Using ONS and Companies House records to research UK customer data platform buyers

What to take away

What ONS datasets reveal about UK customer data platform buyer sectors

The Office for National Statistics publishes the sector backbone you need before you name a single company. Its business data covers counts of businesses by industry, legal form, region and size band. That is the raw material for deciding which sectors can plausibly buy a customer data platform.

For firm counts, start with ONS’s UK Business: Activity, Size and Location tables. The Business Register and Employment Survey (BRES) provides employment statistics, while the Annual Business Survey (ABS) provides financial statistics for businesses. The Inter-Departmental Business Register (IDBR) is an underlying statistical register, not a public list of named prospects; these ONS sources help size and describe a market, not identify firms ready to buy.

A CDP sale usually lands where a firm holds large volumes of named customer records and has a reason to unify them. Retail, financial services, telecoms, utilities, travel and subscription media are candidate sectors to investigate. For example, a retailer joining loyalty and online customer records, or a subscription service bringing billing and engagement records together, may have a use case; neither example alone confirms buying intent.

ONS business data lets you rank those sectors by how many firms actually exist at each size band, rather than by how loudly they market themselves.

ONS tables let you see that spread instead of assuming the whole market sits inside the M25.

Firm size is the filter that changes everything. A CDP implementation needs budget, data volume and someone whose job is to own it. Micro firms rarely qualify. ONS size bands let you draw a line, for example excluding businesses with fewer than ten employees, and defend that line with a published source rather than a hunch.

The same datasets also reveal legal form. Companies, sole traders and partnerships behave differently as buyers. A prospect list built only from limited companies will miss mutuals, cooperatives and public bodies that buy data platforms through different routes.

Start with the ONS Business, industry and trade collection, then open the UK Business: Activity, Size and Location tables for business counts by industry, region and size band. Check the table’s reference period and the ONS release calendar before refreshing your counts.

One caution. ONS data is published at aggregate level. It tells you how many firms exist in a sector, not which ones are ready to buy. That is what the rest of this process is for.

Business, industry and trade data for sector sizing

The wider business, industry and trade collection goes beyond simple counts. It carries turnover, output and structural data that help you estimate how much a sector can spend on data infrastructure.

Use this three-step sizing method:

That third step is where most prospect lists go wrong. A sector with two thousand qualifying firms does not contain two thousand buyers. It contains a smaller number of firms with the data volume, the budget and the internal owner to run a platform. The ONS figures give you the denominator; your own research gives you the numerator.

Trade data adds a useful angle for importers, distributors and manufacturers that sell across borders. A firm with complex cross-border customer records has a stronger case for a unified platform than one selling a single product in a single region.

Sector sizing also protects you from a common trap: building a list around a fashionable vertical because it sounds data-rich, then finding only a handful of firms large enough to buy. The published counts settle that argument quickly.

Use the Business, industry and trade collection to find the relevant ONS tables and publications. Compare sectors using a consistent definition, and record the table and reference period so you can reproduce the counts when the data is refreshed.

Write your sector definition in UK SIC 2007 terms and map each selected code to the closest CRM industry value. Keep the original SIC code in a separate CRM field and record any manual overrides, so your ONS counts can be reconciled with the list.

Companies House filings: accounts, trading status and director records

Companies House is where aggregate sector data becomes named companies. Use its Find and update company information service to search company records, accounts, confirmation statements, director details and PSC information. For repeatable searches, the Companies House API is another named route to company data.

Accounts filings give you firm size in financial terms. Turnover, employee numbers and balance sheet totals appear in filed accounts, subject to the filing regime that applies to the company. Small companies can file abridged or filleted accounts, so some entries show less than you would like.

The filing guidance explains what each company is permitted to submit: Companies House accounts filing guidance.

Trading status is the second signal. An active company with recent filings is a live prospect. A company in liquidation, administration or with overdue accounts is not, or at least not without a very good reason. Confirmation statement dates tell you whether the entity is still engaging with its statutory duties.

Director records add another layer. Directorships show who runs the business, how long they have been in post and whether the same people sit across several related companies. That pattern is common in group structures, where the buying decision may sit at parent level rather than in the trading subsidiary.

Filing history is also a timing signal. A company that has just filed accounts showing revenue growth is more likely to be investing than one that has just filed a decline. You are not predicting the future; you are reading the most recent evidence.

The register’s SIC codes use UK SIC 2007 classifications and let you filter companies by industry. Codes are self-reported and can be stale, so use them as a starting filter, map them to your CRM categories and verify the company’s activity against its own description.

