Why Your Organisation's Knowledge Is Its Most Underused Asset

Ask anyone in a professional firm where to find the planning conditions from a project completed three years ago, and watch what happens. The most likely answer is: "I think Sarah had that — but she left last year. Try the J drive."

This is not a failure of effort or intelligence. It is a structural problem that affects almost every professional organisation: the knowledge the firm has accumulated over years of work is largely invisible to the people who need it most.

The shape of the problem

Institutional knowledge — the decisions made, the precedents set, the hard lessons learned, the procedures refined over time — does not live in one place. It is scattered across shared drives, email threads, project folders, meeting notes, and the minds of individuals who may or may not still work at the organisation.

This matters for three reasons that compound each other.

First, the knowledge is not shared. When a junior colleague needs to know how a similar matter was handled, they cannot easily find out. They either ask around and hope someone remembers, or they start from scratch. Both options are expensive.

Second, the knowledge is not retained when people leave. The departing senior associate, the retiring partner, the project manager who moves firms — they take with them a substantial piece of what the organisation has learned. This is not malicious. It is simply that the knowledge was never properly externalised.

Third, the knowledge that does exist in documents is hard to access. Finding a specific clause in a contract from four years ago, or the exact wording of a planning condition from a comparable project, requires either a well-organised filing system (rare) or someone who remembers where it was (unreliable).

What this actually costs

The cost of not being able to find things is almost always invisible on a balance sheet, but it is very real in practice.

A lawyer who spends two hours searching for a precedent that she suspects exists somewhere is billing that time to a client — or absorbing it. An architect who cannot find the structural specification from a similar scheme has to either redo the research or make do with an inferior starting point. A policy officer who cannot quickly locate the rationale behind a decision made eighteen months ago is exposed when the decision comes under scrutiny.

Multiply these moments across an organisation over a year and the figure becomes significant. Research by various consultancies — though the precise numbers vary — consistently points to knowledge workers spending between 20 and 30 per cent of their working week searching for information. That is one day in five spent looking, not doing.

There is also a less measurable cost: the knowledge that exists but is never applied. A firm that has accumulated twenty years of project records has a competitive advantage in knowing which approaches work, which suppliers are reliable, and which conditions recur in certain types of site. If that knowledge is inaccessible, it is no advantage at all.

Why existing tools have not solved it

Organisations have tried to address this problem in various ways. Document management systems, intranets, wikis, shared drives with careful folder structures — all have been deployed with genuine optimism and have mostly disappointed.

The fundamental problem is that these tools organise documents but do not make knowledge searchable. You can find a file if you know where it is and what it is called. You cannot easily find the answer to "what did we conclude about ground conditions on shallow-piled schemes in flood risk zones?" That question cuts across dozens of documents from multiple projects.

Keyword search helps, but not enough. It returns results, not answers. And it requires you to know which terms to search for — which presupposes a knowledge of what the documents contain that you may not have.

What actually helps

The arrival of large language models — the technology behind tools like ChatGPT — created genuine excitement in professional organisations. Here, finally, was a tool that could read documents and answer questions about them in plain English. The excitement was warranted. The technology is genuinely powerful.

But applying it in a professional context introduces an immediate problem: these tools require you to share your documents with external servers operated by technology companies. For organisations handling client files, commercially sensitive information, or regulated data, this is often not permissible. GDPR restrictions, client confidentiality obligations, and data residency requirements all stand in the way.

The answer is not to avoid the technology. It is to deploy it in a way that keeps your data under your control. The same capability — natural language search across your document archive, with answers drawn from your own knowledge and cited back to their sources — can be provided without your files ever leaving your environment.

The practical implication

For a professional organisation, this means being able to ask a question and get a direct, sourced answer drawn from your own documents. Not a list of search results. Not a generic answer from the internet. An answer that says: "The planning conditions from the Riverside project include the following, per page 8 of the Planning Decision Notice."

That is a qualitatively different experience from any document management system currently in common use. And it makes the accumulated knowledge of an organisation genuinely accessible — to every team member, not just those who were there when the project was completed.