Press release distribution market size in England: what official data can and cannot show

Classification does not describe the purchase

Neither SIC 63.91 nor 70.21 defines press release distribution on its own. ONS places news agencies supplying news, pictures and features to media in class 63.91, and public-relations and communication advice in class 70.21. Those groupings do not identify revenue from transmitting client announcements.

A supplier may use either code, another principal activity, or several codes across a group. The UK SIC 2007 structure sets out the classifications. Provider functions can be compared, but that does not turn the classifications into a niche-wide revenue measure. This article found no official statistic for England that reports revenue from this service alone, and commercial market reports rarely disclose their scope or method, so treat any single published total with caution.

The purchase boundary matters as much as the industry label. Buyer-owned newsroom publishing, direct media outreach, wire transmission, paid placement, monitoring and earned editorial coverage involve different payers and outputs. Combining them would produce an inconsistent total, even if each provider had been identified correctly.

The register is a frame for finding candidates

The ONS business register can help create a broad sampling frame, but it does not report sales of press release distribution. Its value here is as a starting point for locating enterprises and local units, not as a direct count of suppliers, buyers or transactions in the service.

The 2025 release appeared on 24 September 2025 and counts enterprises and local units by industry, region, legal form, employment and turnover band. Its register snapshot is dated 14 March 2025. These details define the coverage and timing of the frame rather than establishing activity in a later period.

The business register methodology says it covers VAT or PAYE businesses, leaving out low-turnover non-employing businesses outside those systems. An enterprise location generally represents its main operating site or head office, while turnover covers the whole UK enterprise. Most turnover data in the 2025 release relates to financial year 2023 to 2024.

The register does not identify service mix, customers in England, distribution fees, paid-placement spend or activity during 2026. An industry count is therefore a proxy with known omissions, not a service measure. Use it to identify candidates, then verify each one against its own website, Companies House record and a contract or invoice.

Supplier records need evidence of activity

Candidate providers to verify include PA Media Group's PA Newswire, PR Newswire (part of Cision), Business Wire, GlobeNewswire and RealWire, alongside smaller UK services and agencies that resell distribution. Being on this list says nothing about size or revenue; each must pass the checks below.

Begin a supplier file with a frozen extraction date and candidate enterprises or local units drawn from disclosed SIC classes. Preserve original codes, then check legal identity, trading name, service functions and availability in England. Deduplicate groups, and record non-response as unknown rather than treating it as proof of inactivity.

Companies House holds each company's filed accounts, officers and group links. Many small companies file accounts without a profit and loss account, so turnover may be missing; record it as unknown, and ask the supplier or use invoices instead.

Keep ownership, duplicate trading names and changes in service as separate fields. Each included legal entity should be traceable to evidence by another researcher. Where a provider performs several functions, include only transactions identified by a contract or invoice. Do not allocate mixed turnover by intuition or code alone.

Buyer records need a documented announcement need

Buyer evidence requires a defined England audience, a set observation period and organisations with a documented announcement need and an authorised decision owner. Record the route each organisation currently uses, including an owned page or direct media contact where no external service is involved.

A broad industry population is not automatically a population of buyers. Record the selected job and route consistently so different service routes are not silently combined. A survey tool can collect the answers, while the population and observation dates still need to be fixed.

Purchase records must use a consistent unit

Capture completed purchases in a common currency, period and unit. Separate transmission, writing, monitoring, paid-media pass-through, internal labour and VAT rather than presenting unlike items as one charge. The budget ledgers provide a way to keep those categories distinct.

Record every amount excluding VAT, so the total does not depend on whether a buyer can reclaim it. Check HMRC's published VAT guidance for the rate on a particular supply, and ask a qualified tax adviser where it is unclear. Separate completed invoices from quotations.

From records to a market estimate

Once the records exist, the estimate is simple arithmetic on a sample, not a guess from codes.

As illustrative arithmetic only, not a measurement: if a stratum holds 2,000 organisations, 5% of the sample bought the service and the mean spend was £1,200, that stratum contributes 2,000 x 0.05 x £1,200, or £120,000. Real results will differ, and a small sample gives a wide range.

A reported figure needs aligned evidence

A result is fit to report only when another researcher can reproduce the extraction and the numerator and denominator match on service, population, geography and period. Report records found, eligible, verified, excluded and unresolved in each file; do not add supplier counts to buyer counts or quotations to completed invoices.

Make uncertainty visible, including non-response, classification error, group structures and businesses omitted from the register. Until the evidence supports an aligned measure, record market size as not directly measured. A defined buyer problem may be a more useful research decision than an unsupported total.

State the service definition, geography rule, dates, source versions and missing populations alongside any result. That record lets readers see where the evidence ends, rather than treating an industry classification or broad register count as a direct measure of England's press release distribution market.