Image

Finding company information is becoming easier. Turning it into reliable company intelligence remains difficult. 

  • EMIS
  • 7 min read

By Artur Swieciak, Head of Company Database, EMIS

AI has made it easier to discover and process company information. Yet for privately held businesses, particularly in emerging markets, finding information is often only the beginning. The harder challenge is establishing which records belong to the same company, how they connect, whether the information is reliable and current, and whether it can be meaningfully compared across businesses and markets. In many jurisdictions, access to the underlying information also remains limited.

I lead the team that builds, validates and maintains our company database across emerging markets. Most of the companies we cover are privately held, including family-owned businesses, mid-sized manufacturers, regional distributors and fast-growing start-ups. Our coverage also includes state-linked enterprises and major corporate groups. Together they account for a substantial share of economic activity and employment, yet they remain among the hardest parts of the economy to assess with confidence from the outside.

For emerging markets, the challenge is not always that information is unavailable. In many cases, it exists but is scattered, inconsistent, local and difficult to access or interpret. In others, limited disclosure requirements mean that important information may not be publicly available at all. And even when information can be found online, turning it into a reliable, complete and comparable company profile requires considerably more work than simply retrieving it. Here is what conventional web research or AI tools often misses, and why it matters.

My team transforms fragmented records into connected, standardized and verifiable company intelligence.

Who actually owns this company?

Ownership is usually the first question and often the hardest to answer. Private companies in emerging markets frequently sit inside layered structures: holding companies, cross-shareholdings between affiliated entities, ownership distributed among family members, nominee shareholders, and entities registered in a different jurisdiction from where the business operates.

A manufacturer in Vietnam may be owned by a Singapore holding company, which is in turn owned by a vehicle in another offshore centre. A business group in Latin America may run dozens of legal entities under a single commercial brand. Each entity may be perfectly visible on its own. The picture only emerges when someone connects them.

Public registries and structured databases can reveal parts of these relationships, but coverage, completeness and reporting dates vary significantly across jurisdictions. Moreover, direct shareholders, ultimate beneficial owners and corporate group relationships are not necessarily the same. A company may belong to a consolidated group without being wholly owned by its parent, while its ultimate controlling interests may be difficult to establish from publicly available records.

Company filings: available does not always mean accessible

In many emerging markets, private companies do file information: registration details, annual accounts, changes in directors or capital. But the rules on what must be filed, when, and how publicly it is shared vary enormously from one country to the next, and sometimes from one region or company type to another within the same country.

Some countries provide extensive digital access to company registration data and financial statements. Others provide only basic incorporation details. Some charge per document, sit behind local-language interfaces, or are accessible only in person. Tax, customs, procurement, licensing and court records each hold a different piece of the story, and they are rarely linked.

Even where substantial information is publicly accessible, retrieving it is only the beginning. Financial statements may follow different accounting standards, cover different reporting periods or represent consolidated rather than individual companies. Registration and activity classifications may differ across jurisdictions. Bringing these records together requires standardization, validation and an understanding of local reporting conventions.

One company, many names

A single private company can appear under multiple names across official registries, corporate websites, financial statements and news sources. Legal names can change following mergers, restructurings or rebranding, while trading names and commercial brands may differ from the registered entity name. Older news coverage, contracts and court records often continue using previous names.

Language adds another layer. A company registered in Chinese, Russian, Thai or Vietnamese can appear in English sources under several transliterations. Legal suffixes differ by country, abbreviations are inconsistent, and generic names are common. Two unrelated firms can share almost identical names, while the same firm can look like three different ones.

For anyone doing due diligence, credit assessment or supplier screening, this is not a cosmetic issue. Merging two different companies, or missing that a new name belongs to a business with a troubled history, can lead directly to the wrong decision.

Reliable entity resolution still depends on consistent identifiers, supporting evidence and validation. Registration numbers, addresses, directors and historical records must be assessed together, particularly when identifiers are missing or information conflicts across sources.

The best information is often local

The most useful signals about a private company are frequently published close to home: a regional business newspaper, a trade association bulletin, a local ratings agency, a provincial government notice, an industry directory. These sources are written in the local language, for a local audience, and assume local context.

Many are available online, but they can be overlooked by conventional research, particularly when searches focus on English-language sources or internationally recognized publications. A plant expansion reported in a regional Brazilian daily, or a licence award announced by a ministry in Indonesia, may never appear in English at all.

AI has made translating and extracting information from local-language sources considerably easier. But interpreting that information correctly still requires understanding which sources are credible. This is where the combination of AI and human expertise creates real value. Our researchers are based across emerging markets and leverage deep local knowledge and long-term source relationships to ensure the quality and reliability of EMIS company intelligence.

Why this matters more in the age of AI

AI research tools have made it much faster to discover, extract and analyze company information. They can process financial statements, interpret local-language documents, identify potential relationships between entities and combine information from multiple sources. Increasingly, they can also work with structured databases, licensed information and internal company records rather than relying exclusively on the open web.

These capabilities are transforming company research, but they do not eliminate the underlying data challenges. AI cannot reliably recover information that was never disclosed, and it can still struggle with incomplete ownership structures, outdated records, conflicting financial figures or companies with similar names.

What changes is how those gaps show up. A traditional search that finds no clear answer often makes the uncertainty visible. An AI tool may instead produce a fluent, confident profile stitched together from partial, outdated or mismatched sources, attributing a former parent as the current owner, or blending two companies with similar names into one.

For private companies in emerging markets, reliable AI research therefore depends on both capable technology and trustworthy underlying information. High-quality, structured and verified data provides the foundation for an more reliable AI-driven analysis.

What good private company data actually requires

At EMIS, building a company database for these markets has taught us that coverage is not only about the number of records. It is about the quality, completeness, consistency and relevance of the information behind each company profile, and how effectively that information connects to everything else that matters.

That means working directly with local registries, credit bureaus, publishers and information partners, many of whose sources are not available on the open web. It means resolving companies to consistent identities across name changes and languages, linking subsidiaries to parents and shareholders, and preserving historical relationships rather than simply overwriting them.

It also means making information comparable. Financial statements need to be interpreted and standardized across different accounting practices, reporting periods and consolidation levels. Company activities need to be classified consistently, and data from different jurisdictions needs to follow common definitions so that users can meaningfully compare businesses across markets.

Technology plays an increasingly important role in this process. Automation and AI can accelerate data collection, document extraction, translation, classification and entity matching at a scale that would be difficult to achieve manually. But local analysts remain essential for validating complex information, resolving ambiguities, understanding market-specific practices and identifying inconsistencies that automated systems may overlook.

None of this is glamorous. Most of it is invisible to the end user, which is exactly the point. When the work is done well, a researcher, credit analyst or AI assistant can start from a structured, connected and well-documented company profile, rather than piecing it together from whatever sources happen to appear in a search.

Finding company information is not the same as turning it into reliable company intelligence. For private companies in emerging markets, that transformation still depends on access to the right sources, consistent data practices, intelligent technology and the people who know how to interpret what the information actually means.

By Artur Swieciak, Head of Company Database, EMIS

Artur Swieciak is Head of Company Database at EMIS, where he leads the development and maintenance of company intelligence across emerging markets, with a focus on data quality, standardization and AI-driven automation. Based in Sofia, he has more than 13 years of experience working with company and financial data, underpinned by a background in accountancy and law.

Learn more about EMIS, leading provider of reliable industry and company intelligence across emerging markets, and explore our recently launched AI assistant ASkISI.