Ordering prototype parts well is mostly about being explicit on the few things that matter and deliberately vague on everything else. Suppliers quote conservatively against uncertainty, so an order that pins down the wrong things and leaves the right ones open costs more and takes longer than it needs to.
What to be explicit about
What the part has to demonstrate, because it changes how a good supplier makes it. The material or, better, what the material has to survive. The dimensions that actually matter and their tolerances. The quantity, now and likely later. And the date the parts are needed, which is different from the date you would like them.
What to leave open
The exact process, if several would answer the question. The stock size and grade within an acceptable range, because the shop's rack decides lead time. The finish on faces that do no work. And the orientation on printed parts unless a property depends on it. Each of these lets the supplier route your job through whatever is fastest and cheapest this week.
What moves the price most
Quantity, because setup dominates and the second part is far cheaper than the first. Lead time, since the fast slot is often a multiple of the standard one. Finishing, which is where identical geometries diverge most. And tolerances, which cost nothing until they cross the threshold the process holds naturally and then cost a great deal.
Ordering habits that waste money
Ordering one part and then needing a second, which pays the setup twice. Choosing the fastest lead time when the review is a fortnight away. Specifying a cosmetic finish on a fit check. Applying a tight general tolerance across a whole drawing. And sending a model with no drawing and then being surprised at what came back.
Questions people ask about prototype parts
How many prototype parts should I order?
More than one. The second and third are far cheaper once the setup exists, and having spares saves a whole lead time when one is destroyed in testing.
Should I specify the process?
Only if it matters. Describing what the part must demonstrate and letting the supplier choose usually produces a faster, cheaper and better matched result.
What information does a supplier actually need?
What the part must do, the material or its requirements, the dimensions that matter with tolerances, the quantity, and the real required date.
Why is expedited delivery so expensive?
Because it is scheduling rather than cost. If the decision the part supports is two weeks away, the standard lead time is free money.