What can honestly be measured about England's paid social agency market

No current official series measures the paid social agencies market size in England. Published data can show how many businesses are based in England, how broad advertising classifications are and how much was spent on social advertising across the UK. It cannot reveal the number or income of English specialist agencies.

The defensible answer is therefore a bounded account estimate, accompanied by direct buying observations. It should not be presented as a national market value.

What to take away

Why the obvious proxy is too wide

The ONS UK SIC 2007 index places advertising agencies, consultants and campaign creation within class 73110. Paid-social specialists may sit there, but so can agencies working across print, broadcast, search or other services. A specialist may also be registered under a different principal activity.

Counting 73110 enterprises consequently gives a broad advertising-supplier proxy. It is not a clean specialist-provider total. Turnover for that class would carry the same service-mix problem and must not be relabelled as paid-social revenue.

The 2025 ONS business activity dataset helps with geography and unit choice. Its IDBR snapshot reports enterprises and local units, and the accompanying notes explain that an enterprise generally follows its main site or head office while a local unit represents a particular operating site. Select one. Adding both would double-count parts of the same organisations.

Start from buyers, not suppliers

For a serviceable-account model, an England denominator beats an agency count. The Department for Business and Trade Business Population Estimates 2025 estimated 4,996,000 private-sector businesses in England at the start of 2025, with a 95% confidence interval of plus or minus 117,000.

The release is official statistics in development. It includes estimates for unregistered businesses and locates businesses by head office.

That 4,996,000 is a population denominator, not a prospect list. It excludes public and not-for-profit bodies from the main private-sector count, includes many organisations with no plausible need or budget, and does not indicate advertising activity. An analyst should choose either this broader estimated population or the registered ONS frame according to the intended customers, never splice the totals together.

Screen the chosen frame against declared conditions: English head office, selected sectors and sizes, a product that can be advertised on the proposed platforms, an accountable buyer, suitable account access and a procurement route. Report:

eligible accounts = reviewed denominator records - documented exclusions

Where only a sample is checked, publish the selection method, sample size, missing records and uncertainty. A low-volume pilot can use a verified list rather than an extrapolation.

Keep spend in its proper geography and unit

IAB UK's 2025 Digital Adspend release reports £11.5 billion of UK social media advertising spend for that year. Its method combined submitted and public data with statistical modelling, and the edition changed methodology. The figure covers media expenditure across the UK. It does not isolate England, identify agency-managed spend or state agency fees.

Direct procurement records answer a narrower question. An AHDB 2025 tender notice specified paid social media campaigns within a wider advertising and promotion requirement. This proves that one documented buyer included the capability. It supplies neither prevalence nor typical contract value.

Publish a range only after observation

A 2027 planning range should come from verified accounts under one method. The lower case can include signed briefs or completed procurements. A working case may add accounts with a confirmed owner, need and timetable. Label both as scenarios, state the observation window and replace assumptions after each research cycle.

Before publication, a named UK statistician should check the denominator and uncertainty, while a commercial reviewer checks revenue units and exclusions. Until those reviews occur, the estimate remains an internal decision aid, not a market-size claim.

Before you act

Common questions

Why can't ONS class 73110 give a clean total for paid social agencies?

ONS class 73110 covers advertising agencies, consultants and campaign creation across print, broadcast, search and other services. Paid-social specialists may sit there, but so can many other agencies. A specialist may also be registered under a different principal activity. Counting 73110 enterprises gives a broad advertising-supplier proxy, not a clean specialist-provider total.

What does the IAB UK 2025 social media adspend figure actually measure?

IAB UK reports £11.5 billion of UK social media advertising spend for 2025. Its method combined submitted and public data with statistical modelling, and the edition changed methodology. The figure covers media expenditure across the UK. It does not isolate England, identify agency-managed spend or state agency fees.

What should be done before publishing a planning range for England's paid social agency market?

A 2027 planning range should come from verified accounts under one method. The lower case can include signed briefs or completed procurements. A working case may add accounts with a confirmed owner, need and timetable. Label both as scenarios, state the observation window and replace assumptions after each research cycle.