A process can have people available and orders waiting, while a single phase prevents you from completing the job. Adding resources elsewhere is likely to create only more unfinished business. The first step is to identify where waiting times accumulate.
In a hypothetical example, a small production prepares many pieces but has only one location for final control. Increasing preparation does not necessarily increase deliveries. You need to check how long the check takes and what errors force you to repeat it.
I propose to follow the path of some orders, recording processing time and waiting time. The two measures describe different problems. A short activity may remain stationary for a long time because of the lack of information, approvals or necessary materials.
The answer can be organizational rather than financial: improving preparation, moving ancillary activities or clarifying priorities. An additional investment makes sense if it intervenes on the actual constraint. After modification, the bottleneck can move. The verification must therefore cover completed orders, quality and overall time, avoiding rewarding a phase for quantities produced that the rest of the process cannot use.