How to Identify Processes Worth Automating
Knowing that automation exists is not the same as knowing where to start. Here is a practical framework for auditing your workflows, scoring your candidates, and picking the right first process.
Most business owners who decide to explore automation face the same problem within a few minutes: they can think of fifteen processes they could probably automate, and no clear idea of which one to start with. So they either pick something arbitrary, spend months planning, or do nothing.
This article gives you a structured way to answer the question: which specific process in my business should I automate first? It covers how to map your current workflows, a simple scoring framework, the holiday test, and the most common mistakes people make when selecting automation candidates.
If you are still exploring what types of tasks are commonly automated, the previous article in this series covers ten of the most common repetitive workflows worth examining first.
Why a process audit is the right starting point
Without an audit, automation decisions tend to be driven by what feels inefficient rather than what actually is inefficient. These are not always the same thing. The process that is most visible or most frustrating is not necessarily the one consuming the most time or causing the most downstream problems.
A process audit gives you a factual basis for the decision. It surfaces hidden time costs, reveals single-person dependencies you may not have noticed, and creates a prioritised list rather than a gut feeling. If your business already uses a connected business operating system, audit data is easier to gather — you already have visibility into where time is being spent.
A good audit does not need to be exhaustive. Spending one focused day with your team mapping workflows and scoring processes is usually enough to identify your best two or three starting points.
The 5-step process audit
Run through this sequence with your team. You do not need any special software — a spreadsheet and an honest conversation are enough.
List every recurring task
Ask each team member to list every task they do more than once a week. Capture it without filtering — at this stage, everything goes on the list.
Map the steps
For each task on the list, write down the steps in order. Who does what, using which tool, and what happens next? Keep it simple — you are looking for clarity, not perfection.
Score each process
Score each task against the five criteria: frequency, time cost, error rate, single-person dependency, and consistency. Add the scores. Higher scores are better candidates.
Apply the holiday test
For your top-scoring processes, ask: would this break if the person who normally does it was unavailable for two weeks? If yes, it is both a dependency risk and an automation opportunity.
Prioritise and pick one
Choose the process with the highest score that is also consistent enough to describe clearly. Start there. Do not try to automate five things at once.
Example audit flow:
The scoring framework
Score each process against five criteria, using a 1–3 scale for each. Add the scores. Processes scoring 12 or above are strong automation candidates. Processes scoring 8–11 are worth examining. Processes scoring below 8 are usually not worth prioritising now.
Frequency
How often does this process happen?
Time cost
How long does it take each time?
Error rate
How often does it go wrong when done manually?
Single-person dependency
Does this break if one person is absent?
Consistency
Does the process follow the same steps each time?
| Total score | Priority | Recommendation |
|---|---|---|
| 12–15 | High | Strong automation candidate — start here |
| 8–11 | Medium | Worth automating — schedule for next round |
| Below 8 | Low | Deprioritise — may not justify the effort now |
The “if I were on holiday” test
The scoring framework tells you which processes consume the most time and happen most frequently. The holiday test tells you which processes are most fragile — the ones that depend on a specific person being available and remembering to do something.
If you and your key team members were away for two weeks with no access to your phones or laptops, which processes in your business would fail to happen?
Processes that would break — because they live in someone's memory, an informal routine, or a habit that was never documented — are your highest-dependency processes. High dependency is both a risk and a signal that the process is ready to systematise.
Single-person dependency is one of the most underestimated operational risks in small businesses. When that person is on leave, sick, or leaves the company, the process either breaks or gets done inconsistently by someone unfamiliar with it. Automation — and at minimum, documentation — removes that vulnerability. This connects closely to the broader risk of knowledge leaving with key employees.
Common examples
- The weekly report that only one person knows how to compile
- The payment reminder that goes out only when someone remembers to check
- The lead follow-up that relies on a salesperson's personal to-do list
- The client onboarding steps that vary depending on who closed the deal
- The approval that sits in someone's inbox with no visibility for the requester
Tools for mapping your processes
You do not need expensive software to map your workflows. The goal at this stage is clarity, not presentation. The simplest tool that helps you and your team agree on what the steps are is the right tool.
Pen and paper
The fastest way to map a simple process. Good for early-stage audits.
Spreadsheet
Add a row per process step with owner, time, tools used, and error notes.
Sticky notes (physical or virtual)
Useful for team workshops — each step on one note, arranged in sequence.
