What SME surveys do and do not tell you about digital advertising
The official UK, US and euro area business surveys we checked do not measure small firms' online advertising. Here is what they do cover, and how to build a benchmark anyway.

The official business surveys we checked do not tell a small firm how its online advertising compares with its peers. The Office for National Statistics in the UK, the US Census Bureau and the European Central Bank each publish useful business data. None of the pages we reviewed measures small firms' use of digital advertising, the channels they choose or where they report difficulty. A benchmark for a media budget will have to come from somewhere else.
What the data shows
The Office for National Statistics publishes a dataset list under its digital economy heading. It describes annual use-of-ICT and e-commerce data for UK businesses, covering how businesses use the internet. The listing gives no advertising figures. A reader learns how UK businesses use technology in general, and nothing about how they promote themselves.
A second ONS listing covers business e-commerce sales by size, location and sector. Its historical series runs only from 2006 to 2010, according to the same dataset list. A series that stops in 2010 can show how online selling moved in those years. It cannot describe the channels a firm uses in 2026. Anyone who quotes it as current evidence is working from old material.
The ONS Business Insights and Conditions Survey is a different kind of product. It publishes weighted fortnightly estimates on financial performance, workforce, prices, trade and resilience. Its topic list does not include advertising. The ONS marks the survey as official statistics in development. A leader should treat its figures as provisional and read the methods note before quoting any one of them.
The ONS Management and Expectations Survey is a second-wave pilot. It covers British management practices and growth expectations. Growth expectations sit closest to marketing in the ONS set, but the dataset list names no advertising topic. A pilot is a weak base for a benchmark, and this one does not reach the question a marketing director would ask.
In the United States, the Census Bureau's Annual Business Survey measures business owner demographics, research and development, innovation and other topics. It is a joint project with NCSES and the NSF, according to the Census Bureau page. The programme page names no advertising topic. Innovation measures are the nearest match a reader will find, and they describe what a firm builds, not how it sells what it builds.
The European Central Bank's Survey on the Access to Finance of Enterprises, known as SAFE, has run quarterly since 2024 Q1. It covers financing conditions, not marketing. Fieldwork is carried out by Ipsos Public Affairs, and the survey covers euro area and EU firms. Its results are broken down by firm size, sector and country. It is the only source here that can place a small firm against others by size, though only on finance.
The 40th round was conducted between 26 August and 2 October 2026, according to the ECB's survey page. The survey reports that firms see further net tightening of bank loan conditions. That is a finding about credit. The survey contains no advertising or digital-channel data, so it cannot tell a firm what other small firms spend on promotion, or where they find it hard.
The Companies House blog post on accounts advice for small companies was first published in 2016. It covers small company accounts filing options from 1 January 2016. It has no bearing on advertising, and we have not used it for any figure. It appears in our sources only because a reader searching for small-firm data may find it first and assume it is relevant.
Taken together, the official sources describe some topics in depth. They cover how firms use technology, how they trade online in an older series, how they run their finances and management, and what they build. Advertising sits outside each of them, at least in the listings and pages we reviewed. The pattern is consistent across the three publishers, which is itself useful information for a planner.
What it means for a leader
When a survey does not measure a thing, the sensible response is not to fill the gap with a guess. An advertising adoption figure that circulates in a board pack without a named publisher is a guess with a footnote. Ask who collected it, when, from how many firms, and what counted as advertising. If those answers are missing, the figure should carry no weight in the budget.
Start from your own records. Pull two years of campaign data: spend by channel, cost per enquiry or cost per sale, and the dates each channel was switched on, paused or cut. Three questions matter more than any national average. Which channel brought in customers who stayed? Which one could you switch off without anyone noticing? Which one are you paying for out of habit?
A useful test is whether you can explain each channel to your chair in two sentences: why it is in the plan, and what result would make you stop it. A channel you cannot defend in those terms is in the plan for the wrong reason, whatever a peer survey says. Put the same test to the digital part of the plan that you would put to a new hire. What does it do, what does it cost, how will you know it worked, and who will tell you if it stops working? A channel with no owner and no measure is a standing cost.
Be wary of figures from advertising platforms and agencies as well. A figure about small firms' habits from a seller of advertising needs its method and sample before it goes anywhere near a budget. Treat it as sales material until you have seen both. The brief behind this article names no such source, and we have not relied on one.
Where your budget depends on a loan covenant, a credit line or a grant condition, check the terms before you move money between channels. Your lender, your accountant or your adviser should confirm what a change in spend means for those obligations. We cannot tell you that from here, and nothing in these sources does either.
If you do want a national benchmark, the job is to find it first and then judge it. Write to the statistical office or survey owner and ask three things: whether advertising is asked at all, which channels are listed, and whether results are published by firm size. A no to the first question closes the matter quickly, which is useful in itself. A yes to all three gives you a source worth reading in full.
Limits of the data
This article rests on four pages: the ONS dataset list, the Census Bureau programme page, the ECB survey page and a 2016 Companies House blog post. We did not review the OECD or any national SME digitalisation survey, because neither was supplied for this brief. A search of those sources may change the picture, and a benchmark built on them would be a different article from this one.
The ONS and Census pages describe datasets at the level of topic lists. A topic list is not the questionnaire. A survey can ask about online promotion inside a question that the listing does not mention. The absence of advertising from these pages is strong evidence about the listings, but it is not proof about the underlying forms. Read the questionnaire before drawing a conclusion about any survey.
Keep the limits in the board paper too. A one-line note stating which sources were checked and which were not saves a director from asking later whether the benchmark is complete. It rarely is. On the evidence available here, no figure in these sources can be quoted as a benchmark for small-firm online advertising. The honest benchmark for now is your own data, read against the questions above.
