AI Adoption: Opportunities and Challenges for Small and Medium Businesses

How to cite

Langer, Christina; Aristidou, Angela; Brynjolfsson, Erik; Vanneste, Bart. 2025. "AI Adoption: Opportunities and Challenges for Small and Medium Businesses." AEA RCT Registry. AEARCTR-0015262. https://doi.org/10.1257/rct.15262-2.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Christina Langer* , Angela Aristidou , Erik Brynjolfsson , Bart Vanneste

* Primary investigator

Registry ID
AEARCTR-0015262
DOI
https://doi.org/10.1257/rct.15262-2.0
First registered
Last updated
Status
In development
Trial dates
2025-01-30 – 2027-12-31
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/15262

Abstract

The impressive rise of Artificial Intelligence (AI) has sparked significant attention and awe but also concerns on what these new technologies may mean for the economy and society. Especially small and medium enterprises (SMEs) exhibit room to grow regarding AI adoption. However, less is known about the concerns SMEs have when it comes to AI adoption, about suitable interventions to help them overcome these barriers, and the impact of GenAI adoption on these businesses.

We aim to study the impact of interventions designed to increase AI adoption in the population of SMEs in the UK. This will give us the opportunity to analyze the effectiveness of different interventions targeted at increasing UK SMEs' AI adoption in a real-world scenario.

Experimental design

The study employs an RCT with firm-level randomization. 300-400 SMEs will participate in six half-day AI adoption workshops that take place between Dec 16 and Dec 18, 2025 (around 50 SMEs per workshop). After the workshops, we will randomize participants to the following treatment arms:

- Control: workshop only
- Treatment arm 1: workshop + free pro GenAI tool access for one year
- Treatment arm 2: workshop + free pro GenAI tool access for one year + free AI consulting for nine months

The study period spans 1 year with multiple surveys.

Design details

Randomization method
Randomization will be done in office by a computer.
Randomization unit
Firm
Clusters
300-400 firms
Observations
300-400 firms
Arms
We plan six workshops with around 50 firms each (300-400 firms in total). We will get sign-up lists for each workshop and randomly assign signed up firms that have completed the baseline survey to one of the two treatment arms or control (workshop only). We expect around 200 SMEs to be in the control group, 50 SMEs in treatment arm 1 and 50 SMEs in treatment arm 2. This allocation was determined by available intervention resources.
Primary outcomes
1. AI adoption metrics (Number of AI tools implemented, percentage of employees using AI, number of business functions using AI; self-reported survey questions) 2. Business performance metrics (revenue, new products or services, number of employees; self-reported survey questions and administrative data from Companies House)
Secondary outcomes
1. Implementation quality (AI action plan completion rate, percentage of staff trained in AI tool usage, implementation timeline achievement; self-reported survey questions) 2. Organizational development (AI readiness scores, data infrastructure development, staff attitude, implementation barriers; self-reported survey questions)

Cited by

No citation data available for this registration yet.

Keywords