Most SMEs do not have an automation problem. They have a prioritisation problem.
The best place to start is often the work nobody wants to do: repetitive, time-consuming tasks that drain attention, create customer friction and make good people feel like workflow engines.
What should an SME automate first?
Look for work that has a repeatable sequence, known inputs, measurable outputs and a clear exception path. Do not begin with the most exciting use case. Begin with the task that quietly costs the most time or morale.
Separate automation from augmentation
Automate defined work that can follow rules and escalate exceptions to a person. Augment higher-value work such as planning, analysis, communication, review and critique. AI can help people do better work without removing human judgement.
A five-question filter
- Is the task repeated often enough to matter?
- Can the steps and decision rules be described clearly?
- Is the source data available and reliable enough?
- Can exceptions be identified and escalated safely?
- Can we measure time saved, quality or customer impact?
Prepare before you automate
Standardise the process first. Remove unnecessary steps. Capture useful data as part of the work. Decide who owns the process, what tolerances apply and when a person must intervene. Automation should make the right work easier to do, not make a broken process faster.
Start with one boring process
Ask your team: Which part of your work would you never choose to do again? If that work disappeared, what would you do instead that would improve outcomes for customers?
The gap between draining work and valuable work is where a good automation programme starts. Begin small, measure the result and use what you learn to choose the next opportunity.
Explore AI adoption and adaptation or assess your readiness.
Expanded for Recalibrate from an article originally published on LinkedIn. View the original article.
Most businesses don’t have an automation problem. They have a prioritisation problem. Where do we start the automation efforts, and what should we target first?
When people talk about “automation”, they often jump straight to shiny use-cases. But the biggest wins usually come from something far less glamorous: targeting the work that quietly drains time, attention, and morale from your team.
You know the stuff. The boring bits. The repetitive, time-consuming tasks. The complex processes that should be straightforward. The friction that customers feel – and staff feel even more – because everyone’s pushing information around to keep the machine moving.
That’s the real target for automation programmes – and you won’t get resistance from your team for automating the boring parts of their work.
Automate vs Augment: two very different jobs for AI
I find it useful to split AI work into two buckets:
Automate Target the work that is boring, repetitive, time-consuming, complex, and friction-causing. These are the tasks where there’s a defined sequence of steps, decisions based on known data (within tolerances), and a clear path to escalate exceptions to a human where that is needed.
Augment Use AI to lift human capability and capacity in higher-value work: planning, analysis, communication, review, critique. This isn’t replacing human judgement or reasoning – it’s strengthening it (or it should be!). Think of it as a smart assistant that helps you think more clearly, act faster, and raise the quality of your outputs.
The trap is automating the “cool” stuff first. The opportunity is automating the boring bits first.
Here’s the shift: automation is becoming the new baseline for SMEs
If a task can be described clearly, repeated reliably, and measured sensibly, then we should assume it’s heading toward automation – even if we don’t automate it today.
That changes how we should run improvement initiatives:
- Standardise and simplify processes first – before even thinking of automation
- Capture the process data as a by-product of the work, not as an additional step
- Make sure you have the data available to make effective process decisions
- Design tolerances and escalation paths if particular cases need a human judgement
- Improve the process continuously using the data
You don’t simply “implement automation”. You need to prepare for it.
This could be the beginning of the end for bureaucracy (and meaningless work)
Bureaucracies have been a fact of life for decades. They thrive when humans are the workflow engine: handoffs, approvals, form-filling, status-chasing, rework, and “please resend that” loops. They survive because SOPs require people to execute each step. Many organisations operate like this – with layers of frustrating work for your teams, and friction for your customers (internal or external).
But if SOPs can be executed by systems (Agents) – with clear rules, audit trails, and human escalation for exceptions – then a huge chunk of bureaucracy becomes optional. That’s not just a cost saving, it helps to create more meaningful work.
Because the more time we take out of admin and paperwork, the more we release people to do the work that actually matters: solving real customer problems, improving services, building relationships, supporting team members, and handling the complex cases where humans genuinely add value.
This is the quiet revolution enabled by automation: less busywork, more value.
A practical challenge
Ask your team two questions:
- What are the boring bits of your work which you never want to do again?
