We build async interview software, so treat this with appropriate suspicion. But we would rather you use it where it works than everywhere, because the second one ends with you concluding the whole category is useless.
Very small candidate pools
Three applicants for a specialist role? Call them. The entire benefit here is parallelism and there is nothing to parallelise. You will spend longer writing good questions than you would have spent on three conversations.
Senior and executive hires
At a certain level the first round is mutual, and quite often you are the one being assessed. Asking a head of engineering with four approaches in their inbox to record answers into a laptop reads as though you do not value their time. You may be right that it is efficient. You will still lose them.
Anything that needs a follow-up question
If the interesting part of an answer is the thing you would probe next, a recording cannot help you. Some roles are like that all the way down — consultative sales, incident response, anything where the skill is the conversation itself.
When you have not written the questions properly
A bad question in a live interview is recoverable; the interviewer rephrases and moves on. A bad question in a recorded round is inflicted identically on every candidate, and you find out when forty people have all answered something you did not mean to ask.
The format has no slack in it. That is its strength when the questions are good and its weakness when they are not.
Where it does earn its place
Volume beyond what your calendar can absorb. Distributed candidates and distributed panels. Rounds where you want every applicant assessed on identical questions. Early-career hiring where CVs cannot separate anyone.
If your round is not one of those, we would rather tell you now than sell you a subscription you resent in a month.
