Businesses are being told to automate more — use AI, connect the tools, remove manual work and become more efficient. But for many owner-managed SMEs, there is a better rule to follow: systemise before you automate.
Automation and AI can reduce repetitive work, improve consistency and give people more time for higher-value activity. But before choosing the technology, there is a more important question: is the work itself clear enough to automate?
For many owner-managed SMEs, the issue is not that they have failed to buy the right software. It is that too much of how the business runs still depends on memory, informal conversations, chasing and personal intervention.
Work moves because somebody remembers to move it. Decisions happen because the owner gets involved. Customer handovers rely on what one person remembers telling another. Problems are resolved because an experienced team member knows the workaround.
At a certain level of activity, that can work surprisingly well.
Growth exposes the weakness.
If the process itself is unclear, automation will not create clarity. It may simply make the confusion happen faster.
That is why systemisation should come before automation.
Not because every SME needs thick procedure manuals or layers of bureaucracy, but because growth requires important work to become repeatable.
Before asking a workflow tool, automation or AI agent to speed something up, you first need to understand what should happen, who owns it, what good looks like and where human judgement is still required.
From “what would break?” to “what needs to become repeatable?”
In last week’s article, we asked a simple capacity question:
If sales increased by 20% next month, what would break first?
The question matters because more growth puts pressure on the parts of the business that are already fragile.
Unclear handovers become bigger delays. Decisions that depend on the owner create more interruptions. Knowledge held by one person becomes a bigger risk. Processes that work through goodwill and experience start to struggle under increased volume.
The natural follow-up question is:
What would need to become repeatable before that growth arrived?
That is the real purpose of systemisation.
It is not about documenting everything. It is about identifying the repeated work that affects capacity, cash, customer experience and owner dependency, then making that work clearer and easier to run.
What does “systemising” actually mean?
For an SME, systemisation does not need to mean turning the business into a corporate machine.
In practical terms, it means making important work less dependent on individual memory and interpretation.
A well-systemised area of the business should answer some basic questions:
- What outcome are we trying to achieve?
- Who owns the result?
- What information is needed?
- What does good look like?
- What decisions need to be made?
- What happens when something falls outside the normal pattern?
- When should something genuinely be escalated?
That last question is particularly important in owner-managed businesses.
The owner often becomes the default exception handler.
Customer unhappy? Ask the owner.
Unusual price request? Ask the owner.
Job needs prioritising? Ask the owner.
Team unsure what to do? Ask the owner.
It can feel efficient because the owner knows the business and can usually resolve the problem quickly.
But over time, it creates a growth ceiling.
The business can only move as quickly as decisions and exceptions can pass through one person.
Systemisation begins to remove that dependence.
Five areas worth systemising first
Trying to systemise the entire business at once is usually a mistake.
Start where repetition and friction are already costing you something.
1. Repeated customer handovers
Handovers are one of the most common places where capacity leaks.
Sales hands over to delivery. Delivery hands over to accounts. A new customer moves from onboarding into ongoing service. A project moves from one team member to another.
When those handovers are informal, the same problems keep appearing: missing information, misunderstood expectations, duplicated conversations and promises that delivery did not know had been made.
Before automating a handover, clarify what a good handover actually contains.
What information must be captured? Who is responsible for it? When does the handover happen? What should the receiving person check? What needs escalation?
Once that is clear, a CRM workflow, template or automation can help.
2. Recurring decisions
Some decisions happen frequently enough that starting again from scratch each time creates delay and inconsistency.
Pricing exceptions. Discount requests. Customer complaints. Scheduling priorities. Credit control decisions. Scope changes.
These decisions may still require judgement. That does not mean they cannot be systemised.
Instead of asking, “Can we automate the decision?”, ask:
Can we make the decision easier and more consistent?
That might mean decision principles, thresholds, examples or clear escalation rules.
Good systemisation does not remove judgement where judgement matters. It gives people a better framework for using it.
3. Work that constantly needs chasing
One phrase reveals a lot about how a business runs:
“Where are we with this?”
If owners and managers are asking that repeatedly, there is usually a workflow issue underneath it.
Ownership may be unclear. The next step may depend on someone remembering. Progress may not be visible. Work may be buried in inboxes, messages and informal conversations.
Chasing creates hidden management work.
When the same issues keep returning, it is usually a sign that the business is fixing the incident rather than the underlying way the work operates.
Before automating reminders, ask why the chasing is necessary.
Where does work stall? Who owns the next step? What should make progress visible? What needs to happen automatically because it should not depend on somebody remembering?
Sometimes the answer is technology.
Sometimes it is simply clearer ownership.
4. Work only one person really understands
Most growing businesses have areas where one person has become indispensable.
Sometimes it is the owner. Sometimes it is a long-standing employee who knows how a particular customer, system or process really works.
That creates risk and limits capacity.
A useful question is:
If this person disappeared for two weeks, what would become difficult, delayed or impossible?
You do not need to document every piece of expertise.
But the business should understand enough of the key steps, decisions and information that work does not simply stop.
