A guide to using ONS business data for UK-specific buyer personas

What to take away

Why ONS business data beats stereotype-led buyer personas

Most UK persona decks are written from memory. Someone lists job titles seen on LinkedIn, adds a pain point, and calls it research. The result reads well and predicts nothing.

ONS business data changes the starting point. Instead of guessing how many finance directors work in Leeds, you look it up. Instead of assuming every manufacturer is a large employer, you check the size band distribution.

Stereotypes fail in three predictable ways. They over-represent London, because that is where the writer's network sits. They over-represent large employers, because those are the ones with visible marketing. And they freeze a market at the moment of writing.

Official statistics do not solve persona building. They bound it. A persona grounded in counts and employment figures is a hypothesis with receipts, which is a stronger position than a hypothesis with a nice photo.

That framing matters commercially. If your persona claims a target market of thousands of firms and the published count shows hundreds, the problem is visible before you spend on campaigns. For a fuller treatment of the hypothesis idea, see how a buyer persona is a hypothesis with receipts rather than a mood board.

There is also a compliance angle. UK B2B marketing sits under the Advertising Standards Authority's CAP Code, and claims about market size or reach should be capable of substantiation. Citing ONS releases gives you a defensible source when a claim is questioned.

ONS publishes across the whole of the United Kingdom: England, Scotland, Wales and Northern Ireland. That matters when your sales territory crosses borders, because business counts and employment patterns differ by nation, not just by city.

Finally, official data is free, versioned and citable. Every release has a reference number and a publication date. When a colleague challenges your numbers, you can send a link rather than an opinion.

UK business counts and industry data for persona sizing

The core dataset here is the UK business counts release, built from the Inter-Departmental Business Register. It tells you how many businesses exist, split by industry, employment size band and geography.

The Business, industry and trade statistics hub is the entry point. From there you can reach activity, size and location data for the sectors you sell into, including VAT and PAYE registered businesses.

Read it in three passes.

Size bands matter more than most marketers admit. A micro business buys software on a card, often without procurement. A firm with 250 or more employees has budget cycles, security review and a buying committee. Same sector, different persona.

Industry concentration is the second useful signal. If 40 per cent of your target sector sits in three regions, your outbound effort should follow that concentration rather than spread evenly across the UK.

Use counts to sanity-check your total addressable market. If your CRM holds 4,000 accounts in a sector where the published count is 6,000 employers, you are either close to saturation or your filters are wrong. Both are worth knowing.

Counts also expose thin markets. Some niches look busy on LinkedIn because the same people appear repeatedly. The register shows the true employer count, which is usually smaller.

For the account-first approach that pairs with this, see define the account before the person when targeting B2B social selling.

One caution: the register counts businesses with VAT or PAYE registration. Genuine sole traders below the threshold may not appear. Treat the count as a floor, not a ceiling.

Regional employment figures and occupational data for persona detail

Business counts tell you how many employers exist. Employment data tells you who works in them and what they do.

The Employment and labour market collection covers employment, unemployment and economic inactivity across the UK. It is the broadest starting point for workforce context.

The Employment and employee types dataset goes further, with breakdowns by occupation, industry, working pattern and region. This is where persona detail becomes concrete.

Three outputs are worth your time.

Occupation data answers questions that job titles alone cannot. Is your buyer a manager or a professional? Are they employed or self-employed? Do they work full time or part time? Each answer changes how they buy and where they spend attention.

Regional variation is real. Greater London, the North West, the West Midlands, Yorkshire and the Humber, Scotland, Wales and Northern Ireland all show different industry mixes and employment rates. A persona written for London rarely transfers unchanged to Belfast or Swansea.

Pay data helps you pitch. If the occupational median for your buyer's role sits in a band where software spend is typically signed off at manager level, a message aimed at board approval will miss.

Working patterns matter for social selling specifically. Part-time and shift-working occupations spend less predictable time on professional networks, so channel assumptions should shift with the data.

Combine this with researching B2B buyers with account frames so the numbers attach to named organisations rather than floating free.

Using ONS geographies to define a target area

"UK and Ireland" is not a target area. It is a wish. ONS geographies give you boundaries that mean something.

The UK geographies guidance explains the classifications ONS uses, from countries and regions down to local authorities, wards and output areas. It also explains how boundaries change over time.

