Influence of SME’s size on internet connectivityMilanka Bogavac, Assistant Professor, Ph.D.Dr. Zoran Čekerevac, Full Professor“Union - Nikola Tesla” University, Belgrade, SerbiaINFLUENCE OF SME’s SIZE ON INTERNET CONNECTIVITYAbstract. This paper presents an overview of the part of the research performed to definethe ways and criteria for determining the degree of success of the digitalization of small andmedium-sized enterprises (SMEs). During the research, an Index of the Digitalization of SMEs(IDSME) was defined. After a brief introduction into the IDSME index, because of the importanceof Internet connections, the authors analyzed whether and how the company size is connected to theInternet connectivity dimension. The research covered 226 companies of different sizes andbusiness orientations from Slovakia, Serbia, and Russia. During the research, the main researchquestion was: What is the relationship between the size of SMEs and their links to the Internetexpressed through the "Connection to the Internet" dimension of the IDSME index? The answer tothis question was found by examining the correlation between these two variables, where the"Connection to the Internet" dimension appeared as the dependent variable. The independentvariable was the number of employees in SMEs.Keywords: Digitalization, SME, Internet, IDSME Index, Business Process.1IntroductionDue to the development of the market and automation, the elimination of bordersand competition, or the inability to adapt to the new conditions, many companies arefailing. Small and medium-sized enterprises (SMEs) have appeared as possible waysof survival of their owners and employees. It is good that SMEs can quickly adapt tonew business conditions, but their size is often a problem because of the size of boththe material and human resources available to them. More about SMEs is discussed inmany literature sources, such as, for example, [1], [2], [3], etc. It can be noted thataccording to the data of the Serbian Chamber of Commerce [4] 340,112 SMEs with837,532 employees operated in Serbia in 2016, making 99.8% of the number ofcompanies, and by number employees 65.7%. In Germany, one of the most developedeconomies in the world, 99.5% of companies in 2018 were classified in the SME group[5, p. 2]. A very similar situation is also in other countries.Digitalization is a term that is widely used nowadays. Many criteria have beencreated to point out to the level of digitalization of individual economies and countries,e.g. DESI - Digital Economy and Society Index [6], I-DESI International DigitalEconomy Index [7], EDI - Enabling Digitalization Index [8], GCI - GlobalConnectivity Index [9], and other indices, but the area of SMEs remained neglected.Considering the data presented and the importance of SMEs, it is very importantto consider the factors that influence and can influence the efficiency and quality of thework of these legal entities. Bearing in mind the ubiquity of information technologies,digitalization is imposed as an undeniable influence on the functioning of small andmedium-sized companies.11

Milanka Bogavac, Zoran ČekerevacThis research was carried out as part of the preparation of the doctoraldissertation [10], a broader study of the impact of business process digitalization onincreasing the performance and sustainability of SMEs. It was carried out in threecountries, Slovakia, Serbia, and Russia. In this paper, the accent is placed on thedimension of the Connection to the Internet. Since authors are not aware that theprevious research on digitalization measurements at the level of SMEs exists, theauthors consider this research a pioneering work in that field.2The SME Digitalization IndexThe SME Digitalization Index (IDSME) allows SMEs to carry out selfevaluation and determine to what level they are digitalized and on what they need topay attention to in their development plans. Its dimensions are Connection to theInternet, Digital Skills, Integration of Digital Technologies, and Internet Usage. TheIDSME measures the progress of SMEs in the digitalization process. It was developedon the principles and recommendations given in the handbook on constructingcomposite indicators [11]. Twenty-eight indicators are grouped into 10 sub-dimensionsand four dimensions. Dimensions and sub-dimensions are weighted, but indicators arenot. The aggregation method is presented in Section 2.2.The structure of the IDSME index and weightings are explained in detail in [10]and [12]. Therefore, they will not be fully discussed here, and only the part that relatesto this research will be shown here.2.1 Dimension – Connection to the InternetThis dimension, the connection to the Internet is today the precondition of allconditions. Except in extremely rare applications, e.g. connecting ATMs or fiscal cashregisters, dial-up connections to the Internet cannot be considered a favorable solution.Therefore, fast internet connections can be considered as a very significant dimensionof IDSME indicators. This dimension is divided into four sub-dimensions, eachfocusing on one aspect of connecting to the Internet: Connectivity to broadband Internet. Connection to the Internet via a public telecommunication network. Internet speed as a factor influences processes in SMEs and focuses on theability (and desire) of a business to make better use of the Internet's capabilities. As the"high-speed Internet" limit, the declared value of the Internet access speed of 30 Mbpsis taken because of the comparability of companies from different countries. At thetime of writing this work, the standard offer in Serbia was 20/4 Mbps and the optimal100/10 Mbps. The boundary speed will certainly be a parameter that will change overtime. As an indicator in the model, a subscription to a fast broadband Internet abovethe limit value in the year under review was considered a limit value subscription. Table1 shows the structure of SMEs' Internet connectivity.Table 1. Results of the survey on the connection of SMEs to the Internet.SMEsMediumDial-up40.4%Cable91.5%Broadband InternetMediumMediumdeclareddeclaredspeed [Mb/s] Mobile speed [Mb/s]117.751.1%41.312

