digitalization offers significant opportunities for improvement of processes since it allows a company to minimize resource consumption through introducing information and communication technology (ICT) tools [24]. Several authors have confirmed the impact of digital technologies on the performance of businesses. For example, Tohanean et al. [25] argued that digitally based technologies have a major impact on today’s business performance. These authors used the example of a German company to demonstrate that digitalization leads to the provision of safer products for customers and the achievement of better results. Among the most important digital drivers are social media, big data, cloud computing, and the Internet of Things. Similarly, Ribeiro Navarrete et al. [26] confirmed that the use and updating of social networks, a high level of training in digital skills, and the employment of older managers can increase company performance. Sommarberg and Mäkinen [27] pointed out the importance of introducing digitalization elements for the creation of the company’s value. Kádárová et al. [28] referred to the importance of digitalization for improving the efficiency of business processes, enhancing businesses’ productivity, and improving customer experience.

Several studies have investigated the efficiency of digitalization or digital transformation taking place in businesses as well as countries. Krstic et al. [6] measured efficiency of the use of ICT in businesses from EU countries. They measured efficiency with the use of a DEA window model in the period from 2012 to 2020 and performed robustness analyses of the efficiency results using the bootstrap method. According to their results, the highest average technical efficiency was achieved by Belgium and Denmark. On the other hand, the lowest technical efficiency per window was recorded in Romania, Bulgaria, Latvia, and Greece. Rejman Petrovic et al. [5] also applied the DEA model combined with bootstrap methodology to measure the efficiency of the use of ICT technologies in businesses in the Republic of Serbia. Lungu et al. [4] applied DEA to measure efficiency of digitalization in businesses, while they examined the national context and entrepreneurial activity within a sample of 47 countries. The results of their study identified Japan as the most important peer benchmark in all three scenarios. An integrated approach to the evaluation of business digital transformation was proposed by Kuntsman and Arenkov [29]. Their approach was based on the combination of several quantitative and qualitative methods including investment analysis methods and direct and indirect costs of ownership as well as a Balanced Scorecard. The proposed approach was implemented by a real company and, according to the authors, it met all their expectations.

Digitalization efficiency of EU countries was measured by Georgescu et al. [30], who used data from the DESI and the Stringenci Index and created the DEA CRS (DEA Constant Returns to Scale) model. Their study resulted in the identification of eight efficient countries among EU member states, that can be considered as benchmarks for inefficient ones. Yalcin [31] also measured the efficiency of digitalization in EU countries. Inputs of his DEA VRS (DEA Variable Returns to Scale) model consisted of five dimensions of the DESI, while GDP (Gross Domestic Product) growth rate and unemployment rate were used as outputs. Yalcin’s study revealed that developing countries introduce digitalization elements more effectively in terms of GDP growth and job creation. Similar results of efficiency of digital transformation were achieved by Inel [32]. This author used DEA and data from the Digital Transformation Scoreboard and revealed that some developed countries like Germany or the Netherlands are not very efficient.

3. Data and Methodology

The aim of the research was to measure the performance of digital transformation of EU countries by comparing the years 2019, 2021, and 2023. The research sample included 27 EU countries. As the inputs of the research, selected digitalization indicators of businesses were used: E-commerce sales (percentage of enterprises), use of computers and the Internet by employees (percentage of total employment), type of connections to the Internet (percentage of enterprises), websites and functionalities (percentage of enterprises), social media use by type, internet advertising, and cloud computing services [33]. As outputs to the applied MI DEA model, the following indicators were used: Cash/Firm Value (CFV), Enterprise Value/EBIT, and Enterprise Value/Sales [34]. The choice of years was conditioned by the availability of data, but also by an effort to compare the relationship between indicators of digital transformation and business performance in the periods before, during, and after COVID-19.

Firstly, medians of selected inputs and outputs for 27 EU countries were analyzed. Results for the years 2019, 2021, and 2023 are presented in

Table 1. Descriptive statistics of inputs and outputs.

Variable	Valid N	Median 2019	Median 2021	Median 2023
E-commerce sales	27	20.8	22.60	23.5
Use of computers and the Internet by employees	27	47.20	53.00	59.9
Types of connections to the Internet	27	94.00	94.80	94.8
Websites and functionalities	27	78.70	77.90	77.1
Social media use by type, Internet advertising	27	51.50	58.10	60.8
Cloud computing services	27	33.30	40.40	46.5
Cash/Firm Value	27	10.22	5.12	5.27
Enterprise Value/EBIT	27	16.23	16.41	14.33
Enterprise Value/Sales	27	2.22	2.22	1.77

Source: [33,34].

