Skip to content
The best first automation is rarely the biggest, cleverest or most complicated process in the business.
A better starting point is work that happens frequently, follows a reasonably consistent pattern and consumes time without requiring much human judgement.
In other words: start with the boring stuff.
Look for repeated work
Good automation candidates often hide in ordinary tasks:
- Moving information from a form into another system.
- Creating standard records or documents.
- Sending reminders when something is waiting.
- Notifying someone when a status changes.
- Compiling recurring information for a report.
- Copying the same information between applications.
- Chasing a predictable next step.
None sounds particularly transformational. That is often why they make good starting points.
Repeated small improvements can remove friction without forcing the business to redesign everything at once.
Automate the queue, not the confusion
A frustrating process is not automatically a good process to automate.
If three people complete the same task differently, nobody knows which system contains the correct information or exceptions are dealt with from memory, automation may simply make the inconsistency happen faster.
Before building anything, the business should be able to describe:
- What starts the process?
- Which information does it rely on?
- Who owns it?
- Which rules determine what happens next?
- What are the exceptions?
- What should the finished result look like?
If those answers are unclear, simplify the process first.
Current Microsoft automation guidance follows much the same lifecycle: strategy and planning come before design, implementation, testing and ongoing refinement. Automation is a managed process, not simply the act of building a flow.
Keep people where judgement adds value
Automation does not need to remove the person from the process.
A system might collect information, check that required fields are present, create a record and notify the right colleague—while leaving the important client or commercial decision to them.
AI can extend this further by summarising information, interpreting less structured input or preparing a response. But the principle remains the same: the higher the consequence of a decision, the more carefully the role of human judgement should be considered.
Pick something you can measure
Before automating, establish what happens today.
How often does the task occur? How long does it take? How much waiting or rework does it create? What happens when something is missed?
The benefit may not simply be hours saved. Faster responses, fewer handovers, greater consistency and better visibility can be equally valuable.
A useful question for a small business is:
Can we clearly describe the trigger, information, rules, exceptions, owner and desired outcome of this process?
If not, automation is probably premature.
If we can—and the work happens often enough to matter—we may have found a very good place to start.
The best automation does not make a confused process faster. It makes a clear process easier to repeat.