Cross-Border Company Data: How to Source and Compare It

Cross-border company data is information about businesses in more than one country, held in a form you can compare: legal identity, ownership, financials, activity, risk and the people who run them. The hard part is not finding records. It is knowing how complete, current and comparable they are from one country to the next. Milda.ai covers over 420 million companies and 50 million verified contacts in one platform, so a team can research several markets without stitching together a separate source for each.
TL;DR
- Worldwide company information is only as good as its weakest country. Coverage, reporting rules and update speed differ widely between markets.
- Record counts alone mislead. Ask what share of real companies a database covers in the countries you care about, and which kinds of firms are missing.
- Official research shows commercial global databases skew toward larger, older firms and lag behind filings. Plan for that in any analysis.
- Compare databases on coverage, freshness, entity matching, depth (ownership, financials, risk, signals), contact quality and privacy compliance.
- For contact data, the lawful basis and the local rules matter as much as accuracy.
Table of Contents
- What cross-border company data actually includes
- Why international company data is harder than domestic data
- What the research says about coverage and bias
- How to compare global company databases
- How different teams use global company data
- Handling privacy when data crosses borders
- A practical routine for a multi-country research project
- Where Milda.ai fits
- Our perspective
- FAQ
- Sources
What cross-border company data actually includes
A useful international company record has several layers, and most buyers need more than one.
- Identity: legal name, registration number, address, legal form, status (active, dissolved), and a way to match the entity across sources.
- Firmographics: industry, size, headcount range, revenue band, location and founding year. Our guide to firmographic data explains why these fields are a first filter and not the whole picture.
- Ownership and structure: shareholders, parents, subsidiaries and ultimate beneficial owners.
- Financials: filed accounts where a country requires public filing, estimates where it does not.
- Risk and compliance signals: status changes, adverse events, sanctions exposure.
- Activity and growth signals: trade flows, patents, publications, hiring, funding, expansion.
- People: decision-makers and verified business contacts.
The same company can appear in different forms in different countries. A holding company, a local operating subsidiary and a brand name can all be separate records. Good cross-border data resolves them into one structure. Our primer on company master data covers how that resolution works and why it decides whether a database is usable.
Why international company data is harder than domestic data
In the United States you can lean on a fairly standard set of filings. Elsewhere, what exists depends on local law. Each country decides which companies must register, which must publish accounts, and how quickly. The Library of Congress research guide on non-U.S. companies warns that the amount of information and the ease of access vary from country to country, and that researchers often need to consult multiple sources.
There is also no single global identifier. A Bank of Spain occasional paper on identifying all companies worldwide notes that public administrative registries for identifying corporations are not standard practice in many countries. That limits what a single global code, such as the legal entity identifier, can achieve on its own. Matching records across sources is real work.
What the research says about coverage and bias
The most useful thing a buyer can learn before choosing a database is how its coverage is shaped. Three findings from independent research are worth knowing.
Coverage is a sample, and the sample is skewed
A 2020 OECD working paper on a widely used commercial global company database found that coverage ranges from over 50% of firms in the best-covered countries to very few observations in others. The covered firms are disproportionately larger, older and more productive, even within size classes. The authors conclude that such data suits studies of top performers and multinationals better than comparisons of whole national economies.
For a corporate development or investment team, that tells you something practical. If your targets are mid-sized or fast-growing private companies, you need to check how many of them a database holds in each target country, not trust a global headline number.
Freshness lags behind reality
An NBER working paper on using a major commercial company database for research documents a reporting lag of roughly two years: a 2010 filing appeared fully only in 2012 editions. It also found that companies that stop reporting can be erased even if still operating, and that download caps can create missing data. The authors stress that these biases need documented handling before analysis.
The lesson is not that databases are unreliable. It is that financial fields built from statutory filings are backward-looking by design. Forward-looking questions, such as which companies are expanding right now, need event-driven signals on top of filings.
Different databases can lead to different conclusions
A peer-reviewed study of the “database effect” in international business research compared several commercial sources. It found that coverage claims differ widely, from tens of thousands to over 90% of world market capitalization. Differences in country coverage and delisting rules can lead to different conclusions, especially for developing countries. How a database defines its sample matters as much as how many records it holds.
The same issue of uneven country coverage and sampling bias is covered from the benchmarking side in this piece on cross-company benchmarking and federated analytics, which is useful if you compare internal metrics against peers.
How to compare global company databases
Compare options on the criteria below, and test each on your own target list, not on a vendor’s sample.
| Criterion | What to ask | Why it matters |
|---|---|---|
| Country coverage | What share of registered companies is held in each country you target? | Coverage varies sharply by country, so a global total hides gaps. |
| Company size and type | Does it include small, private and young companies, or mostly large filers? | Research shows skew toward larger, older firms. |
| Freshness | How often is each field refreshed, and from what source? | Filing lags can reach years. |
| Entity resolution | Does it link subsidiaries, brands and parents into one structure? | Prevents double counting and missed ownership links. |
| Depth | Ownership, financials, trade, patents, risk, growth signals | Decides whether one tool answers the whole question. |
| Contacts | Are decision-makers verified, and how? | Bounced or stale contacts waste outreach. |
| Access | Web app, browser extension, API | Determines whether data reaches the workflow where you need it. |
| Compliance | Is data source-verified and GDPR-compliant? | Required if you contact people in Europe. |
A practical test: take 50 companies you already know well, spread across at least four countries and three size bands. Check how many are found, how many are correctly structured, and how current their key fields are. That sample tells you more than any feature list.
How different teams use global company data
The same dataset serves different jobs, and each job stresses different fields.
Market analysts map competitors and size markets across countries. They need consistent industry classification and broad coverage of private firms, so that a market is not understated in countries where reporting is thin. Our market research use case shows this workflow.
Investment firms source targets and run early diligence. They need ownership, shareholder links, financial indicators and growth signals, especially for private companies. See our overview of private market intelligence.
Business development and sales teams need the right companies and the right people. Matching a company to its decision-makers, then confirming the contact with a second source, is the routine we describe in finding decision-makers for B2B sales. The sales and GTM research use case covers the workflow end to end.
Innovation and scouting teams look for partners with specific technology. Patents, publications and technology portfolios say more about capability than an industry code does.
Government competitiveness teams need a view of an industry or a country’s firms, including how they connect to foreign owners and trade partners. Here, coverage of smaller domestic firms matters most, which is exactly where commercial samples tend to be thin.

