Illustrative example. Gintaras runs a 14-person logistics services company in Klaipėda. He knows his team spends too much time on manual work: someone moves data between systems every day, someone emails customers by hand, someone spends two hours every Friday afternoon putting together the weekly report. Gintaras sees the problem. But ask him “what should you automate first?” and he stops. “I don't know where to start,” he says. “There are too many processes, and something is off in all of them.”
This is a typical situation. Owners sense that automation could help, but they have no structured way to find out where exactly. The result is either nothing happens (paralysis by analysis) or the wrong process gets automated and the investment never pays back. In this article I give you a concrete framework: how to run a business process inventory in 2–3 hours, 12 questions to shortlist candidates and a priority matrix to choose your first project.
Why automation projects often start with the wrong process
The first automation project is often chosen not by objective criteria but by whatever annoys the owner most right now. That makes sense psychologically, but it is a strategic mistake. An annoying process does not always mean a big time cost or clear automation potential.
Too complex for a first project. Most owners want to automate a complex, multi-step process with many exceptions, such as complaint handling or project coordination. Complex processes need a large upfront investment and a long rollout. A successful first project should be simple, clearly defined and show results quickly.
Too small to pay back. The other extreme is automating a process that takes 15 minutes a week. It is technically possible, but the savings will never cover the investment. Automation pays off where a process repeats often and takes a lot of time.
No clear user. The process gets automated, but the team does not use it or works around it, because nobody involved them in the decision. The best inventory is done together with the people who run the process every day. They know the pain points better than the owner does.
A process inventory framework: 4 columns in Google Sheets
An inventory does not have to be complicated. Start with a simple Google Sheets file with four columns; that is enough to pick your first project:
These four dimensions let you compare processes quickly and see where the time cost is highest. The goal is to collect 20–50 processes from everyone on the team within a day or two. Not a perfect analysis, just enough to decide on the first project.
- 1List your processesGoogle SheetsProcess, owner, frequency, hours/week
- 212 questions for each one4+ points = candidate
- 3Place them on the priority matrixReturn and complexity
- 4Start with one quick win
12 questions to find automation candidates
Once the inventory table is filled in, ask these 12 questions about each process. Every “yes” is one point. Processes with 4+ points are automation candidates; 7+ points are priorities.
- 1. Which process repeats more than 5 times a week? Frequency is the first signal for automation. If a process is rare, the costs will dominate.
- 2. Which process spans several tools? Moving data from one system to another by hand is a classic automation candidate.
- 3. Which task keeps getting put off “for later”? If the team systematically postpones a certain job, the process is frustrating and people are looking for a way around it.
- 4. Which process depends on a single person? A single point of failure: when that person goes on vacation, the process stops. Automation removes the dependency.
- 5. Which process misses deadlines? If certain tasks are late again and again, automation can make sure the steps happen on time.
- 6. Which task takes more than an hour to teach a new hire? Complex manual processes with many steps can usually be standardized and automated.
- 7. Which process produces errors? Errors point to human-factor risk, and automation removes repeated mistakes.
- 8. Which process involves copying data by hand? Copying between systems is prime automation territory. It is boring for people, and the chance of error is high.
- 9. Which process has to run at night or on weekends? If customer service or order processing is needed outside working hours, automation is the cheapest solution.
- 10. Which process directly affects revenue but keeps slipping? If an important task (for example, following up a lead) is regularly missed, that is a direct loss of revenue.
- 11. Which process needs a manager's approval at every step? Approval flows are a common candidate: email notifications, approval tracking, escalation.
- 12. Which process do customers complain about? Complaints about slow replies, wrong information or missed deadlines clearly show where automation would add value for the customer.
How to prioritize automation: ROI vs complexity matrix
Once you have scored the candidates, sort them on two axes: potential return (time saved or revenue impact) and implementation complexity. That gives you four quadrants:
Quick wins (do first)
High return + low complexity. For example, automatic reminder emails, reports from existing data, sending invoices.
