AI Marketing
AI marketing statistics UK: adoption is up, revenue is not yet
Short answer: The ONS found that 29% of UK businesses used at least one AI technology in June 2026, up 8 percentage points in a year. Marketing is the joint most common area for it, named by 72% of firms using or planning AI in DSIT research. But in the same research, 77% of AI-using firms had not yet seen a change in revenue. So the adoption rate is not the number to act on. Your own enquiries and jobs won are.
UK AI adoption figures are real and rising. But they measure who has switched a tool on, not who has earned from it. A small business should read them as context, not pressure to buy.
This page sits under our guide to AI marketing for UK small businesses.
The official count: 29% of UK businesses, and rising quickly
The ONS Business Insights and Conditions Survey, published on 2 July 2026 (Wave 159, live 15 to 28 June 2026), found that 29% of businesses used at least one type of AI technology in June 2026. That is 8 percentage points higher than in June 2025.
The most adopted technology was text generation using large language models (17% of businesses), then visual content creation (14%), up 12 and 11 points since September 2023.
The ONS follow-up, Artificial intelligence in UK businesses: 2023 to 2026, adds size bands. Among businesses with 10 or more employees, self-reported use rose from around 12% in late 2023 to around 35% in June 2026.
| Size band | Share using AI | Source |
|---|---|---|
| All businesses | 29% | ONS, June 2026 |
| 0 to 9 employees | 28% | ONS article, 20 July 2026 |
| 10 or more employees | around 35% | ONS article, 20 July 2026 |
| 250 or more employees | 49% | ONS, June 2026 |
Source: the two ONS publications above.
Two cautions. The ONS labels these official statistics in development and advises caution. The Wave 159 response rate was 26.7% of 38,637 sampled businesses. The ONS article also leaves the 0 to 9 band out of some estimates, because that band’s large survey weights can distort totals.
Why the ONS, DSIT and the FSB land on 29%, 16% and 55%
Three numbers for the same question look like a contradiction. They are not, for three reasons.
First, when they asked. DSIT’s fieldwork ran in early 2025, the FSB’s in November 2025 and the ONS’s in June 2026. Adoption was rising across that stretch, so an older survey reads lower.
Second, who they asked. The ONS sampled businesses of all sizes. DSIT, in its AI Adoption Research by IFF Research and Technopolis Group, held 3,500 telephone interviews with businesses of at least 5 employees, between 12 February and 2 May 2025. The FSB, a membership body that lobbies for small firms, surveyed 904 small business owners.
Third, what counts as using AI. Pasting a prompt into a chatbot once a week may be a yes in one survey and a no in another.
Our verdict: ONS for how many, DSIT for what for and what it returned.
Marketing is the first place UK firms point AI
DSIT asked the 919 businesses using or planning AI which areas it applied to.
| Business area | Share of firms using or planning AI |
|---|---|
| Marketing | 72% |
| Administration | 72% |
| IT | 64% |
| Operations | 53% |
| Sales | 49% |
| Customer service | 46% |
Source: DSIT, AI Adoption Research (GOV.UK page last updated 13 February 2026).
Among the 700 businesses actually using AI, 85% used natural language processing and text generation, which fits the ONS finding on text generation (17%) and visual content creation (14%).
The ONS article does not break out marketing at all. So DSIT’s figure is the only official one on marketing among the sources read for this page, and it dates from early 2025.
Our reading is that marketing leads because drafting text and images is the cheapest thing to try, not because it is where a return has been proven.
Wide but shallow: 1.6 tools and a minority of staff
The ONS calls adoption to date “relatively shallow”. Among adopting businesses with 10 or more employees, the average number of AI technologies rose from around 1.4 in late 2023 to around 1.6 in June 2026.
Only 15% of businesses with 10 or more employees said more than half of their employees use AI in daily work. DSIT found most firms bought ready-to-use tools: for text generation, 71% bought and 14% built in house.
There is also a gap. The ONS reports that 55% of employees say they use AI for work or education, against around 35% of businesses with 10 or more employees reporting use. Our reading, an inference, is that staff are using AI the business does not count or control. That is where trouble starts, and we cover which AI marketing mistakes a UK regulator can fine you for separately.
The revenue question: 77% have seen no change yet
DSIT asked the 700 businesses using AI what had changed. 75% reported improved workforce productivity. 77% had not yet seen a change in revenue, and 12% reported an increase.
The FSB report The Confidence Code reads more warmly. Among users, 59% said AI improved productivity, 24% reported higher revenues and 22% said it helped them grow. The FSB also estimates adopters have seen an average 3% increase in revenue, and that wider adoption could add more than £42 billion a year to the UK economy. Those are the FSB’s own estimates, from a lobby group’s own survey, not measured outcomes.
Do not compare them directly. The surveys ran months apart and asked different firms.
Our opinion: time saved is real. Revenue is not yet shown. Anyone quoting a return multiple for AI marketing to a UK small business is not quoting official data.
What UK adopters actually spend: a £2,000 median
DSIT asked about AI-related spending in 2024, among businesses using or planning AI that could give a figure. The mean was £19,000. The median was £2,000. And 31% reported spending nothing at all.
The gap between the two averages means a handful of large spenders drag the mean up. A small firm should read the median, which describes the typical business that could put a number on it.
Tool spend is only part of the bill. An agency fee sits on top, and our guide to what digital marketing costs in the UK gives that context.
