A senior partner hands in her notice on a Monday morning. She has been with the firm for eight years. By Wednesday, you have realised she is the only person who knows how the firm’s largest retained client likes to receive financial updates, who the actual decision-maker is in their procurement function, and why a standard proposal format will not work for that sector. None of that is written down anywhere. It lives in her head. In three weeks, it leaves.
What is tacit knowledge, and why is it hard to see?
Tacit knowledge is the know-how that people carry without writing it down: how to handle a difficult client conversation, which workaround fixes a recurring issue, what the unwritten rules are for sign-off on a particular type of work. It sits alongside your documented processes without replacing them. Because it leaves no trace on any system, businesses typically only notice it when the person who held it has gone.
A pricing policy, a compliance checklist, or a step-by-step onboarding guide can be written down, filed, and found. Tacit knowledge is everything else.
An account director who reads a client’s mood before a quarterly review and adjusts the agenda accordingly. The operations lead who knows which supplier is genuinely reliable and which one needs chasing, without anyone telling them. Neither consults a manual; the know-how accumulated through experience and sits in them rather than in any document.
For a firm of five to fifty people, this concentration in individuals is the structural risk. When that person leaves, years of context you cannot easily reconstruct leaves with them.
Why does losing it hurt a small firm so hard?
The problem is proportionality. In a large organisation, the loss of one person’s know-how is diluted across hundreds of colleagues who hold related experience. In a ten or twenty-person firm, the same loss can remove a third of the client-facing knowledge the business runs on. The people carrying the most critical knowledge are rarely the ones easiest or cheapest to replace.
ONS data for 2024 recorded around 859,000 UK job-to-job moves in Q4 2023, above pre-pandemic levels. CIPD’s Labour Market Outlook for winter 2023-24 found that 47% of UK employers had hard-to-fill vacancies, with professional and technical roles the most difficult to replace. That combination means you cannot assume you can buy back the knowledge a departing senior person held. You will not find someone with equivalent client relationships and process instincts through a job board in under six months.
The productivity cost of the gap is also measurable. Gartner has estimated that employees spend an average of 2.5 hours a day searching for information or recreating work that already exists somewhere in the organisation. Panopto’s research with YouGov put the productivity loss from inefficient knowledge sharing at around £38 million per year for every 1,000 employees. For a firm where critical processes sit with one or two individuals, the gap between “they are still here” and “they left last month” is felt from the first week.
Where does tacit knowledge actually live in your business?
In a small services firm, tacit knowledge concentrates in four places. Client context: who the real decision-makers are, what history or sensitivities exist, why one client needs handling differently from another. Process workarounds: the informal fixes your team has developed around system limitations or supplier quirks. Risk judgement: the instinctive read on which situations need escalating. And what some practitioners call glue work.
Glue work is the informal coordination that keeps a small team functioning. The person who notices when a client relationship is drifting before it becomes a complaint. The ops lead who knows which supplier to call when the usual one lets you down. None of this appears in a job description. It is invisible until the person doing it is gone.
A useful diagnostic: sit with your senior team for 60 to 90 minutes and map two lists. First, the ten recurring tasks that only one or two people can currently carry out. Second, the ten situations where the business would genuinely struggle if a particular person left tomorrow. That list shows you where the knowledge exposure is concentrated, and where to focus first.
When should you start capturing it, and when is the risk lower?
The urgency depends on three things: how process-dependent your service delivery is, how stable your team has been, and whether any key person is approaching a natural transition. A firm where every engagement is genuinely unique and one-off gets less return from intensive process documentation than one where the same services repeat across clients. The diagnostic is quick: how many of your critical delivery steps currently depend on one person?
Start now if three or more critical processes sit with one person. Start now if any senior member of the team is above 55 or has recently shown interest in other opportunities. And start now if your firm has regulatory obligations that depend on documented processes, because regulators tend to find gaps at exactly the moment when a key person has recently moved on.
The practical starting sequence: search first for what already exists (proposal templates, onboarding checklists, training slides, process emails sent to new starters) and standardise these into a simple playbook structure. Then run structured knowledge interviews, a 45-minute conversation with each key person using consistent questions: walk me through the last time you did this, step by step; where do things commonly go wrong; what do you look for before moving on. Record with proper notice and consent under UK employment data protection rules, and use the transcript to draft the process description. Research on structured debriefs in professional settings has found they improve subsequent task performance by 20 to 25%.
Integrate capture into your joiner, mover, and leaver processes. NCSC guidance on insider risk explicitly recommends these frameworks as a way to prevent critical knowledge leaving the organisation. New starters reviewing processes in their first 90 days often identify gaps more clearly than people embedded in the routine.
The most persistent failure mode combines a promising start with poor maintenance. Firms run a documentation sprint, produce a set of process guides, and let them sit unread until they become inaccurate. Research on knowledge management suggests procedural documents can become unreliable within 12 to 18 months without upkeep. Whatever you build needs a named owner keeping it current.
What else connects to this for a UK services firm?
Knowledge capture connects to three areas that a UK services firm needs to plan for from the start: data protection law, sector regulation, and AI tools. Getting these right at the outset is considerably easier than retrofitting compliance later, particularly if the capture process has already absorbed personal data or client-confidential material that should have been handled more carefully.
Under UK GDPR and the Data Protection Act 2018, any process involving personal data needs a lawful basis and must collect only what is necessary for running the business. The ICO’s employment monitoring guidance makes clear that staff must be told what is being recorded, why, and how the information will be used. If you are recording knowledge interviews or using AI tools to summarise documents containing identifiable personal data, that transparency is a legal requirement.
For FCA-regulated firms, including financial advisers and insurance brokers, the SYSC sourcebook requires that systems and controls are documented well enough that no single role-holder’s departure leaves the firm unable to evidence its compliance.
AI tools can accelerate the capture process. Transcription software turns a 45-minute knowledge interview into a draft process description in minutes. Retrieval tools built on your internal documents let staff query your procedures rather than tracking down the same colleague. The governance constraint is straightforward: do not put client-confidential or identifiable personal data into consumer AI products without a clear data-processing agreement. For many small firms, the business-grade tools already in use carry considerably less risk than consumer alternatives with less transparent data terms.
The starting point for all of this is simpler than it looks. Sit with the people who carry the most critical know-how in your business and ask them to walk you through what they do. Record it. Draft the process from the transcript. File it somewhere the whole team can reach. Three months from now, your business will be able to answer a question it currently cannot.