Start with the largest unit that matches your sales structure.

Boundary changes are the trap. Local authority mergers, ward revisions and parliamentary boundary reviews mean a saved list from three years ago may not match current geographies. Always check the vintage of the geography you are joining to.

Postcodes are a practical bridge. ONS publishes lookups that convert postcodes to a range of geographies, which lets you tag CRM accounts with a consistent region rather than a free-text city field.

Once accounts carry a standard geography, reporting gets easier. You can compare pipeline by region against employer counts by region and see where coverage is thin.

This is also where sizing B2B social selling in England from the bottom up becomes possible, because each region has a countable employer base.

Pick one geography level and hold it. Mixing local authority counts with region-level targets creates numbers that cannot be reconciled, and the persona loses credibility the moment someone checks.

Building a persona from ONS datasets step by step

This is the working method. It takes an afternoon with the datasets open, plus the primary research later.

Step eight is the one people skip. A persona with visible gaps is more useful than one that pretends to be complete, because it tells you what to go and ask.

Keep the working file. When a new release lands, you update one row rather than rewriting the document.

Where ONS data stops and primary research begins

ONS data describes populations. It cannot tell you what a specific buying committee argued about last quarter.

Here is the boundary in practice.

Everything in the right-hand column needs interviews, surveys or sales-call analysis. Budget for it, because a persona built only from published statistics will be accurate about the market and silent about the buyer.

Sampling matters. Six interviews with people who match the ONS-derived profile beat thirty with whoever replied. Use the official data to define who qualifies, then recruit to that definition.

Governance matters too. UK GDPR applies to personal data you collect during interviews and surveys, and the Information Commissioner's Office enforces it. Keep consent records, store responses securely and delete what you no longer need. When you are choosing a survey tool for B2B audience research, check data residency and retention controls before features.

If you sell into regulated sectors, add a compliance check. Financial services buyers answer to the Financial Conduct Authority, and their language reflects that. Health, energy and public sector buyers have their own constraints.

Worked example: an ONS-grounded persona for a UK B2B seller

Suppose you sell workforce scheduling software to logistics and warehousing operators. The buyer is operations, not HR.

Step 1: industry frame. You choose the warehousing and support activities for transportation group rather than the whole transport and storage section, because your product fits warehouse shift patterns.

Step 2: business counts. The published count shows a large employer base nationally, heavily weighted towards micro and small employers, with a smaller group of large operators. You note that the large-employer segment is small in number but concentrated in specific regions.

Step 3: employment figures. Employment data shows the sector employs a substantial workforce, with a high share of full-time roles and a meaningful self-employed minority. Occupation data points to managers and process, plant and machine operatives as the dominant groups.

Step 4: geography. Using ONS regions, you find the largest clusters in the West Midlands, North West England, Yorkshire and the Humber, and around Greater London and the South East. You pick region as your reporting level.

Step 5: persona skeleton.

Step 6: gaps for primary research. Who approves spend above the manager's limit? What triggers a review of scheduling tools? Which incumbent systems are in place? How do they measure success?

Step 7: test. You check the persona against ten accounts in your CRM. Eight fit. Two are single-site operations below the size band, so you create a separate micro-employer variant rather than stretching the main persona.

That is the whole method. Published counts set the boundaries, employment data fills the detail, geography keeps it honest, and interviews close the gaps.

Keeping personas current as ONS releases update

Personas rot quietly. A document built on 2022 figures still looks fine in 2026 until someone checks the source.

Set a review rhythm tied to publication, not to the calendar. Business counts and employment releases update on known schedules, and revisions are normal. When a release lands, check three things: total counts in your industry, the size band split, and the regional distribution.

Most updates change nothing material. Occasionally one shifts a segment enough to change targeting, and that is the value of the review.

Record the release date and reference in the persona document itself. Anyone reading it should be able to see how old the evidence is without asking.

Watch for classification changes. SIC codes and geography boundaries are revised periodically, and a code that described your sector five years ago may now sit in a different group.

Keep a short changelog. One line per update, noting what changed and what you did about it. This takes minutes and prevents the slow drift that turns a grounded persona back into a stereotype.

Finally, pair the refresh with your primary research cycle. If interviews happen annually, align the ONS review to the same point so the document is updated once, coherently, rather than in fragments.

Common questions