Flow situation after a reduction, or before, inside the valve, is determined by actual flow, pressure, temperature and medium and the valve geometry, plus geometry after the valve. Flow turbulence/loss might be calculated approximaty based on earlier data from experience. Measuring and observing with the actual flow condition is the only sure way.
Other ways: 1. See what competitors have done. 2. Wich constructions are 'best selling' on the market. 3. Test with an actual valve with different adapted reductions mounted. 4. Better yet: Cut off valve ends and mount together with different end-redution pieces.
There is no general rule for a reduction size, but common practises for some different 'typical' applications. If you want to avoid too turbulent situations and cavitation (in or just after the valve) you must be conservative.
You might get some suggetions if you publish more details!
But way bother with large reductions? It will of course inflence and lower the production cost under some conditions, and more if number of valves is large.
PS. Reduction to 80% will in most (not all!)common cases be OK, lower has to be calculated/tested.