Every automation article tells you to "start with the right process." None of them tells you how to find it. Process analysis for automation is the structured practice of mapping, measuring, and ranking business processes to determine which ones deliver the highest return when automated. Without it, you are investing on instinct — and instinct is expensive.
This article gives you a concrete five-step framework that takes you from a vague "we should automate something" idea to a prioritised shortlist of the three processes that will save the most time and money. No theory lectures — a method you can apply this week.
Why Process Analysis Is Non-Negotiable
Most automation projects that fail do not fail because of bad technology. They fail because the wrong process was automated. A company spends €8,000 automating a reporting task that takes two hours a week, while order processing — twenty hours a week, full of errors — stays untouched.
This pattern stems from three thinking traps:
- Visibility bias: You automate the process you complain about most, not the one that costs the most
- Complexity avoidance: You skip the big, messy process and pick something "easy"
- Technology-first thinking: You start with the tool ("let's use Zapier") instead of the problem
A process analysis takes half a day to two days, depending on the size of your business. That investment pays for itself the moment it steers you away from a low-impact project. Our complete guide to automating business processes covers the full landscape — this article zooms in on step zero: the analysis that precedes everything.
Step 1: Map Your Processes
Before you can prioritise, you need an inventory. This does not mean drawing an org chart or producing a 47-page Visio diagram. It means listing every recurring activity with four core data points.
Walk through every department or function and ask these questions:
- What do you do repeatedly? (daily, weekly, monthly)
- How much time does it take per instance?
- How many times per week/month does it happen?
- Who performs it? (and is that the best use of their time?)
Record each process in a spreadsheet. Be specific — not "admin work," but "check and approve invoices in accounting software" or "transfer customer details from web form to CRM." The more specific, the better you can score in the next step. For insight into which tools you might eventually use, read our comparison of workflow automation tools.
Expect 15 to 40 processes in a business with 10 to 50 employees. That is normal. Not all of them are suitable for automation — you filter those in the next step.
Step 2: Score Each Process on Five Criteria
Here is where it gets concrete. Rate each process from 1 to 5 on these five criteria:
| Criterion | Score 1 (low) | Score 5 (high) |
|---|---|---|
| Frequency | Monthly or less | Multiple times per day |
| Time spent | Under 30 min/week | Over 10 hours/week |
| Error rate | Rarely goes wrong | Regular mistakes with real impact |
| Rule-based | Many exceptions, grey areas | Clear if/then rules |
| System readiness | Fully manual (paper, phone) | Already digital, in existing software |
The total score per process ranges from 5 to 25. Processes above 18 are direct candidates. Processes below 10 go on the shelf.
Why "rule-based" matters: A process with many exceptions requires custom development or AI to automate. That is not impossible, but it is more expensive and slower to implement. Start with highly rule-based processes — they deliver the fastest results.
Why "system readiness" matters: Processes that already run in software (CRM, accounting package, email) are far easier to automate than processes that live on paper or in someone's head. Connecting a digital process costs €500–€3,000. Digitising a paper process first and then automating it costs multiples of that.
Step 3: Calculate the Potential Savings
Scores alone are not enough — you need to know what each automation is worth in euros. That makes the business case tangible and helps secure internal buy-in.
Use this formula per process:
Annual saving = (hours per week x 48 weeks x hourly cost) x automation percentage
Example: an employee spends 8 hours per week on order processing. Hourly cost (including employer contributions): €45. Automation can handle 70% of the work.
8 x 48 x €45 x 0.70 = €12,096 per year
Compare that to the actual costs of business automation. A no-code workflow runs €500–€3,000 one-off. That means a payback period of under three months. For a detailed return calculation, see our article on calculating AI ROI.
Save 10 hours per week on manual process analysis and bottleneck identification
Step 4: Prioritise with the Impact-Effort Matrix
Now you combine two dimensions: how much does it deliver (impact) and how much work does implementation require (effort)?
| Low effort | High effort | |
|---|---|---|
| High impact | Start here — quick wins that free up hours immediately | Strategic projects — plan these for phase two |
| Low impact | Nice-to-haves — only if you have spare capacity | Avoid — little return, lots of work |
High impact comes directly from your scores and savings calculations. Effort is determined by:
- How many systems need to be connected?
- Is custom software required or will a no-code tool suffice?
- Are there dependencies (modify system A before system B works)?
- Is there resistance from the team that currently runs the process?
Select a maximum of three processes for your first automation round. More leads to fragmentation and delays. One successful project delivers more value than five half-finished ones.
Step 5: Document and Validate
Before you request a quote or purchase a tool, document each selected process in detail:
- Trigger: What starts the process? (customer sends email, order arrives, first of the month)
- Steps: Every action, in sequence, including decision points
- Input: What data goes in? (form, email, invoice, spreadsheet)
- Output: What is the end result? (approved invoice, CRM record, report)
- Exceptions: When does it deviate from standard? How often?
Have this documentation validated by the people who actually execute the process — not the manager who thinks they know how it works. There is always a gap between the "official" process and reality. That gap is precisely where the bottlenecks hide.
With this documentation you can have a focused AI consulting conversation or engage a business automation partner — and get a realistic estimate instead of a rough guess.
Common Mistakes in Process Analysis
Three pitfalls we encounter regularly with small businesses:
1. Starting too broad. You want to map everything, including processes you might automate two years from now. Focus on the top-10 time drains. The rest comes later.
2. Not involving the people who do the work. If the employees who execute the process are not part of the analysis, you miss the workarounds, the exceptions, and the actual time spent. Do not send a survey — sit next to them.
3. Ignoring cost differences. A process that takes 20 hours a week but is done by an intern has a different business case than a process that takes 5 hours a week but occupies a senior employee. Always calculate with full loaded hourly costs.
Getting Started: Your First Process Analysis
You now have a concrete five-step plan. The investment is modest — half a day of conversations, an afternoon of scoring and prioritising. The result is a grounded automation strategy instead of a guess.
Start today with step 1: open a spreadsheet and note the first ten processes that come to mind. Ask your team to add five more. Score them, calculate the savings, and you will know exactly where your first automation project delivers the highest return.
Need help running a process analysis? Or have you already completed the analysis and are looking for a partner to implement?
Learn more about business automation?
View service