5. Work that directly affects cash, capacity or customer experience
Not every process deserves the same amount of attention.
A task that happens twice a year and causes little consequence may not need systemising.
A process affecting quoting, onboarding, scheduling, invoicing, credit control, delivery or customer issues probably does.
Those are commercially important workflows.
When they are unclear, the consequences show up as delays, margin leakage, stressed teams, customer frustration and owner intervention.
That is where systemisation creates real value.
A practical sequence for systemising work
Systemisation should start with reality, not theory.
The aim is not to design the perfect process. It is to understand what actually happens now and make it better.
1. Clarify the outcome
Start with the result the workflow exists to create.
A sales-to-delivery handover, for example, does not exist so someone can complete a handover form.
It exists so the delivery team understands what was sold, the customer receives what they expect and nothing important is lost between teams.
Start with the outcome.
2. Map what actually happens now
Look at the real workflow.
Where does it begin? Who touches it? What information gets passed on? Where does work sit waiting? Where are decisions made? Where do people chase? Where does the owner get pulled in?
The formal process and the real process are often very different.
That gap is usually where the friction lives.
3. Remove unnecessary complexity
Before making the process repeatable, simplify it.
Are people entering the same information twice?
Are approvals happening because they have always happened rather than because they are useful?
Are too many people involved?
Does work move backwards because something was not clarified earlier?
Systemisation should remove friction, not add it.
4. Clarify ownership and decision rights
Every important workflow needs clear ownership.
That does not mean one person performs every step. It means somebody is accountable for the result and for making sure the work keeps moving.
People also need to know what they can decide themselves.
If everything still requires owner approval, the process may have been documented, but it has not really become independent.
5. Make the workflow repeatable
Only once the work is understood should you define the simple, repeatable version.
Keep it usable.
The best workflow is not necessarily the most detailed. It is the one people actually follow.
6. Choose the tool last
Only now should you decide whether the business needs:
a checklist, a template, clearer delegation, a workflow tool, simple automation or AI.
The technology should fit the workflow.
Not the other way round.
Not everything should be automated
Systemising something does not automatically mean automating it.
Some work happens too infrequently.
Some requires human judgement, empathy or relationship context.
Some should be simplified rather than automated.
And some should stop altogether.
That distinction matters because businesses can easily become fascinated by what technology can do rather than asking what the business needs.
The goal is not maximum automation.
The goal is a business that runs better.
Where AI can genuinely help
This is where AI becomes interesting.
Once the workflow is clear, AI can support repetitive, information-heavy or coordination-based work.
This is also where organisations are increasingly finding the greatest value from AI: not simply adding tools to existing work, but redesigning workflows, reducing unnecessary handoffs and clarifying where human judgement and accountability still sit.
It might summarise information, classify enquiries, prepare drafts, monitor activity, prompt next steps or assemble information for a decision.
Agentic AI can go further by carrying out parts of a defined workflow rather than simply responding to an individual prompt.
But the same principle still applies.
AI needs clear inputs.
It needs boundaries.
It needs ownership.
And it needs a clear definition of what good looks like.
If the business cannot explain the workflow, it is not ready to automate it intelligently.
AI might prepare a customer handover summary, for example. But the business still needs to know what information matters.
It might monitor overdue actions. But somebody still needs to own the action and decide what happens when it is overdue.
It might draft a credit control message. But the business still needs rules around tone, timing, customer context and escalation.
The useful sequence remains:
Clarify the business problem → understand the workflow → define the judgement → choose the enabler.
That is how AI becomes useful rather than distracting.
The owner’s role is to stop being the workaround
In many owner-managed businesses, the owner is not simply leading the company.
They are also acting as the glue.
They remember what was promised. They know which customer needs special handling. They chase the late actions. They make the awkward decisions. They connect sales, delivery, finance and operations.
That may have been necessary when the business was smaller.
It cannot be the permanent model for growth.
At some point, the owner’s job changes from rescuing work when it gets stuck to designing the work so that it does not need constant rescue.
That is why systemisation is not an administrative exercise.
It is part of building a business that can grow without depending on the owner.
Create capacity, not bureaucracy
Systemisation has a reputation problem.
People hear the word and imagine procedure manuals, complicated flowcharts and endless process documentation.
Growing SMEs do not need that.
They need enough structure to make the business easier to run.
Enough clarity for people to make good decisions.
Enough consistency for customers to receive what was promised.
Enough visibility for work to move without constant chasing.
And enough independence that the owner does not have to intervene every time something falls outside the obvious path.
Before buying another tool or asking AI to fix the problem, ask a simpler question:
What repeated work is still too dependent on memory, chasing, interpretation or the owner?
That is where systemisation should start.
Not everywhere.
Not all at once.
And not for the sake of process.
Start with the work that affects capacity, cash, customer experience and growth. Make it clearer. Make ownership stronger. Make it repeatable.
Then decide whether technology can genuinely make it better.
If you are planning for growth and want to understand where workflow, capacity or owner dependency may become a constraint, a focused review can help identify which parts of the business need to become more repeatable first.
