Bring the use case. We will tell you if it is real.
The first conversation is a stress test, not a pitch. Most candidate use cases do not survive it, and that is the cheapest possible place to find that out — an hour, rather than a quarter and a steering committee.
Five questions, in this order.
Come with the volume numbers if you have them: how many times a year the trigger fires, and what one item costs across everyone who touches it. A weekly average is the one number nobody knows about their own job.
- 01
What triggers the work
The event that starts it, and how often that event happens in a year.
- 02
What it produces
The artefact at the end, and who has to accept it before it counts as done.
- 03
What one item costs today
Across everyone who touches it — drafting and review, because the review cycles are where the estimate always goes wrong.
- 04
Which GxP flavour applies
Set by the process the system touches, not by the company. It decides the whole validation posture.
- 05
What would have to be true
The data that must exist, the integration that must hold, and the person who owns the output.
You may not need the conversation.
A good deal of what gets asked in a first call is already written down. If your question is about validating a non-deterministic system, what Annex 22 actually says, or what a use case is worth, start with the writing — it is free and it is dated.