Decide what to delegate.
Delegation is worth it when the value of the time you recover exceeds what the delegation costs, and the arithmetic has to include the friction. Someone new to a task takes longer than you do, and you spend time reviewing and correcting — so you never recover the full number of hours. Accounting for both means the decision rests on the genuine net rather than on the appealing but wrong comparison of your rate against theirs.
Net value of delegating
Delegate hours = task hours x (1 + rework); cost = delegate hours x their rate; your review time = task hours x rework x 0.5; net = (task hours - review) x your value - cost
Because it should reflect what you would do with the time instead. If the recovered hours go into work that generates revenue or unblocks a team, the value is well above your salary rate; if they go into email, it is below.
Thirty to fifty percent for the first few cycles of a genuinely new task, falling to under ten percent once the person is competent. Delegation that looks marginal at first often becomes clearly worthwhile by the fourth or fifth repetition.
Then either the task is not worth delegating at that rate, or you should delegate it permanently rather than occasionally. The rework cost is largely one-off, so a negative first-month figure can still be right over a year.