What is a business glossary and why does nobody maintain it?

A business glossary is a company's shared collection of business term definitions: what exactly "order fulfilled", "deal margin" or "churn" mean in the language of the people who work with those concepts every day, not in the technical schema of a table. It is a linguistic artifact, not a data inventory: it says what words mean, while a data catalog says where the data those words describe actually lives. The difference looks subtle, but it is not: two teams can look at the same table and mean two different things by "fulfilled," and no technical catalog resolves that disagreement, because the problem is not finding the data, it is agreeing on what it means. The glossary almost always starts inside a data governance project or a business intelligence initiative, with the stated goal of getting finance, marketing and product to speak the same language before each one builds its own metric.

Why the glossary dies after a few months

The problem is almost never creation, which usually goes fine in a well-facilitated workshop, but maintenance. The people who need to maintain it are data stewards and business owners, who use that term and bear the consequences when it is wrong, not the IT team that built the tool: IT can provide the platform, not the authority to decide whether an "order fulfilled" is counted at shipment or at invoice. Without a named owner and a review trigger, the glossary stalls at launch day: a reorganization changes teams, a regulation introduces a concept, an ERP change moves the threshold that flags that state without the entry getting updated. Updating it feels like extra work with no visible payoff: the cost of a stale definition is diffuse and delayed, writing it costs immediately.

A living glossary versus a dead list

A living glossary has a periodic review with a real deadline, an owner per term who approves or corrects the entry each cycle, and it lives inside the tools people already use: a tooltip in a BI dashboard, a reference from a single source of truth for metrics, not an isolated document nobody reopens after launch week. A dead list is the opposite: hundreds of entries, often contradicting each other, that nobody believes anymore. The question that exposes the difference is not "does a glossary exist" but "who signed off on the last review". A formal glossary is not always the right call: in a small organization where everyone already uses the same words, a disciplined shared spreadsheet is enough, and a dedicated tool with no owner willing to maintain it is just an abandoned license fee.

An enterprise example

A manufacturing company had two definitions of "order fulfilled" in use at once: logistics counted it at shipment, finance counted it at invoice date, with a gap of several days between the two. Each fed a different report to the same supply-chain committee, which received two different fulfillment rates for the same period without knowing it. The fix was not technical: naming a business owner for that definition, having the committee formally approve it once, and linking it to the data catalog so every report pulled it from the same source. The glossary alone would not have changed anything: the accountable owner did.

Why it matters for decision-makers

A business glossary is cheap to start and expensive to let die, because its usefulness is proportional to how often it gets updated, not to how many terms it holds. Whoever funds a governance initiative should evaluate who is accountable for reviewing entries and on what cadence, not the number of terms defined at launch: without that answer the glossary looks finished and was never really started. The real return is not having a document, it is preventing two functions from deciding differently while believing they are talking about the same metric.