London leads, Scotland lags, construction trails everyone
DSIT found London at 20% using AI, against 16% overall. In Scotland, 84% of businesses neither used nor planned to use AI, against 80% overall. We did not find comparable adoption rates for Wales or Northern Ireland in the sources read here, so we quote none.
By industry, the ONS has information and communication at 58% and construction at 13%. These are two different surveys, but on their face the gap between sectors is far wider than the gap between London and everywhere else.
Among businesses with no plans to adopt, 81% told DSIT the biggest factor was no identified need. The ONS found 41% of businesses with 10 or more employees reported no barriers to adoption. For many firms, not using AI is a decision, not an oversight.
Where £40.5bn of UK digital ad spend went in 2025
IAB UK’s Digital Adspend 2025, published on 3 March 2026 with Oliver Wyman, put the UK digital ad market at £40.5bn in 2025, up 10%.
| Line | 2025 spend | Growth |
|---|---|---|
| Search | £17.9bn (44% of total) | up 6% |
| Social | £11.5bn (28% of total) | up 21% |
| Video | £9.3bn | up 20% |
| Retail media | £3.8bn | up 18% |
Video spans social platforms, broadcasters and publishers, so the lines overlap and do not add to the total. In IAB UK’s industry sentiment survey, 48% of respondents named AI and automation as a defining force for the sector over the next decade.
This is market spend by all advertisers, not a small business benchmark. For smaller budgets, see digital advertising for UK small businesses.
A hypothetical: a Stockport joinery firm reads these numbers
This is an invented example to show the arithmetic. It is not a real client.
A four-person joinery firm in Stockport spends the DSIT median, £2,000 a year, on AI tools for quote replies and write-ups of finished jobs. Assume, purely for illustration, a 30% margin and an average job of £3,400.
To cover £2,000 at a 30% margin, the firm needs about £6,667 of extra sales (2,000 divided by 0.30). At £3,400 a job, that is about two extra jobs a year (6,667 divided by 3,400 is 1.96).
So the test is not whether the firm has “adopted AI”. It is whether the firm can point to two jobs it would not otherwise have won. To check, log reply time and jobs won before and after, then compare with the same months last year.
Numbers we would not put in a pitch deck
We left some figures out on purpose.
- Vendor return-on-investment multiples. No stated method, so nothing to check.
- Any UK figure on AI Overviews and clicks. We read none with a stated UK method.
- The FSB’s £42bn. A forecast from a membership body, not a measurement.
- Any “X% of marketers use AI” poll from a software company. The seller wrote the question.
How to use the rest, in three instructions:
- Quote the ONS figure for how many firms use AI, with its caution label.
- Use DSIT’s median spend and revenue findings for what a typical adopter gets, and say they date from early 2025.
- Measure your own enquiries and jobs won before and after any tool. That beats every survey above.
We sell this kind of work through our AI automation service, so we have an interest. That is a reason to be strict about the evidence.
Where each figure comes from
- ONS, Business insights and impact on the UK economy, 2 July 2026.
- ONS, Artificial intelligence in UK businesses: 2023 to 2026, 20 July 2026.
- DSIT, AI Adoption Research, by IFF Research and Technopolis Group, page updated 13 February 2026.
- FSB, The Confidence Code, 2026 (press release linked).
- IAB UK, Digital Adspend 2025, 3 March 2026.
The FSB figures come from a survey run by a membership body, and we read them through press coverage of the report (Business Times, 23 June 2026), not the report itself. The ONS figures are official statistics in development.
Frequently asked questions
What percentage of UK businesses use AI in 2026?
The ONS found 29% of UK businesses used at least one AI technology in June 2026, up 8 percentage points on a year earlier. Among firms with 10 or more employees the figure was around 35%, and among those with 250 or more it was 49%. The ONS labels the data official statistics in development.
Why do UK AI adoption surveys give such different figures?
They asked different people at different times. DSIT phoned firms with at least 5 employees in early 2025 and found 16%. The FSB surveyed small business owners in November 2025 and found 55%. The ONS surveyed businesses of all sizes in June 2026 and found 29%. Definitions of using AI also differ.
What do UK businesses use AI for most?
Marketing, jointly with administration. DSIT found 72% of businesses using or planning to use AI named marketing, then IT at 64%, operations 53%, sales 49% and customer service 46%. Among users, 85% used natural language processing and text generation. The ONS does not break marketing out.
Has AI increased revenue for UK businesses?
Not yet shown in official data. DSIT found 77% of AI-using businesses had seen no change in revenue and 12% reported an increase, while 75% reported better workforce productivity. The FSB, a lobby group, says 24% of small firm users report higher revenues. The two surveys ran months apart and asked different firms.
How much do UK businesses spend on AI?
DSIT found a median of £2,000 on AI in 2024, among businesses using or planning AI that could give a figure. The mean was £19,000, which suggests a few large spenders pull it up, and 31% reported spending nothing. Most bought ready-made tools rather than building their own.
Which UK regions and industries use AI most?
London led in DSIT’s research at 20% using AI against 16% overall, while 84% of Scottish firms neither used nor planned it. The ONS found information and communication at 58% and construction at 13%. We did not find comparable rates for Wales or Northern Ireland in the sources read here.
Want to know whether your own AI tools are winning you enquiries, or only saving you time? Get a free AI marketing audit. No jargon, no pressure.
Nexiiom Team
AI-powered marketing for growing businesses. We write about what actually works: automation, ads, websites and AI search.