Barriers to Adopting AI Technology in SMEs: A Multiple-case Study on Perceived Barriers Discouraging Nordic Small and Medium-sized Enterprises to Adopt Artificial Intelligence-based Solutions

Axel Aarstad & Michal Saidl

Student thesis: Master thesis

Abstract

The focus of this thesis is Artificial Intelligence (AI) technology adoption constraints in small and mediumsized enterprises (SMEs). Today, only 5% of SMEs in Europe have engaged in the use of AI technology. Compared to larger organizations, SMEs are vastly underrepresented and face the risk of losing their competitiveness. The issue was addressed by exploring the following research question: “Why are some SMEs hesitant with adopting AI technology?” Previous literature and research on AI application in business, technology adoption in SMEs, and Digital Transformation in SMEs was reviewed which led to ten concepts that potentially affect the outcome of an AI adoption decision process: AI Value Perception, AI Black Box, Data Ecosystem Requirements, Strategy and Resources, Digital Transformation Capabilities, Organization Readiness, Management Support, AI Talent, Risk Perception, AI Technology Accessibility. The concepts were used in combination with the TechnologyOrganization-Environment (TOE) framework as a research lens. As the next step, four objectives related to the research question were set with the main one being: “Explain what factors come into play, discouraging SMEs from engaging in AI-investments”. Subsequently, the following research methods were applied as they were relevant for this study: an exploratory and pragmatic approach, both abductive and inductive reasoning, multiple-case study research design, qualitative data collection strategy, and qualitative data analysis through coding, theming and categorizing. The data was collected using non-standardized semi-structured, open-ended interviews from eight representatives of four Nordic SMEs. The interviewed representatives were executives, senior employees or decision-makers that would be involved in a technology adoption decision. The interviews were recorded and transcribed using the Otter.ai tool and analyzed using software NVivo 12, Microsoft Word, and Microsoft Excel in three phases. The analysis process led to the result of 65 themes representing perceived barriers preventing SMEs to engage in applying AI technology. A hypothesis of the 20 most significant barriers hindering SMEs to adopt AI technology was constructed (see chapters 6.6 and 6.7). These found barriers were (1) Lack of AI competence, (2) Dependency on external help, (3) Lack of IT competence or knowledge, (4) No or little prior AI experience, (5) AI or technology scepticism, (6) Change resistance, (7) Unclear benefits of an AI initiative, (8) Competing priorities, (9) Employee age, (10) Firefighting, (11) Lack of AI understanding, (12) Resources constraints, (13) Lack of clear business case and strategy, (14) Insufficient employee training, (15) Financial constraints, (16) Incompatibility of an AI solution with an organization's legacy IT systems or processes, (17) Not following AI trends, (18) Price of an AI solution, (19) Risk of losing reputation and damaging customer relationships, (20) Tasks or processes that are challenging to streamline. This preliminary study contributes to identifying perceived barriers to engage with AI technology that specifically apply to SMEs and invites researchers to further study this field as it is not sufficiently researched.

EducationsMSc in Business Administration and E-business, (Graduate Programme) Final Thesis
LanguageEnglish
Publication date2019
Number of pages146