Role ambiguity
Who is responsible for the result, and who can intervene when the normal path fails?
AI readiness
AI cannot reliably operate a business it cannot read.
SOPs are only part of the context. Useful automation also needs explicit roles, decision rights, rules, systems, handoffs, ownership, exceptions, and escalation paths. If those are illegible to your operators, they will be illegible to an agent.
The missing context
A clean procedure can still sit inside an illegible operation.
Missing decision rights
Which decisions can the system make, which require review, and who owns the exception?
Tacit rules
What standards and conditions are applied by experienced people but absent from the workflow?
System context
Where does information come from, which record controls, and what downstream work depends on the update?
Hidden exceptions
What makes a case nonstandard, and how is it identified before the wrong action is taken?
Unclear handoffs
When does ownership move, what must be true, and how does the next role know the work is ready?
Model before automating
Make the operating boundaries explicit first.
The goal is not to document everything. It is to give people and machines enough structured context to act safely, escalate correctly, and expose uncertainty.
- 1Model the current operation
- 2Name ownership, rules, and decision boundaries
- 3Identify stable and unstable work
- 4Select automation candidates
- 5Design human review and exception handling
- 6Update the model as the system learns
Focused diagnostic
Already have an SOP?
The AI Readiness Score examines the procedure itself. It is a useful starting point, not a substitute for mapping the operating context around it.