1
Start
- Observable trigger
- Required information
- Known source
← Back to resources
Automation Readiness Guide
Before You Automate Anything: Map the Process, Handoff, and Exceptions
Before connecting tools or adding AI, explain the real process, write the handoff, identify human decisions and unusual cases, and test whether one repeated task is clear enough to automate safely.
Guided clarity 3 min read
Automation readiness gates
A task must pass every gate before it earns automation
Repetition alone is not readiness. Clarity, ownership, exceptions, human review, and recovery make a contained test safer.2
Decide
- Stable rule
- Human judgment marked
- Expected result
3
Handoff
- Context
- Owner and timing
- Definition of done
4
Recover
- Exceptions
- Error detection
- Correction path
Automate→Clarify first→Keep human
Automation gets exciting quickly: a form can create a task, a task can send an email, and an AI tool can draft a response. But software cannot reliably repair a process the business cannot explain. Start with one real task and describe the manual version. Capture what begins it, what information is needed, where decisions happen, what the next person must receive, and what happens when the normal path does not fit.
Before the tool, understand the work
Automation should begin with the process people actually perform—not the cleaner version everyone wishes they followed.
Watch the real process
Walk through the last three examples rather than describing the ideal process. Record every step, tool, decision, wait, workaround, correction, and side conversation. The untidy details reveal where the process depends on memory or hidden knowledge.
Name what starts it
The starting event must be observable and specific: a complete form arrives, a payment is verified, a due date passes, or a status is approved. “When needed” is not a usable trigger because the software cannot see it consistently.
Identify what must be present
List the information required before the task can begin, what is optional, where each detail comes from, and what should happen when something is missing or conflicting. A reliable automation needs reliable inputs.
Make the handoff safe
The automation is not finished when it sends a notification. The next person still needs enough context and ownership to continue the work without detective work.
Write the five-part handoff
A useful handoff transfers the story of the work—not merely an alert. Write the manual version first so the system has a clear blueprint for what to create, send, assign, or update.
- Context: what happened before this step?
- Owner: who is responsible now?
- Timing: when does it need attention?
- Outcome: what should be true when it is complete?
- Exception: where does it go when the normal rule fails?
Keep judgment visible
Stable rules can often be supported by automation. Fit decisions, complaints, pricing exceptions, legal or financial decisions, access changes, safety concerns, and relationship-sensitive communication should remain visible to a responsible person. The goal is not to remove people; it is to stop wasting their attention on predictable setup work.
Plan for the path that does not go normally
The happy path is only half the process. A trustworthy system also knows when to pause, alert someone, preserve context, and recover.
Map unusual cases before launch
Review recent examples for missing information, duplicate records, unavailable owners, urgent requests, tool outages, conflicting results, and outputs that simply look wrong. Decide which cases can follow a different rule and which must stop for human review.
Design recovery, not just failure
For each likely problem, define how the process detects the issue, who receives the alert, where the original information remains available, how the record is corrected, and how the work safely resumes. A small automation with a clear recovery path is more useful than a clever one nobody can repair.
Readiness gate
A task earns automation by becoming clear and recoverable
Each gate protects the next one. An unclear trigger, weak ownership, unstable rule, or missing exception path changes the responsible outcome.- Clear start Observable trigger →
- Dependable inputs Required information is available →
- Stable work Rules and ownership are understood →
- Safe recovery Exceptions reach a person
Decide what is ready
Repetition makes a task worth examining. It does not automatically make the task safe or useful to automate.
Run the readiness check
Answer each question with yes, partly, or no. Several “partly” answers usually mean the business should clarify the process before connecting tools.
- Is the trigger clear?
- Are required inputs known and usually available?
- Are the rules stable and understandable?
- Is ownership visible?
- Is the expected result specific?
- Are exceptions routed to a person?
- Can errors be detected and corrected?
- Is the information appropriate for the proposed tool?
Choose the right outcome
Automate when the task is clear, low enough risk, measurable, and recoverable. Clarify first when ownership, definitions, information, or exceptions are weak. Keep it human when judgment, sensitivity, or consequences make automation irresponsible.
Test one boring step
Begin with a contained step such as task creation, confirmation, routing, a reminder, a status flag, or a draft for review. Watch real results, correction time, and missed context before adding more complexity. The safest first automation is usually useful, visible, and pleasantly unexciting.
Visual guide
A task is ready for automation when the rules are clear enough to trust.
Your action plan
Run an automation-readiness check
A complete pre-automation worksheet and a decision to automate, clarify first, or keep the task human.
- 1 Choose one repeated task.
- 2 Walk through the last three real examples.
- 3 Write the trigger, required information, steps, decisions, and result.
- 4 Create the five-part handoff.
- 5 Mark what must remain human.
- 6 List unusual cases and recovery steps.
- 7 Score the eight readiness questions.
- 8 Decide: automate, clarify first, or keep human.
- 9 Test one low-risk step and review the results.
Related MethodMade support
Prepare one task for safe automation
MethodMade can document the real process, clarify handoffs and exceptions, evaluate readiness, and design a contained automation that supports human responsibility rather than hiding it.
A natural next step
Keep going when you’re ready.
The guide or story below builds naturally on what you just read.
Related practical guide
The Intake and Follow-Up System Small Businesses Forget to Build
A form collects information; a working system makes the request visible, confirms receipt, assigns ownership, tracks status, prompts follow-up, and recovers when the normal path breaks.
Read guide →Related experience story
Spreadsheet to Operations Platform
A spreadsheet-based workaround had become a recurring operational dependency. I turned the real process underneath it into a searchable internal application that preserved existing work and removed approximately 295 hours of manual entry each month.
Read story →