Using ONS and Companies House data to find UK promotional merchandise markets

What to take away

Which ONS datasets reveal promotional merchandise markets

Start with the ONS business counts release, which publishes the number of businesses by industry, size band and region. That is your first filter: how many firms exist in each sector you sell to.

A London marketing agency buying branded drinkware looks different from a West Midlands manufacturer buying workwear. The counts let you compare both on the same basis rather than on anecdote.

The UK business activity, size and location bulletin is the practical entry point, because it breaks counts down far enough to compare Greater London with the West Midlands, Greater Manchester or West Yorkshire. You can read the latest release here: UK business; activity, size and location.

Business demography is the second dataset. It tracks creations and closures, so you can see whether a sector is adding firms or losing them. A rising count of new companies in a sector is a weak but real signal of onboarding spend, including welcome packs and event merchandise.

The current release sits at Business demography, UK.

The Annual Business Survey is the third. It covers turnover, purchases and approximate gross value added for broad sectors, which is how you move from counting firms to estimating what they spend.

Its methodology and quality notes matter more than the headline tables, because they explain coverage and revision practice: Annual Business Survey QMI.

For a wider view of what buyers are doing with their budgets, our piece on demand signals covers the softer indicators that sit alongside these tables.

Using SIC codes to isolate promotional products businesses

SIC codes are the classification system used across official UK statistics. Every registered company and every ONS table uses them, which makes them the only reliable way to join datasets.

The full list is published by government and is worth bookmarking, because codes are revised periodically and old mappings break: Standard industrial classification of economic activities (SIC).

You need two sets of codes. The first describes your own side of the market: printing, advertising, wholesale of clothing, and similar activities that promotional merchandise suppliers and distributors register under.

The second describes your buyers. A distributor selling branded tech accessories might target section 62 (computer programming and consultancy) or section 70 (head office activities). A supplier of site kit targets section 41 (construction of buildings) and section 43 (specialised construction).

A worked example makes this concrete. Suppose you sell branded workwear in the North West.

That sequence turns a national statistic into a call list. It also keeps you honest: every name on the list traces back to a code and a source.

Two cautions. Companies pick their own SIC codes at incorporation and often never update them, so a firm may be classified under something stale. And some codes cover a mix of activities, so a code match is a lead, not proof of need.

Reading Companies House filings for sector and region signals

Companies House is the register of UK companies. Its free search lets you filter by SIC code, location and status, which is the closest thing to a live view of the buyer market.

Use it for four things. First, incorporation dates: new companies are setting up offices, uniforms and client gifts. Second, registered addresses: they cluster in business parks and city centres you can visit.

Third, accounts. Small companies file abridged or micro-entity accounts, so you will not get turnover for most prospects. You will get filing timeliness, which is a rough proxy for administrative competence.

Fourth, officer and name changes. A new marketing director or a rebrand often precedes a spend on branded material, and the filing history shows both.

Regional targeting works best when you combine the register with what you know about local industry. Greater Manchester and West Yorkshire have dense professional services and manufacturing bases. Scotland has a large public sector and energy supply chain. Wales and Northern Ireland have significant food, agri-tech and public bodies.

Those patterns are visible in the register if you filter by postcode area and code. They are also the reason a single national pitch rarely lands: a Belfast public body and a Solihull engineering firm buy for different reasons.

Sizing regions with business population estimates

ONS employer surveys miss businesses with no employees, which is a large share of the UK private sector. Business population estimates fill that gap by combining survey data with administrative sources.

The 2023 release is the reference point for the private sector total and its split between employers and non-employers: Business population estimates 2023.

Use it as the denominator. If a region holds a certain share of UK businesses in your target codes, that share is a ceiling for your realistic prospect count, not a sales forecast.

Three practical uses stand out. It tells you whether a region is under-served relative to its business count. It gives a defensible figure when you pitch a regional campaign to a client. And it flags where non-employer businesses dominate, which usually means low average order values.

The limits are real. Non-employers include landlords, contractors and hobby businesses with no marketing budget. Treat the estimate as an upper bound and apply your own conversion assumptions separately.

Combining ONS output data with buyer sector research

The Annual Business Survey gives turnover and purchases at broad industry level. That lets you estimate a sector's total spend, then apply a marketing or promotional share to reach a serviceable market.

Work in three layers. Layer one is the count of firms from the ONS business counts. Layer two is average turnover per firm from the survey. Layer three is your assumed share of turnover going to promotional products, which comes from your own sales data, not from official statistics.

Only layers one and two are sourced. Layer three is a judgement, and you should label it as one in any proposal.

This is where market sizing discipline matters. A figure built from a count, an average and a stated assumption survives scrutiny. A round number pulled from a blog does not.

Pair the output data with qualitative research on the sector itself. Trade bodies such as the British Promotional Merchandise Association and the Chartered Institute of Marketing publish material on buyer behaviour. Sector press tells you what a construction firm or a software company is currently spending on.

Compliance research belongs in the same pass. The Advertising Standards Authority governs claims in promotional copy, the Information Commissioner's Office governs the data you hold on prospects, and the Office for Product Safety and Standards covers many of the goods you print. None of that changes the numbers, but it changes what you can say and sell.

Building a target list without overclaiming the data

A target list is a filter stack, not a mailing list. Build it in steps and record the source of every field.

The checklist below is the minimum before you call the list a market.

Before you commit budget to a region, our competitor research explained guide sets out how to check who else is already selling there.

Limits of official statistics for promotional merchandise

Official data tells you how many firms exist and roughly what they turn over. It does not tell you what they buy, from whom, or when.

Coverage lags. Business counts and the Annual Business Survey are published well after the period they describe, so a fast-moving sector can shift before you act.

Classification is imperfect. SIC codes are self-selected and rarely updated, and many firms fit more than one code.

Geography is coarse. Registered addresses are often accountants' offices, especially for smaller companies, so a postcode can mislead you about where the buying decision sits.

Financial detail is thin. Most small companies file abridged accounts, so you cannot read budget from a filing.

Finally, official statistics describe the whole economy, not your niche. They bound the market. They do not define your share of it. Use them to narrow, then use your own pipeline data to decide.

Common questions