Illustrative example, not client data
Example company: a 140-person UK manufacturer processing roughly 600 technical quote requests each month. This note demonstrates the shape and candour of the diagnostic output. All names and figures are fictional.
Recommendation
Go, narrowly.
Pilot quote-intake automation for one product family. Do not automate scheduling or supplier negotiation yet.
Quote preparation is the only opportunity with a clear operational owner, sufficient source material, a safe human approval point, and payback inside twelve months under conservative assumptions.
Opportunity 1
Go
Why: A repeatable document workflow with measurable handling time, good historical examples, and a human approval point before any quote is issued.
Next: Run a four-week pilot on one product family. Measure preparation time, extraction accuracy, and exception rate.
Opportunity 2
Wait
Why: The value is plausible, but routing and machine-capacity data are inconsistent. Automating around the current data would hide rather than fix the constraint.
Next: Standardise three scheduling inputs and collect eight weeks of clean exception data before revisiting.
Opportunity 3
No
Why: Low transaction volume, high commercial judgment, and material relationship risk. The likely saving does not justify the controls and review burden.
Next: Keep negotiation human-led. Use existing tools only for meeting summaries and comparison preparation.
Conservative cost model
Payback assumptions
Decision required
Approve a fixed-scope pilot for one product family, with named operational ownership and baseline measurement completed before build starts.
Talk about a diagnostic