Recruiters resist structured interviewing because it sounds like reading from a script to someone whose career you are handling. The research says something more interesting: the structure is what creates room for the conversation.
There is a fair objection to structured interviewing, and it is worth stating properly before dismantling it: a person's career is not a form, and treating a conversation as a data-collection exercise is how recruitment earns its reputation for coldness.
The objection is fair. The conclusion drawn from it is wrong, and unusually for this industry, we have decades of evidence about why.
The most cited work here is Schmidt and Hunter's 1998 meta-analysis, which pulled together eighty-five years of research on selection methods. It put the predictive validity of a structured interview at 0.51, against 0.38 for an unstructured one.1
More recent work by Sackett and colleagues revised those figures downward, and in doing so made the case for structure considerably stronger: structured interviews at 0.42, unstructured at 0.19.2 On those numbers the free-flowing conversation predicts performance barely better than chance.
The gap is not that structure is slightly better. It is that the unstructured interview measures something, and that something is mostly rapport.
This is the uncomfortable part. An unstructured interview reliably identifies people who are good at unstructured interviews, which correlates with confidence, similarity to the interviewer, and comfort in exactly the setting most candidates find least natural. It is not that it measures nothing. It is that it measures the wrong thing, consistently.
Here is what the objection misses. In an unstructured interview, the interviewer is doing two jobs at once: deciding what to ask next, and listening to the answer. The first job consumes most of the attention, which is why interviewers so often ask a question the candidate has already answered.
Fix the questions in advance and one of those jobs disappears. What is left is listening. The best structured interviews feel more conversational than unstructured ones, because the interviewer is not visibly composing their next move while the candidate is speaking.
It is a smaller change than people fear. Structure means the same core questions in the same order for every candidate on a role, scored against agreed criteria, with notes taken at the time rather than reconstructed afterwards.
It does not mean reading verbatim, refusing to follow an interesting thread, or declining to answer the candidate's questions. Follow-ups are fine. The rule is that everyone gets the same starting points, so that when you compare two people you are comparing their answers rather than your memory of two different conversations.
There is a second finding worth knowing: structured processes reduce the influence of bias, because they narrow the space in which impressions substitute for evidence. For a South African desk that is not only an ethical argument, it is a defensibility one. If a client asks why one person progressed and another did not, a scored answer to a shared question is a record. A recollection that someone "seemed like a better fit" is not.
Structure is not the opposite of treating people as people. It is what stops the process quietly favouring the candidates who most resemble the person interviewing them.
The same questions and the same scoring on every candidate, captured while you talk rather than typed up afterwards.