A surprising amount of business knowledge never appears in a process manual. It exists in the judgement of colleagues who know which customer needs a different approach, how an unusual invoice should be handled or why an apparently sensible shortcut caused trouble last time.
This informal knowledge is valuable. It also creates a dependency. When the person who holds it is absent, busy or leaves the organisation, the work can become slower and less reliable. The immediate temptation is to document everything. That would be a mistake for most small businesses.
Not every process deserves the same attention
An organisation with twenty people may have hundreds of recurring activities. Creating detailed instructions for every one would consume time better spent serving customers. The appropriate level of documentation depends on the consequences of getting the work wrong, how frequently it changes and how easily another person could take it over.
I would prioritise processes where failure affects revenue, safety, legal obligations, customer confidence or the ability to continue trading. Payroll, order acceptance, customer onboarding, critical supplier changes and collection of payments often justify attention. A low-risk internal preference usually does not.
The important first step is to identify a small number of essential outcomes rather than begin with an organisation-wide documentation programme.
Follow the work, not the department
A process described only from one team's viewpoint may miss the point where problems arise. Sales may believe its responsibility ends with a signed agreement, while delivery requires information that has not been collected. Finance may then have to reconstruct the agreed commercial terms before raising an invoice.
In an illustrative growing consultancy, onboarding requires a sales handover, contract checks, resource allocation, system access and an initial customer meeting. The difficulty is not that any single step is particularly complicated. It is that there is no dependable confirmation that all necessary steps have been completed before work starts.
A one-page process map can make the dependencies visible. Show the trigger, key steps, hand-offs, decision points and definition of completion. Ask the people doing the work to challenge it. If the map describes what managers think happens rather than what actually happens, it has limited value.
Separate instructions from judgement
Good process design recognises that not every situation can be reduced to a checklist. A standard order may follow a consistent path. A disputed contract, unusual customer requirement or suspected fraud may require a professional judgement and an escalation route.
I would distinguish three things: a normal route for routine work, clear checks that protect essential outcomes, and exceptions requiring someone with suitable authority. That is often enough to provide consistency without creating an impractical manual of every possible scenario.
The process owner should also be clear. Someone must be able to approve a change, resolve conflicting practices and check whether the process remains useful. Without ownership, documentation gradually stops reflecting reality.
Knowledge transfer is a practical test
A document can be beautifully formatted and still fail when somebody tries to use it. The better test is whether a competent colleague who does not normally perform the activity can complete a routine example with appropriate support.
This need not become an elaborate audit. A short handover, observation exercise or holiday cover rehearsal can reveal missing access, unclear definitions and unspoken customer expectations. Where mistakes would create material harm, test safely in a controlled environment rather than by exposing a live customer or service to unnecessary risk.
It is worth recording the exceptions encountered during the test. They are often more revealing than the normal sequence because they show which decisions still depend on tacit knowledge.
Technology is not a substitute for clarity
Many businesses turn to a new workflow system or integrated application when a process becomes inconsistent. The technology may help, particularly where a stable process requires a clear record, timely reminders or repeatable checks. However, automating an ambiguous process can distribute the uncertainty more efficiently without resolving it.
The OECD's 2026 SME digitalisation research identifies time, maintenance costs and skills among the barriers to making effective use of digital tools. That research concerns a non-representative international sample, so its findings should not be treated as a precise UK benchmark. The practical implication remains worth considering: processes, capabilities and responsibilities matter alongside the technology. [1]
Before buying a workflow product, decide what good performance looks like, who owns the process and which evidence shows it has worked. A shared checklist may be sufficient initially. If volume, complexity or control requirements increase, more sophisticated tools can be assessed against a defined need.
Keep documentation alive
Processes change when suppliers, services, people or customer commitments change. A document that is not maintained can be more dangerous than an acknowledged gap because it gives a false sense of confidence.
For each critical process, record a named owner, review trigger and last validated date. A material incident, important organisational change or repeated failure should prompt review rather than waiting for an arbitrary annual cycle. The aim is to support the business, not to accumulate a library of untrusted paperwork. For wider policy context, the UK government’s SME plan outlines its approach to supporting small-business growth. [2]
Prioritise knowledge by consequence
A useful knowledge review can begin with three categories. First, activities that stop essential work if a named person is unavailable. Second, activities that continue but become materially slower, more expensive or prone to mistakes. Third, useful preferences or expertise whose absence would be inconvenient but not critical.
This distinction prevents an organisation from treating every undocumented habit as an emergency. The first category may justify immediate cover planning and a controlled handover test. The second may call for a concise guide and a named deputy. The third can often be addressed during normal coaching and team development. The decision should also consider how likely the dependency is to be tested and how much mitigation would cost.
Make the record usable under pressure
Documentation for critical work should answer questions that arise when normal arrangements are disrupted. What triggers the work? What must be checked before action is taken? Where does the required information live? Who has authority to decide an exception? What evidence confirms successful completion?
A complex process might require detailed technical instructions. Many ordinary business activities need less: a short checklist, a contact and escalation point, and links to the records used by the team. The form should follow the need. A document that takes longer to find than the task takes to complete is unlikely to be consulted regularly.
Ownership matters after the first version is created. Someone should be able to notice that a process has changed and update the working guidance. A practical review trigger, such as a recurring error or a change to a critical supplier, is generally more useful than accumulating documents with arbitrary revision dates.
The test remains operational: could a competent colleague, with reasonable preparation and access, carry out the normal activity and recognise when to seek help? That tells the business more than the number of pages written.
Sources and further reading
[1] OECD, Empowering SMEs in the age of AI: 2026 D4SME Survey Read online
[2] UK Department for Business and Trade, Backing your business: plan for SMEs Read online
Sources accessed 6 November 2026. All examples explicitly described as illustrative are hypothetical.