Build a simple screening table before you contact anyone:

That table turns a pile of search results into a defensible list. It also separates records, claims and unknowns so you know which fields you have actually verified.

People with significant control as ownership evidence

PSC records show who owns or controls a company. A person with significant control usually holds more than 25 per cent of shares or voting rights, or otherwise exercises significant influence. The rules are set out at People with significant control.

For prospect research, PSC data answers a question that job titles cannot: who ultimately decides. In a subsidiary, the marketing director may run the process, but the group finance function may hold the budget. PSC records point you towards the controlling entity or individual.

They also expose ownership chains. A trading company may be owned by a holding company, which is owned by an overseas parent. Each hop changes the sales motion, the procurement rules and sometimes the data protection posture.

Overseas entities appear on the register too, often through a registered overseas entity route. That matters for UK GDPR and PECR compliance questions, because data processing arrangements and accountability can sit in a different jurisdiction from the customer base.

PSC data is not always complete. Some companies report a statement instead of a named person, and some entries are out of date. Treat a missing PSC as a research task, not as evidence of anything.

Use PSC records to segment your list by ownership type: independent, private equity backed, group subsidiary, or public sector. Each type buys differently and each has a different approval chain.

Dissolved company search as a screening step

Use Companies House’s Find and update company information service to check whether a company is dissolved before treating it as a live prospect. The register retains records of dissolved entities, so a name in an older dataset may no longer represent a trading business.

Search every prospect in Companies House’s Find and update company information service before it reaches a campaign. Check its status and company number to catch businesses that have ceased trading, names reused by another business, and group entities folded into a parent.

Name reuse is the subtle one. A dissolved company's name can be taken by a new company years later. Without checking the company number, you may attach old research to a completely different business.

Always match on company number, not name. The number is stable; names change with rebrands and ownership shifts. A prospect list keyed on names alone will decay faster than one keyed on registration numbers.

Dissolution also clusters. When a sector goes through consolidation, several mid-sized firms disappear in a short period. If your list was built two years ago and never re-screened, it may be carrying a meaningful share of dead entries.

Make re-screening a scheduled task, not a one-off. Quarterly is enough for most lists, and it keeps bounce rates and wasted outreach down.

Building a prospect list without manufacturing a persona

A prospect list should be a set of verified organisations, not a fictional character with a job title and a pain point. The distinction matters because personas drift into invention, while lists stay anchored to evidence.

The method is straightforward. Define the recruitable population from ONS sector and size data. Pull the named companies from Companies House. Screen them for trading status and ownership. Then, and only then, describe the roles you expect to find inside them.

Use this seven-step sequence:

The final step is where discipline pays. If you write the persona first, you will unconsciously select companies that fit it. If you build the list first, the persona becomes a summary of real evidence.

A checklist for each entry:

This approach also keeps you honest about coverage. You will know which sectors you excluded and why, which is useful when someone asks why the list looks narrow. It is narrow because the evidence is narrow. That is a feature, not a flaw.

For a fuller treatment of the method, see how to research without manufacturing a persona. The principle is the same whether you are working at national or regional level.

Release timing and the ONS release calendar

ONS data is not continuous. It arrives on a published schedule, and your research cycle should follow it rather than fight it.

The ONS release calendar lists upcoming statistical publications with dates. Check it before you plan a quarterly list refresh. If a business population update lands in the same week, you can rebuild counts with fresh figures instead of stale ones.

Timing affects three things. Sector counts shift when new business population estimates are published. Regional breakdowns shift when subnational data is revised. Firm size distributions shift more slowly, but they do move.

There is also a lag to respect. Published statistics describe a period that has already passed. A count published this month may describe the business population several months ago. For fast-moving sectors, treat the ONS figure as a baseline and update it with Companies House filings, which appear closer to real time.

Plan your year around a small number of anchor releases. Two or three refresh points are enough for most B2B research teams. Between them, use Companies House changes as the live feed and ONS as the structural frame.

If you publish any figure externally, cite the release and its date. UK audiences, and the regulators who watch marketing claims, expect numbers to be traceable. The Advertising Standards Authority and the Competition and Markets Authority both take a dim view of unsupported statistics.

Finally, remember the compliance frame around the list itself. UK GDPR and the Data Protection Act 2018 govern how you hold and use business contact data, and PECR governs electronic marketing. The Information Commissioner's Office publishes guidance that applies whether your prospects are sole traders or large companies. Build the list with that in mind from the start.

Common questions