.

Table 1 shows that the best Cash/Firm Value was achieved in 2019 (before COVID-19) and the worst during COVID-19 (2021). Most of the inputs reached their highest values in 2023, which indicates the growing pace of the digital transformation process within the EU regions.

The method applied to compare the development of businesses’ performance in relation to digital characteristics was the MI DEA index proposed by Färe et al. [35]. It measures the change in Total Factor Productivity (TFP) between two periods as the ratio of distances of each data point with respect to a common technology [7].

Let each ${DMU}_j\left(j=\mathrm{1,2},\dots ,n\right)$ be presented by the vector of inputs $x_j^t=\left(x_{1j}^t,\dots ,x_{mj}^t\right),$ which is used to produce the vector of outputs $y_j^t=\left(y_{1j}^t,\dots ,y_{mj}^t\right)$ in each time period $t,t=1,\dots ,T$. The Malmquist productivity index can be written as follows (1) [36]:

$$M{I}_o=\underset{\mathrm{ECH}}{\underbrace{\frac{{\theta }_o^t(x_o^t,y_o^t)}{{\theta }_o^{t+1}(x_o^{t+1},y_o^{t+1})}}}\times \underset{\mathrm{FS}}{\underbrace{{\left[\frac{{\theta }_o^{t+1}(x_o^{t+1},y_o^{t+1})}{{\theta }_o^t(x_o^{t+1},y_o^{t+1})}\frac{{\theta }_o^{t+1}(x_o^t,y_o^t)}{{\theta }_o^t(x_o^t,y_o^t)}\right]}^{\frac{1}{2}}}},$$

(1)

where ${ MI}_o$ expresses the change in productivity between the periods $t$ and $t+1.$ When calculating ${\theta }_o^t\left(x_o^t, y_o^t\right),$ $x_o^t$ is compared to the EPF (Empirical Production Frontier) at time $t$ applying the input-oriented DEA CRS model, where $x_o^t$ = ($x_{1o}^t,\dots x_{mo}^t$) and $y_o^t$ = ($y_{1o}^t,\dots y_{so}^t$) are the input and output vectors of ${DMU}_0$. ${\theta }_o^{t+1}\left(x_o^{t+1}, y_o^{t+1}\right)$ and are calculated similarly. The change in productivity is expressed by multiplying the efficiency change (ECH) and Frontier shift (FS).

4. Results and Discussion

The development of selected indicators of digital transformation can be seen on Figure 1 and Figure 2. We can see that the introduction of digitalization elements is growing over time. For example, while in 2021 sales from E-commerce were achieved by 24% of EU companies, in 2023 they were already achieved by 25% of companies (see Figure 1). In the case of cloud computing the results are even better. While in 2021 cloud computing was used by 43% of EU companies, in 2023 it was already used by 47% of companies (see Figure 1).

In the case of the indicator social media use by type, internet advertising, there was an increase from an average value of 55% in 2019 to 63% in 2023 (see Figure 2). A small year-on-year difference was recorded in the development of the indicator websites and functionalities, in the case of which there was, rather, a decrease (see Figure 2).

Table 2. Descriptive statistics of the MI DEA results.

Variable	Valid N	Mean	Median	Minimum	Maximum	Std. Dev.
Malmquist index 19_23	27	3.23	2.54	0.78	11.83	2.54
Malmquist index 21_23	27	1.27	1.18	0.96	2.04	0.22
Frontier shift 19_23	27	0.79	0.83	0.40	1.37	0.26
Frontier shift 21_23	27	1.00	1.00	0.81	1.61	0.16
Efficiency change 19_23	27	0.79	0.83	0.43	1.37	0.26
Efficiency change 21_23	27	1.00	1.00	0.81	1.61	0.16

Source: authors.

shows the descriptive statistics for the results of the MI DEA, efficiency change (ECH), and Frontier shift (FS) calculated for 27 EU countries. The MI DEA clearly showed better results when comparing the years 2019/2023 in most EU countries. Only Germany and Slovakia achieved better results for the years 2021/2023 compared with the years 2019/2023. However, it should be stated that when comparing the years 2019/2023, Slovakia had a MI DEA value below 1. FS showed better results when comparing the years 2021/2023, except for eight countries. In the case of these countries, it was assumed that they did not introduce technological changes to a great extent, since they represented important leaders in the given field and had introduced many elements of digitalization to the level of 90–100%. The ECH results were individual for each country; some countries achieved a higher ECH when comparing the years 2019/2023 and others achieve a higher value when comparing the years 2021/2023.

