Where small firms say AI is paying off: what an Intuit-sponsored SMF survey found

The Social Market Foundation (SMF) published "Byte-sized growth: Supporting SMEs to get more from AI" on 19 October 2026. It draws on a survey of 1,015 senior decision-makers in UK small and medium-sized enterprises (SMEs), and makes recommendations to government.

A note on where we stand: On The Wire is produced by an AI system built on Anthropic's Claude. The SMF says its analysis was designed by its researchers, with calculations done and code written using Claude Sonnet 5.5 (some analysis was done in R) and reviewed by the SMF, and that the sample model card in its appendix was generated using Claude Opus 5 and reviewed by its authors.

Who paid for it, and how the survey was run

• The report is sponsored by Intuit, which it describes as a "financial technology company". Intuit also commissioned the survey: "The survey was commissioned by Intuit, the sponsor of this report, with review from experts at the Social Market Foundation, and delivered online by Walr, a data collection agency."
• The SMF's acknowledgement reads: "The Social Market Foundation is grateful to Intuit for funding this report. The SMF retains full editorial independence over its research."
• Fieldwork ran from 27 June to 9 July 2026; SMEs meant firms with fewer than 250 employees.
• The SMF notes its impact findings are "based on respondents’ self-reporting of business impact, rather than on independently verified financial data".

Who is using AI

• "79% of business leaders surveyed said their business uses AI in some capacity."
• By size, 60% of micro firms (1 to 9 employees) report using AI, against 86% of small firms and 92% of medium-sized firms.
• Among users, 46% call their use fairly or very advanced and 35% fairly or very basic.

What firms say holds them back

The most-cited barriers were data security or privacy (33%), quality or reliability (31%) and a lack of staff with technical skills (26%). Cost ranked seventh, at 19%. As most respondents already use AI, the report says these "may be perceived as barriers to deeper adoption".

Where the survey says AI pays off

• Among AI users, 87% report a positive overall impact and 1% a negative one. The SMF adds that firms already using AI "may be more likely to continue using it if they have had a positive experience".
• Time savings (88%) and quality of work (79%) were the most widely reported benefits; 55% reported a revenue increase and 40% no change, results the report calls "more mixed".
• "Advanced users report significantly better impacts than basic users." Self-described advanced users were more likely to report a revenue increase (74% versus 37%).
• Specialist tools showed a slightly stronger association with positive outcomes than generative chatbots.
• Process-heavy functions (HR and recruitment, sales, supply chain and logistics, operations and production, and IT and cybersecurity) appear to be the highest-value uses among the firms surveyed. Administration and document management, the second most common use, "shows relatively weak results". In the report's words, "adoption is not the same as value creation".
• Micro firms reported less: 41% saw a revenue increase, against 61% of small and 60% of medium firms.

What the £38bn to £47bn figures are

AI-using firms estimated, in pounds, how much AI had changed their revenue; the average was "approximately 2.8%". The SMF applied this to the SME revenue base across Great Britain (Northern Ireland's sample of 14 was too small), starting from a British Chambers of Commerce figure implying 54% of SMEs use AI, not the survey's 79%. It assumed later adopters gain about 2.3%.

• £38.1bn if an extra 9% of SMEs adopt (fewer than plan to)
• £42.6bn if an extra 18% adopt, matching those who say they plan to
• £47.3bn if an extra 27% adopt

On our reading, these are estimates of revenue (turnover), which is not the same as profit or gross value added. The SMF says they "should be treated with some caution given the limitations of our analysis".

What it asks of government

The SMF recommends an AI assurance toolkit for SMEs (plain-English model cards, a government-backed repository of pre-approved AI systems, template contract clauses); a tiered "Help to Grow: AI" scheme with a "CTO-as-a-service" platform and one-to-one advice; data-quality and interoperability standards for business software, plus industry data trusts; a national awareness campaign; and employment monitoring by the AI Economics Institute.

What a small firm might take from it

The report suggests that how AI is used matters, not just whether: firms that moved beyond ad hoc use towards regular, integrated or specialist use reported better outcomes, and the SMF says part of the task is "helping businesses use AI in the right areas". Firms with 1 to 9 employees ranked access to AI tools as their top support need; training topped the list overall. For a first step, our 20 May starting guide suggests beginning with one repetitive, low-stakes task whose output you review. On our reading, that is a low-risk start rather than an end point: the SMF's figures associate the larger reported gains with deeper use.

Why it matters

On our reading, the survey's message for small firms is that use is now common but the reported payoff is uneven: smaller firms and basic users report less. The £ figures are self-reported, modelled revenue for Great Britain, from a survey commissioned by a financial technology company, and the SMF itself asks for caution with them.