A conference room pilot is a working, end-to-end demonstration of the client's own process, in a real environment, run in front of the people who will use it — early enough to change the design. Every clause in that sentence is load-bearing.
It is not a demo. A demo shows the product; a pilot shows THEIR product being made in THEIR system. That distinction is the whole technique, and it is the difference between a session that changes your design and a session that produces polite nodding.
Why it works, and this is specific to this industry. Manufacturing clients are concrete thinkers. A design document produces polite agreement — people read it, it looks reasonable, they sign it, and they have not really engaged with it because it is describing something they cannot picture. A work order completing on screen with their part number on it produces real feedback, including the objections nobody raised in the workshop. That is not a criticism of the client; it is how people work. Show somebody their own process and they will tell you what is wrong with your version of it in about ninety seconds.
What you need to build first — and this is why the pilot sits here rather than at the end of the course. An organization, a calendar, units, one or two real items, one real structure, resources, one work definition, some stock, and enough cost setup to value it. That is roughly the minimum path, and it is why this sits beside quick setup, which is how you get there in a week rather than a month.
What it surfaces, every item of which you would otherwise find in UAT: process steps nobody documented, the operator's actual sequence versus the engineer's assumed one, costing expectations Finance never stated, data quality problems in the item and BOM extract, and genuine terminology mismatches between plant and system — where you and they have been using the same word for different things for six weeks.
When to run it: after the design workshops, before the full build. Early enough that changing your mind is cheap.
And the commercial argument, because you will have to make it. Sell it as risk reduction, not as a nice-to-have. A nice-to-have gets cut when the plan gets tight; risk reduction gets a line and a number — and the number is a week against the cost of rebuilding a design after it has been configured.