- If we removed them, what work would you rather be doing that would genuinely improve outcomes for customers?
That gap – between what frustrates and drains energy and what creates value – is where the best automation programme starts.
Originally published on LinkedIn. Source: https://www.linkedin.com/pulse/automate-boring-bits-hamish-duff-nlxoe
Most SMEs do not have an automation problem. They have a prioritisation problem.
The best place to start is often the work nobody wants to do: repetitive, time-consuming tasks that drain attention, create customer friction and make good people feel like workflow engines.
What should an SME automate first?
Look for work that has a repeatable sequence, known inputs, measurable outputs and a clear exception path. Do not begin with the most exciting use case. Begin with the task that quietly costs the most time or morale.
Separate automation from augmentation
Automate defined work that can follow rules and escalate exceptions to a person. Augment higher-value work such as planning, analysis, communication, review and critique. AI can help people do better work without removing human judgement.
A five-question filter
Prepare before you automate
Standardise the process first. Remove unnecessary steps. Capture useful data as part of the work. Decide who owns the process, what tolerances apply and when a person must intervene. Automation should make the right work easier to do, not make a broken process faster.
Start with one boring process
Ask your team: Which part of your work would you never choose to do again? If that work disappeared, what would you do instead that would improve outcomes for customers?
The gap between draining work and valuable work is where a good automation programme starts. Begin small, measure the result and use what you learn to choose the next opportunity.
Explore AI adoption and adaptation or assess your readiness.
Expanded for Recalibrate from an article originally published on LinkedIn. View the original article.
Most businesses don’t have an automation problem. They have a prioritisation problem. Where do we start the automation efforts, and what should we target first?
When people talk about “automation”, they often jump straight to shiny use-cases. But the biggest wins usually come from something far less glamorous: targeting the work that quietly drains time, attention, and morale from your team.
You know the stuff. The boring bits. The repetitive, time-consuming tasks. The complex processes that should be straightforward. The friction that customers feel – and staff feel even more – because everyone’s pushing information around to keep the machine moving.
That’s the real target for automation programmes – and you won’t get resistance from your team for automating the boring parts of their work.
Automate vs Augment: two very different jobs for AI
I find it useful to split AI work into two buckets:
Automate Target the work that is boring, repetitive, time-consuming, complex, and friction-causing. These are the tasks where there’s a defined sequence of steps, decisions based on known data (within tolerances), and a clear path to escalate exceptions to a human where that is needed.
Augment Use AI to lift human capability and capacity in higher-value work: planning, analysis, communication, review, critique. This isn’t replacing human judgement or reasoning – it’s strengthening it (or it should be!). Think of it as a smart assistant that helps you think more clearly, act faster, and raise the quality of your outputs.
The trap is automating the “cool” stuff first. The opportunity is automating the boring bits first.
Here’s the shift: automation is becoming the new baseline for SMEs
If a task can be described clearly, repeated reliably, and measured sensibly, then we should assume it’s heading toward automation – even if we don’t automate it today.
That changes how we should run improvement initiatives:
You don’t simply “implement automation”. You need to prepare for it.
This could be the beginning of the end for bureaucracy (and meaningless work)
Bureaucracies have been a fact of life for decades. They thrive when humans are the workflow engine: handoffs, approvals, form-filling, status-chasing, rework, and “please resend that” loops. They survive because SOPs require people to execute each step. Many organisations operate like this – with layers of frustrating work for your teams, and friction for your customers (internal or external).
But if SOPs can be executed by systems (Agents) – with clear rules, audit trails, and human escalation for exceptions – then a huge chunk of bureaucracy becomes optional. That’s not just a cost saving, it helps to create more meaningful work.
Because the more time we take out of admin and paperwork, the more we release people to do the work that actually matters: solving real customer problems, improving services, building relationships, supporting team members, and handling the complex cases where humans genuinely add value.
This is the quiet revolution enabled by automation: less busywork, more value.
A practical challenge
Ask your team two questions:
That gap – between what frustrates and drains energy and what creates value – is where the best automation programme starts.
Originally published on LinkedIn. Source: https://www.linkedin.com/pulse/automate-boring-bits-hamish-duff-nlxoe