Working document

Is this process expensive enough to fix?

A practical workbook for testing one recurring process before you book a call or buy automation.

We evaluate the process, not the person. Good teams often compensate for broken handovers, duplicate work, and gaps between systems.

Time: 20–30 minutes
Workbook: 8 printable pages
Outcome: Proceed, narrow, or stop

Start with one process. Specific beats comprehensive. ↓

Choose the work

Pick one recurring process.

Choose work where skilled people spend time on steps that do not need their judgment: copying data, chasing status, rebuilding documents, fixing avoidable errors, or maintaining a spreadsheet beside the official system.

Process nameWhere does the process start?Where should it end?Which team or role owns the outcome?Which other teams or external partners touch it?

Make the friction concrete

Name the pain in observable terms.

Describe what people do and what the business experiences. “The system is annoying” is hard to improve; repeated waiting, rework, errors, handovers, and complaints can be measured.

What makes this process painful today?Where is the largest waste of time or money?Which pattern dominates: manual work, errors, waiting, rework, unclear ownership, or complaints?Who feels the problem most strongly?What happens if nothing changes in the next 6–12 months?

Strong signal

“Three people spend about ten hours every week reconciling this, and month-end slips by two days.”

Weak signal

“It would be interesting to automate something here.”

Build a first baseline

Estimate what the process costs each year.

This is a decision estimate, not an audited calculation. Use realistic role-level figures and include repeated licences, external support, and avoidable rework.

Hourly employer costgross monthly salary × 1.25 ÷ 150

Monthly labour costpeople × process hours × hourly cost

Annual process cost(labour + additional costs) × 12

Role	People	Hours / month	Gross € / month	Monthly cost
				—
				—
				—
Additional monthly costs

Estimated monthly total—

Estimated annual total—

Triage guide · rule of thumb

Annual process cost	Usually means
Below €10k	Often too small, unless the fix is extremely simple
€10k–€30k	Possible when complexity is low
€30k–€100k	Good Process Discovery candidate
€100k+	Strong candidate when stakeholders and IT are manageable

Test delivery friction

Run a quick complexity check.

A valuable process can still be the wrong first project. Select one level per row. The score is directional: it helps expose where discovery needs more care.

Signal	Low	Medium	High
Departments	1	2–3	4+
Systems	1–2	3–5	6+ / legacy
Data sensitivity	Low	Customer / finance	Employee / legal / regulated
Exceptions	Few	Some	Many / unclear
IT dependency	Low	Moderate	Formal approval required

Complexity signal—

Which dependency or exception could surprise the project?

Protect what matters

Decide what should not be automated.

The strongest automation boundary is designed before implementation. Keep judgment, accountability, sensitive communication, and unusual exceptions visible to a person.

Where is human judgment genuinely important?What already works well and should remain intact?What change would create justified internal resistance?Which actions or decisions always need an explicit approval?

Use the evidence

Make a provisional decision.

The goal is not to justify automation. The goal is to choose the responsible next step based on value, ownership, and delivery risk.

Turn the worksheet into action

Choose the smallest credible next step.

Keep the completed workbook. It is useful whether you continue internally, speak with AutoMates, or decide not to automate this process.

1

Discovery fit call

A free 30–45 minute conversation to test fit, rough value, and obvious constraints. It is not the full process assessment.

2

Paid Process Discovery

A fixed-scope decision document: current process, cost baseline, constraints, opportunity, and proceed / narrow / stop recommendation.

3

Implementation

Only when the case holds up. Scope, controls, expected savings, and the definition of done are agreed before the build starts.

Send the completed worksheet

Submitting creates a clear completion signal and lets AutoMates connect your answers to the contact you entered. You can also keep everything only in this browser.

NameWork emailCompany