Handling privacy when data crosses borders
Company data is mostly not personal data. Contact data is. If you reach people in Europe, the GDPR applies to how you process their details. Regulators accept that direct marketing can in some cases rest on legitimate interests, but only after a documented balancing test. The UK regulator’s guidance, for example, says you must apply a three-part test and comply with other marketing laws. That guidance applies to the UK; EU member states and other countries have their own rules and enforcement, so confirm the rules for each market you contact with your privacy counsel or data protection officer.
This is why the provenance of contact data matters. Milda.ai describes its data as source-verified and GDPR-compliant, which means your team starts from compliant records instead of building that assurance itself.
A practical routine for a multi-country research project
- Define the entity universe. List the countries, size bands and industries. Write down what “a company” means for your purpose, for example operating companies with more than 10 employees.
- Check coverage per country before you rely on totals. Run your known-company sample. Note where results thin out.
- Resolve structures. Group subsidiaries under their parents so that a group is counted once and ownership chains are visible.
- Layer signals on filings. Use financials for history and growth signals for direction. A two-year filing lag is too slow for timing outreach or deals.
- Cross-check material findings. For a decision that carries money or reputation, confirm key facts with a second source, such as the local registry or the company’s own filings.
- Document the sample. Record the countries, date and filters, so a colleague can reproduce the numbers and so gaps are visible when you present results.
- Deliver to where work happens. Pull data through a web app for research, a browser extension while reading a company website or profile, or an API into your CRM or models.
Where Milda.ai fits
Our data platform gives research, sourcing and sales teams one place to work across more than 420 million companies and 50 million verified contacts. It brings together firmographics, ownership and shareholders, financials, risk assessments, trade data, patents, publications, event-driven growth signals and competitor mapping. You can reach it through search, the browser extension or the API. The data coverage page shows what is included.

Our positioning is deeper coverage from proprietary and multi-source data, which in our own measurements surfaces up to three times more companies than other sources. Pricing adapts to team size and data usage level across Browser Analyst, Strategy Terminal and Data Solutions. See the pricing page for the product lines.
If you want to see a record before you decide, open a sample profile such as Trendex, Device Solutions or Conductive Technologies.

To discuss your countries and use case, start from milda.ai.
Our perspective
Most teams we talk to start by asking how many companies a database holds. The better question is how many of the companies that matter to them it holds, in the countries where they work, and how those companies connect to each other.
We built Milda.ai around that idea. A single source always has blind spots, because every country reports differently. Combining proprietary and multi-source data is how we aim to surface more companies, including the private and smaller firms that filings-based sources tend to miss, and to attach ownership, growth signals and verified contacts to each.
Data should also arrive where decisions are made: in a research workspace, in the browser while you read a company’s site, and through an API into your own systems. And it should be source-verified and GDPR-compliant from the start, so teams spend their time on strategy and relationships rather than on cleaning and justifying records.
Vilius Stanciauskas, Founder and CEO
FAQ
What is cross-border company data?
It is information about companies across multiple countries, organized so you can compare them. It typically covers identity, ownership, financials, industry, risk, growth signals and contacts. Its value depends on how well records from different national sources are matched and kept current.
Is there one global source for worldwide company information?
No official single source exists. Registration and filing rules are set country by country, and a Bank of Spain paper notes that public corporate registries are not standard in many countries. Commercial platforms combine many national sources, so the practical question is how well each platform does that for your target countries.
How current is international company data?
It depends on the field. Financials built from statutory filings can lag by a year or more, and one NBER study found a lag of about two years in a major commercial database. Event-driven signals and contact data can be refreshed faster, so check the refresh rate of each field you rely on.
What should I compare when choosing an international company database?
Compare country-by-country coverage, the share of small and private firms, freshness, entity matching, depth of ownership and financial data, contact verification, delivery options such as API, and privacy compliance. Test with a sample of companies you already know across several countries.
Do I need to worry about GDPR when using global company data?
Yes, if you contact individuals in Europe. Company-level data is generally not personal data, but names, emails and phone numbers are. Use contact data from a source that is verified and GDPR-compliant, and confirm the rules for each market with your data protection officer.
Sources
- Non-U.S. Companies, Library of Congress company research guide
- How to identify all companies worldwide, Bank of Spain
- Coverage and representativeness of Orbis data, OECD
- Using Orbis for research, NBER working paper w21558
- Theoretical and empirical advances in international business research (database effect)
- When can we rely on legitimate interests? ICO
- Cross-company HR benchmarking with federated analytics