Influence of SME’s size on internet %49.4%108.757.8%36.0Total - have53.5%75.2%115.056.2%36.5Source: Authors’ research Possibility to work from a remote location (or the possibility to work on aremote computer) is an option that has been seriously talked about for some twentyyears, and which gets its importance in time especially in crises such as with COVID19. This sub-dimension focuses on the willingness of SMEs to organize and usedistance working. As a criterion, the percentage of employees is taken here about thetotal number of employees. It is obvious that if corrections are not introduced, theremay be unjustified favorability for micro-companies to others. Therefore, the modelset the limit to 10% of the total number of employees. With ten or more percent ofthose who worked remotely the surveyed SMEs achieved the maximum number ofpoints according to this criterion. Survey results related to the possibility to work froma remote location for this set of SMEs are shown in Table 2.Table 2. Survey results related to the possibility to work from a remote 0%56.6%52.2%Criterion A6.9%13.4%54.3%10.4%Criterion B3.3%6.5%27.7%5.0%Criterion A – % of those who worked remotely related to the number ofemployees in SMEs that enable distance workCriterion B – % of those who worked remotely related to the number ofemployees in SMEsSource: Authors’ research2.2 AggregationThe aggregation of indicators in the Internet connectivity dimension wasconducted on the way that the indicators were aggregated into the sub-dimensions, andthen the weighted values of the sub-dimensions were aggregated into dimensionInternet connectivity.The aggregation of indicators into dimension was done as follows:1.Before aggregation, the normalization of the indicators was done to unifythe units. It was performed using the min-max method by the linear projection of eachindicator on a scale in the range 0 to 1. The zero corresponds to the minimum valueand the one to the maximum value (see Table 4).2.In the second stage, the indicators are aggregated into sub-dimensions bysimply summing their values.3.In the third stage, the sub-dimensions were weighted using the w valuesshown in Table 4 and by summing aggregated into the Internet connectivity dimension.13

Milanka Bogavac, Zoran ČekerevacTable 3. Structure of inclusion the indicators into the dimensions Connection to theInternetDimension1.Connectionto ection to a Possession of1a. Connectivity tofixedanactivebroadband InternetbroadbandconnectionInternet1a2.Connection to a Possession of(w 25%)mobileanactivebroadbandconnectionInternet1b. Connection to1b1.the Internet via aPossession ofConnection topublicanactivethe Internet viatelecommunicationconnectionphone linenetwork (w 15%)min max0101011c1.1c. Internet speedSubscription to(w 30%)fast BB accessDeclaredaccess speed 30Mbps011d. Possibility toworkfromaremotelocation(w 30%)% ofemployeeswho used thisopportunity inthe last threemonths0101d1. Users ofthe workingfrom remotelocation optionSource: [12]In the case of aggregation of the Internet connectivity dimension (ICD) for thei SME, the following formula was applied [12]:thICD(SMEi) Connection to broadband Internet(SMEi) * 0.25 Connection via a public telecommunication network(SMEi) * 0.15 Internet speed(SMEi) * 0.3 Share of distance working(SMEi) * 0.33Methodology of ResearchThis research is exploratory with elements of explanatory and descriptive. Theresearch used an empirical method as an analytical method that enables reliableconclusions on the interdependence of certain observed elements and trends inindividual phenomena. The statistical analysis covered the relevant data that enabledthe detection of the legality of mass events covered by this analysis.14

Influence of SME’s size on internet connectivityOf the general methodological methods in this research, analytical and syntheticmethods, induction and deduction, as well as historical and comparative methods, wereused. When necessary, multivariate analysis was also used. The research is based onchecking hypotheses. Although the samples should be randomly selected, it was notpossible to provide absolute randomness of the samples in this survey because theSMEs’ addresses were mostly obtained by Internet mining. It is acceptable because thetopic of work is focused on studying the impact of digitalization on the work of SMEsthat use the Internet in some way.3.1 The research question and results of the researchBy applying the presented methodology to the sample of 226 respondents, theICD dimension for each SME was calculated, and the results were grouped accordingto the size of the SME and the country in which it operates. In the calculations, theRegression option was used within the Data Analysis of the MS Excel program, a partof the MS Office 365 software package.For SMEs that were reliably identified to work in a particular country, the resultswere grouped and the average ICDs for a particular country and all SMEs covered bythe survey were calculate.During the research, the research question (A) arose: What is the relationshipbetween the size of SMEs and their links to the Internet expressed through the"Connection to the Internet" dimension of the IDSME index?The answer to this question was found by examining the correlation betweenthese two variables. The subject of research is the dimension of "Connection to theInternet", so it appears as the dependent variable (y), and the independent variable (x)is the number of employees in SMEs. For this analysis, the significance limit has beenadopted at the level of 0.05 (5%). The following indicators were considered in the"Connection to the Internet" dimension:Is SME connected to a broadband Internet access network?Is SME connected to the mobile Internet?Is SME connected to the Internet via a fixed telephone network?Whether at least one connection method provides an Internet access speedgreater than or equal to 30 Mbps?The share of those who worked from a remote location in the previousthree months in the total number of employees.The null hypothesis for the research question A was defined as:HA0: There is no strong relationship between the size of SME and its connectionto the Internet expressed through the "Connection to the Internet" dimension of theIDSME Index.The alternative hypothesis for the research question A is then:HA1: There is a strong connection between the size of SMEs and its connectionto the Internet expressed through the "Connection to the Internet" dimension of theIDSME index.The results of statistical data processing for the relationship between the size ofSMEs and its connection to the Internet expressed through the dimension "Connectionto the Internet" of the IDSME index for this set of SMEs are shown in Table 5.15