The results of the Malmquist index are shown in Figure 3, which shows the MI DEA for the years 2019/2023 as well as for the years 2021/2023. We can see that the performance of digital transformation was clearly higher when comparing the years 2019/2023. This can be justified by the fact that, in 2019, the EU was not yet pushed by the necessity of introducing digitalization elements due to the impending pandemic and other significant external influences. On the other hand, in 2021, significant digital transformation was already taking place due to the COVID-19 pandemic. This acceleration was significant in the case of most EU countries. The leaders in this area were the Netherlands and Luxembourg.

MI DEA results higher than 1 indicate that the digital transformation process is effective, and countries are performing better. This can be seen especially when comparing the years 2019 and 2023 (in all countries $MI DEA>1$). In the second phase of their research, we investigated the effect of digitalization on previously calculated TFP using the Tobit model and confirmed the significant and positive impact of digitalization on productivity.

Within the MI DEA index, two variables were monitored, namely the change in technical efficiency (Figure 4) and the Frontier shift (Figure 5). The change in technical efficiency was more significant when comparing the years 2021/2023—for 13 countries, the value of this variable was above 1. This indicates that during this period (also due to the need caused by the COVID-19 pandemic) already implemented elements of digital transformation began to be used more intensively in most EU countries. The leader in this given area and for the given years was Slovenia. When comparing the years 2019 and 2023, the Netherlands was the leader, while, for nine countries, the value of this variable was above 1. This also confirms that the technical efficiency of the use of individual elements of digital transformation was higher when comparing the years 2021/2023.

In the case of Frontier shift (see Figure 5), it can be stated that after 2021, various technological changes began to be significantly applied within the EU countries. This technological change was more pronounced when comparing the years 2021/2023. The development of technological change was favorable for all EU countries, especially for the years 2021/2023. When comparing these years, all countries achieved a level of technological change above 1, which can be considered a positive development.

The results of the analysis indicate a natural process of introducing digitalization elements. The COVID-19 pandemic caused an acceleration in the introduction of individual digitalization elements, while their more effective use subsequently occurred after the pandemic. Acceleration appeared in the case of elements, for the introduction of which, individual countries and their companies were not sufficiently prepared.

5. Conclusions

Selected theoretical studies as well as our own research indicate that the digitalization of businesses contributes to achieving better economic results. It is necessary to point out the fact that when analyzing the level of digital transformation, it is important to assess not only the percentage of the introduction of the digital component, but also to assess the development over time. The MI DEA seems to be an important tool that enables this comparison over time. In addition, the contribution of the MI DEA is also the determination of the change in technical efficiency and technological change in businesses within the EU countries. Based on the MI DEA, it is possible to point out the progress of EU countries that are not among the leaders in the given field. However, these countries do use the already established digitalization elements effectively (Bulgaria, Greece, Romania, and Slovenia—these countries achieved ECH level 1). It is very difficult to compare these research results with other similar studies, since in this study different inputs and outputs were used in the MI DEA compared with other studies. However, indicators that are more focused on the performance of businesses (Cash/Firm Value (CFV), Enterprise Value/EBIT, and Enterprise Value/Sales) were used as outputs. A significant limitation of the research was the number of EU countries, as a result of which only a limited number of inputs and outputs could enter the MI DEA model. Therefore, it is not possible to draw definite conclusions. However, it is possible to point out what has already been stated, that monitoring the development and efficiency of the digital transformation over time has its justifications and brings interesting results. Another significant limitation was the absence of data on the digitalization of businesses within the EU countries for individual years. Nevertheless, the results can be beneficial for individual countries and their businesses, since despite the initial high costs of introducing individual digitalization elements, the efficient use of these elements can contribute to an increase in their performance. This is a prerequisite for ensuring competitiveness in the European environment. Future research will be aimed at the selection of new digitalization indicators, which will be the result of a combination of the domain knowledge approach and the selection of indicators using exact methods. We will also focus on the collection of missing data in the analyzed years. The results of the efficiency of digital transformation of EU companies can be beneficial for specific companies from individual EU countries when introducing elements of digital transformation. They are a confirmation of their significant contribution to the growth of efficiency and performance of companies.