Long-term wins (plan)
High return + high complexity. For example, full CRM integration, AI agents, ERP automation. Worth investing in, but after the quick wins.
Postpone
Low return + low complexity. Technically simple, but the return is small. You can automate these later, once the infrastructure is in place.
Reject
Low return + high complexity. Expensive and gives little back. Do not make one of these your first pilot.
Your first project should always be a quick win: small, clearly defined, with a result you can see quickly. A successful first project builds trust in automation among the team and management, which is the emotional groundwork for what comes next. For how to calculate the return, read the article on process automation ROI in 90 days.
Illustrative example: a services company inventory with 47 processes, 3 quick wins, ~€1,200/month
For example, an IT services company with 22 employees would typically complete a process inventory in two days: on day one, interviews with team members and filling in the table; on day two, prioritization together with management.
Typical inventory results. The team might identify, for example, 47 different processes, 31 of them recurring. After applying the 12-question framework, 14 processes would score 4+ points and become automation candidates. Of those, 3 would land in the quick wins quadrant.
Three quick wins:
- Generating weekly reports: 5 hours/week of project managers' time. An automated Make.com workflow collects data from Jira and Google Sheets and sends a summary to management every Monday morning. Estimated savings: ~€480/month.
- Onboarding emails for new clients: 3 hours/week of the sales team's time spent on manual follow-up emails. An automated sequence based on CRM statuses. Estimated savings: ~€360/month.
- Tracking sent invoices: 2 hours/week of the accountant's time monitoring payments and sending reminders. Automatic reminders after 7 and 14 days. Estimated savings: ~€360/month.
- 1Data is collected from work systemsJira · Google Sheets
- 2The summary is built automaticallyMake.com
- 3Every Monday morning, to managementEmail
From the example above: the process takes 5 hours a week of project managers’ time, so it pays back fastest.
Likely overall result. All three quick wins can be automated in 3 weeks. Estimated investment: ~€2,400. Estimated monthly return: ~€1,200. Break-even: about 2 months (an illustrative calculation, not a client case). An inventory also typically surfaces longer-term opportunities, such as full CRM automation or an AI agent for technical support, which are worth planning for a second phase.
Common mistakes in a first process inventory
Even with a structured framework, an inventory can fail. Here are three classic mistakes:
Aiming too high. Owners often want to automate everything at once: all 47 processes in two months. That ends in more stress, sloppy work and disappointment. The goal of the inventory is to pick one first project, not to draw up a two-year roadmap.
Leaving employees out. If only the owner or a manager does the inventory, the biggest pain points will be missed. The people who run the processes every day know where the time is really wasted, and it is not always where the owner thinks. You need at least one 30–45 minute conversation with each team.
Not measuring the baseline. “It takes a lot of time” is not enough. Before automating you need concrete numbers: exactly how many hours a week, how many repetitions, what an employee hour costs. Without this data you cannot calculate ROI or justify the investment to management. Tracking time for at least 15 minutes a week over one week gives you a sufficient basis.
Process inventory template and a 30-minute step-by-step plan
Here is a minimal plan for starting your process inventory today, even if you only have 30 free minutes:
- Min 1–5. Create a Google Sheets table with four columns: Process, Owner, Frequency, Hours/week.
- Min 6–15. Write down everything you do by hand every day or every week, at least 10–15 processes. It does not need to be perfect; what matters is starting.
- Min 16–20. Ask the 12 questions for each process and count the points. Mark the processes with 4+ points.
- Min 21–25. Place the selected candidates on the priority matrix (ROI vs complexity). Identify the quick wins.
- Min 26–30. Send the table to your team and ask them to add to it within the next two days. Schedule a 15-minute meeting to discuss it.
You can get the template for free: write to me and I will send you a ready Google Sheets file with formulas, the question list and the priority matrix. Once it is filled in, you can send it back: I will review it and reply by email with the process that looks best to automate first, and if we need to go through the details, we will arrange a call. More about the automation service: business process automation with Make.com and n8n.
If after the inventory you decide you need a more advanced AI agent (not just workflow automation), see the AI agent development and deployment service.