Simple flowchart tools
Whimsical, Lucidchart, or even a basic drawing tool work well.
Whatever tool you use, aim to capture: the trigger (what starts the process), the steps in order, who is responsible for each step, which tools are used, and what happens when something goes wrong.
Quick wins vs strategic automation
Once you have your scored list, you will likely notice two types of candidates: quick wins and strategic automations. Both have a place, but they are different in nature and should be treated differently.
Quick wins
- Simple, well-defined process
- Takes hours or days to implement
- Immediate, visible time saving
- Builds confidence for the next project
- Example: automating payment reminders
Start here if your team has no prior automation experience.
Strategic automation
- Touches multiple systems or teams
- Takes weeks to plan and implement
- High long-term ROI
- May require custom software or integrations
- Example: end-to-end lead-to-invoice flow
Approach after you have proven value with a quick win.
The typical recommendation for a business that is new to automation is to start with a quick win — even if a strategic automation looks more impressive on paper. A quick win completed is worth far more than a strategic project that takes months and never ships. You can read more about where automation fits versus AI in a separate article.
Common mistakes when selecting automation candidates
These are the patterns that most often lead to automation projects that do not deliver value or do not get finished.
Automating the exception, not the rule
If a process has different outcomes based on complex judgement calls most of the time, it is not ready to automate. Focus on processes where the normal case is clear and predictable.
Automating a broken process
Automation makes processes happen faster and more consistently. If the process is broken, you will get a faster, more consistent broken process. Simplify first, then automate.
Starting with the most impressive, not the most impactful
It is tempting to automate something visible or technically interesting rather than the unglamorous task that consumes the most time. Follow the score, not the novelty.
Skipping the documentation step
If you cannot write down the steps of a process clearly enough that a new employee could follow them, the process is not ready to automate. The documentation exercise itself often reveals gaps you did not know were there.
The clearest signal that a process is ready to automate: you can write the steps down, clearly and completely, without using the phrase “it depends.” If the process is clean enough to document, it is clean enough to automate.
Frequently asked questions
How long does a process audit take?
For a small or medium-sized business, a basic process audit typically takes one to two days. This includes a few hours of interviews with key team members, a review of common workflows, and time to document and score what you find. You do not need expensive consultants or weeks of analysis to get a useful starting point. A simple spreadsheet and honest conversations with your team are usually enough.
Do I need a consultant to identify automation opportunities?
No. Many businesses can identify their best automation candidates using a simple scoring framework and a structured conversation with their own team. A consultant is useful if you want an outside perspective, have a complex operation, or are unsure how to prioritise across many potential candidates. For most small and growing businesses, starting internally and then bringing in specialist help for implementation is the more practical path.
What is the 'holiday test' for automation?
The holiday test asks a simple question: if you and your key team members were away for two weeks with no internet access, which processes in your business would break? The ones that would break — because they depend on someone remembering to do something, tracking something manually, or making a judgement call that only one person makes — are your highest-dependency processes. High dependency often signals a good automation candidate, or at least a process that needs better documentation and systematisation.
What makes a process a bad automation candidate?
A process is a poor automation candidate if it happens rarely, varies significantly from case to case, requires genuine human judgement, or is poorly defined to begin with. Automating a process that has exceptions as often as it has normal cases usually creates more problems than it solves. Before automating anything, the process should be consistent enough that you can describe the steps clearly without 'it depends' appearing more than once or twice.
Should I automate everything at once?
No. Trying to automate too many processes at once is one of the most common reasons automation projects stall. Pick one process — ideally the one scoring highest on your audit — and automate that first. Get it working well, measure what it saves, and let that build confidence and clarity for the next one. A single well-implemented automation often delivers more value faster than an ambitious multi-process overhaul that never ships.
Article series
Business Automation Series
This article is part of a series covering business process automation — from identifying what to automate, to understanding costs, tools, and implementation. New articles will be added as the series grows.
- 10 Repetitive Business Tasks Worth Automating
- How to Identify Processes Worth AutomatingYou are here
Not sure which process to automate first?
We work with businesses to map their workflows, identify their highest-value automation candidates, and build the systems that make it happen — whether that means a workflow automation, an integration, or custom software. The conversation starts with your business, not a technology pitch.
No-obligation conversation
Desi Script helps businesses find and fix their highest-cost workflows.
We've helped businesses across sectors identify the manual processes costing them the most time, and replace them with reliable automations that actually get used.
