Institutional knowledge
Institutional knowledge is the collective know-how of an organisation — facts, decisions and their reasons, client and market context, relationships and ways of working — held in its people, documents and systems, which lets it keep functioning when individuals change.
What counts as institutional knowledge?
It is broader than documentation. Typical examples:
- Facts and records — contracts, product specifications, prices, policies.
- Decisions and their reasons — why a supplier was dropped, why a client was advised against an option, what was tried and failed.
- Context — the history of a client relationship, how a regulator tends to read a rule, which sources in a field can be trusted.
- Know-how — how a process really works, as opposed to how the manual says it works.
- Relationships — who to ask, who knows whom.
Researchers distinguish explicit knowledge, which can be written down, from tacit knowledge, which people use without being able to fully articulate it — Michael Polanyi's observation that "we can know more than we can tell". Nonaka and Takeuchi's SECI model describes how organisations convert between the two through socialisation, externalisation, combination and internalisation.
Why does institutional knowledge matter?
Because it walks out of the door. When an experienced person leaves, retires or moves to another team, the knowledge only they held leaves with them. The risk is often described by the bus factor: how many people would have to disappear before a project stalls. For many small firms and teams, the honest answer is one.
The costs show up in familiar ways: new hires take longer to become productive, clients have to explain their history again, the same research is repeated, and old mistakes are made twice because nobody remembers why a path was abandoned. Conversely, organisations that keep their knowledge reachable can reuse it — one engagement's research informs the next, and a decision can be defended years later because its basis is on record. See consulting knowledge management for how professional-services firms approach this.
How do organisations capture institutional knowledge today?
- Wikis and documentation in tools such as Confluence or Notion — see knowledge base.
- Document management and shared drives, often organised by client or project.
- Process documentation — runbooks, checklists, onboarding guides.
- People-based practices — mentoring, pairing, handover periods, exit interviews, communities of practice.
- Decision logs that record what was decided, by whom and why.
- AI search and assistants that answer questions across the existing documents.
Each works for part of the problem. Documentation captures explicit knowledge; mentoring and pairing are still the main ways tacit knowledge transfers.
Why does institutional knowledge get lost anyway?
- Written once, never updated. Pages go stale, people stop trusting them, and then stop reading them.
- The why is missing. Documents record outcomes, rarely the reasoning and sources behind them.
- Research evaporates. The reading behind a recommendation — articles, reports, regulatory texts — lives in one person's browser tabs and is gone when they leave.
- Findability. Even when the knowledge was written down, nobody knows it exists or which words to search for.
- Documentation is extra work. If capturing knowledge competes with billable or urgent work, it loses.
How can a small team keep its institutional knowledge?
A few practices go a long way, and they work without a dedicated knowledge manager:
- Capture as a side effect of work. Save sources while you research, record decisions when you make them — not in a separate documentation sprint.
- Keep reasons and sources together. A recommendation should link to the evidence it rests on.
- Use one shared place per topic or client, so knowledge does not fragment across personal folders.
- Make retrieval easy: search by meaning, plus an assistant that answers with citations.
- Review what changed. Sources move, regulations are amended; knowledge that was true last year needs a check.
- Pair people for tacit knowledge. No tool replaces working alongside someone experienced.
How does Pith help with institutional knowledge?
Pith addresses one specific slice: the knowledge that comes from reading and research, and the decisions built on it. In a shared workspace, everything members save — articles, PDFs, highlights — feeds one cited wiki, so the research behind the team's work stays with the team when someone leaves. Teammates join without seat fees; the workspace pays per AI action from a prepaid credit.
For client work, clients group the sources of an engagement, and Decisions record what you recommended, for which client, and which sources it rests on. Pith re-checks those sources about weekly and flags one that changed, went offline or was withdrawn. Search, chat and your own assistant over MCP answer questions from the shared memory with citations.
Pith does not replace process documentation, HR onboarding or the mentoring through which tacit knowledge moves. For a team wiki of hand-written documentation, Confluence or Notion remain the natural tools — see Pith vs. Confluence.
Frequently asked questions
What is an example of institutional knowledge?
Knowing why a long-standing client prefers a particular reporting format, which regulator contact responds quickly, or why a vendor was dropped three years ago. None of it is usually in a manual, and all of it matters when someone new takes over.
What is the difference between institutional knowledge and tribal knowledge?
The terms overlap. Tribal knowledge usually means undocumented know-how held by a small group, while institutional knowledge covers everything an organisation knows, documented or not.
How do you prevent loss of institutional knowledge when employees leave?
Start well before the notice period: capture sources and decisions continuously, keep them in shared rather than personal spaces, pair people on critical work, and use handover time for the tacit parts that documents cannot carry.
Can AI preserve institutional knowledge?
AI can make written knowledge far easier to find and summarise, and it can turn saved research into structured pages. It cannot capture what was never written down, and its answers need sources so people can check them.
Is institutional knowledge the same as a knowledge base?
No. A knowledge base is one container for the explicit part of institutional knowledge. Much of what an organisation knows also lives in people, habits and relationships.
See also
Last reviewed: · Licensed CC BY 4.0 · cite freely with attribution to Pith.