Milanka Bogavac, Zoran ČekerevacTable 5.Results of statistical data processing for the relationship between the sizeof SMEs and its connection to the Internet expressed through the dimension"Connection to the Internet"Regression StatisticsMultiple R0.07754R Square0.006013Adjusted R Square0.001576Standard icients Errort Stat P-value 95%95%Intercept0.085100.00309227.52 8.283E-74 0.079010.0912X Variable 1 4.636E-05 3.983E-05 1.16 0.24563 -3.212E-05 0.000125For this case, the value of the Pearson’s coefficient of correlation is ρ 0.07754,which is significantly less than 0.3 and very close to zero. The P-value is greater thanthe limit value of 0.05, so the results obtained for this set of data do not reject the nullhypothesis HA0.4CONCLUSIONSIn analyzing the results obtained, it should first be pointed out that the IDSMEindex is primarily intended for self-evaluation of SMEs, allowing them to see theirachievements and weaknesses, and to establish plans for improving their businesses.This analysis of the influence of SME size on Internet connectivity conductedusing the appropriate IDSME index dimension showed that there is no strongcorrelation between the size of SMEs and their links to the Internet expressed throughthe dimension "Connection to the Internet" of the IDSM index. The probable reason isthat the sample covers the whole range of SMEs that differ in their origin, maturity,size, and activity, so it is not possible to detect significant interdependence. If thesample included a huge number of similar SMEs, it would be possible to establish amore significant dependency. Obtained results show that each SME must be connectedto the Internet in its way, adapted to their current needs. It can be expected that largercompanies have a greater need for faster Internet access, but also that the need for fastInternet access depends not only on the number of users but also on the type of businessactivity.Although the research results indicate that there is no dependency between thesize of the company and the analyzed characteristics, the results are significant becausethey are part of a wider mosaic that enables the company to look at its position in thebusiness environment and to select elements that need to be significantly influencedand elements that can be less affected. If planning for its growth, the company can seethat over-investing in this dimension will not have a significant effect on the company'sgrowth, although it will most likely have a positive impact on the business results.References1. H.-J. Chang, "A brief history of capitalism. How have we got here?," in Economics,the user's guide, Bloomsbury Press, 2014, p. 384.16

Influence of SME’s size on internet connectivity2J. Bolton, Report of the Committee of Enquiry into Small Firms, London: HMSO,1971.3. J. A. Schumpeter, The Theory of Economic Development: An Inquiry into Profits,Capital, Credit, Interest, and the Business Cycle, New Brunswick, New Jersey:Transaction Books, 1934; 1983; 2008.4. N. Čokorilo, S. Nikolić, S. Milošević, D. Lazić, M. Nikolić, M. Trajković and S.Đurić, "Izveštaj o malim i srednjim preduzećima i preduzetništvu," 11 ed 15 10 2018].5. EC, 2019 SBA Fact Sheet - Germany, European Commission, 2017.6. F. Benini, «The Digital Economy and Society Index (DESI),» 26 10 2018.[Онлайновий]. Available: .7. P. Foley, D. Sutton, I. Wiseman, L. Green та J. Moore, International DigitalEconomy and Society Index 2018, Luxembourg: Publications Office of the EuropeanUnion, 2018.8. E. Hermes, «Enabling Digitalization Index 2018: Measuring digitagility,» 15 032018. [Онлайновий]. Available: x-2018-measuring-digitagility.html.9. .10. M. Bogavac, Research on the Influence of Digitalization on Small and MediumEnterprises, Belgrade: ICIM , Faculty of Business and Law, "Union - Nikola Tesla"University, 2019.11. OECD, Handbook on constructing composite indicators: methodology and userguide, Paris: OECD Publishing, 2018.12 . M. Bogavac and Z. Čekerevac, "IDSME Index – New Method for Evaluation ofSMEs Digitalization," MEST Journal, vol. 7, no. 2, 15 July 2019.17

Economy Index [7], EDI - Enabling Digitalization Index [8], GCI - Global Connectivity Index [9], and other indices, but the area of SMEs remained neglected. . Integration of Digital Technologies, and Internet Usage. The IDSME measures the progress of SMEs in the digitalization process. It was developed on the principles and recommendations .