CompanyGraph
Thomson Reuters Corporation
TRI · Canada
Runs the legal research database Westlaw, built on a classification system that has organized every major court decision since 1896.
Thomson Reuters runs Westlaw, a legal research database built around the West Key Number System — a taxonomy that lawyer-editors have used since 1896 to classify every Anglo-American court decision into numbered topic categories, linking each new ruling backward through 125 years of related precedent. Because a bar-admitted lawyer-editor must read and classify each new decision before it enters the citation graph, the rate at which new law becomes searchable is capped by how many qualified editors are on staff rather than by any technical limit. Law firms have written Westlaw's specific search commands into associate training programs, corporate legal departments have embedded its citation formats into brief templates, and bar prep courses teach it as standard methodology — so switching to a rival system would mean simultaneously retraining every lawyer, rewriting every template, and rebuilding every workflow at once. The one thing that could unwind all of that is an AI tool capable of running jurisdiction-spanning precedent searches directly from raw court text, bypassing the classification layer entirely, because if law firms accepted that as reliable enough, the 125-year editorial taxonomy would stop being the necessary gateway between a new decision and a usable research result.
How does this company make money?
Law firms pay annual subscription fees priced by how many attorneys the firm has and how heavily they use the service. Corporate legal departments pay per-seat licenses, with extra charges when they retrieve more documents than their plan covers. Law schools pay educational subscriptions priced per student. All three are recurring fees, renewed on a schedule, rather than one-time purchases.
What makes this company hard to replace?
Law firms have built Westlaw's specific search commands and Key Number research methods into the training programs they run for new associates — switching means rewriting those programs and retraining every incoming lawyer. Corporate legal departments have Westlaw citation formats embedded inside brief templates and litigation management systems, which would all need to be rebuilt. Bar exam preparation courses teach Westlaw research methodology as the standard way to do legal research, so even law students arrive already trained on Westlaw's system before they enter the workforce.
What limits this company?
Every new court decision has to be read and classified by a bar-admitted lawyer-editor before it can connect to the existing citation graph. That step cannot be handed to software or split across unskilled workers. So the speed at which new law enters Westlaw is capped by how many qualified lawyer-editors are on staff — not by servers, bandwidth, or any technical limit.
What does this company depend on?
Westlaw depends on court systems filing decisions electronically through PACER and state court databases, because that is how new decisions arrive. It depends on its lawyer-editor staff holding bar admissions to classify those decisions. It depends on the West Key Number System taxonomy itself remaining internally consistent since 1896. It also depends on Westlaw platform infrastructure to deliver content, and on licensing agreements with law publishers and bar associations to include their material.
Who depends on this company?
Large law firms rely on Westlaw to trace how courts in different jurisdictions have ruled on the same legal question — without it, that cross-jurisdiction research would have no systematic path. Corporate legal departments use it to prepare for litigation and monitor compliance with case law; losing it would remove their standardized research capability. Law schools use it to teach students how to conduct legal research; without it, that structured training infrastructure disappears.
How does this company scale?
Giving an additional user access to the existing Westlaw archive costs almost nothing — the content is already there, and one more login does not require more editorial work. What does not scale cheaply is the classification of new content: every new court decision still requires a human lawyer-editor to read it and assign it a West Key Number. So revenue can grow quickly by adding subscribers, but keeping the product current requires growing the editorial workforce at roughly the same rate that courts produce new decisions.
What external forces can significantly affect this company?
Federal courts have been shifting to mandatory electronic filing through CM/ECF, which changes how Westlaw acquires new decisions and could disrupt document intake workflows if those systems change. Legal content licensing agreements with Canadian and UK sources are subject to regulatory changes in those countries, which could affect what international case law Westlaw can include. U.S. law school enrollment trends matter directly because law schools are a subscription customer base — fewer students means weaker demand for educational licenses.
Where is this company structurally vulnerable?
If AI tools trained directly on raw court text — skipping the West Key Number classification layer entirely — became reliable enough that law firms trusted them for jurisdiction-spanning precedent searches, those firms would no longer need Westlaw's editorial taxonomy to find the law. The switching cost that currently locks training programs, brief templates, and litigation workflows to Westlaw would dissolve, because the thing those workflows depend on would no longer be the necessary path to get research done.