Every step, every part
The formulas are B1 to B3 on the formulas page.
What interview rounds are for
A hire that does not work out is expensive for the employer and worse for the person hired. Interview rounds exist to make that less likely, and the first ones usually do.
Each round after that has less left to catch, because the earlier rounds already caught most of it. Its costs do not shrink the same way. The seat stays empty for another week, and the candidates who have other offers are the ones most able to stop waiting.
So the useful question is not whether to interview carefully. It is which round stops paying for itself, on your own figures.
What this cannot tell you
- How much of the remaining risk a round really catches. That is your estimate, and it is the figure the answer turns on. The page shows how high it would need to be.
- How fast your candidates receive other offers. It differs by role, by market and by season.
- Rounds that do different jobs. A work sample and a fourth conversation are not the same kind of round, and this treats every round alike.
- Roles where a mistake is costly enough that every round pays. Put that cost in and the page will say so.
- Anything about one person. It weighs a process, not a candidate.
- How many people a step really turns away, or how many it catches. Both are your estimates. Measuring them on a sample is the way to replace the guess.
- Steps that are required by law, by a contract or by safety. Those are not a matter of arithmetic.
- Steps that differ. This treats every step alike, and in practice one step often does most of the catching.
- Whether the people turned away would have been worth what the others are.
- Customers who pay a fee once and do not come back. Put what a returning customer is worth, not one order, to count them.
- What a requirement costs the people who went and met it: the years and the fees behind a degree or a licence are outside this sum.
- Anything about one person. It weighs a process, not the people going through it.