What to expect in a mock interview

A mock interview deliberately copies the real one's skeleton, so nothing in it should surprise you, and that is the point. The standard arc has four phases.

• Setup, 1 to 2 minutes. The interviewer confirms the target role and what kind of round you are simulating: behavioral, technical screen, or a mixed panel. With an AI interviewer this is where your resume and the job description get loaded.
• Introduction, 2 to 3 minutes. Almost every session opens with some form of tell me about yourself, because almost every real interview does. Treat it as a scored question, not a warm-up.
• The question block, 15 to 30 minutes. Expect 5 to 8 primary questions drawn from the role, each with 1 or 2 follow-ups that probe whatever you just said. The follow-ups are the realistic part; a session without them is a quiz, not a mock interview.
• Feedback, 5 to 10 minutes. The interviewer walks through what worked and what did not, ideally per answer rather than as one general impression.

Total: 20 to 45 minutes depending on depth. Shorter than most real onsites, and intentionally so, because the value per minute is in the question block and the feedback, not in duration for its own sake.

First-timers usually ask some version of how seriously am I supposed to play this, and the answer determines most of the value: fully. The session only trains what you actually do in it.

Answer in complete spoken sentences, the way you would with a stranger deciding your offer. Do not restart answers when they wobble; recovering mid-answer is precisely the skill the real interview demands. Do not ask the interviewer what the right answer was until the feedback phase, and do not narrate your nervousness; manage it, which is what you will have to do on the day. If the session is remote, run it with the camera on and the setup you will really use, so the logistics get rehearsed along with the answers.

One deliberate exception to realism: it is fine, and useful, to ask for a question to be repeated or rephrased. That is allowed in real interviews too, and practicing the ask removes the hesitation to use it when it counts.

The script above holds across all three common formats; what varies is who sits across from you and what the feedback is worth.

A peer, a friend or fellow candidate, supplies presence and unpredictability but usually soft, encouraging feedback; useful for reps, weak for diagnosis. A professional coach gives the sharpest human judgment and costs accordingly, which in practice limits most candidates to 1 or 2 sessions placed late in preparation. An AI interviewer sits in between: SubcueAI's mock interview generates questions from your actual resume and the target job description, asks them out loud through a speaking interviewer, pushes follow-ups based on your answer, and ends with scored feedback per session. It will not read your body language the way a coach can, and it is available at midnight at a per-question credit cost, which changes how you schedule practice.

The formats stack sensibly rather than competing: AI rounds for volume and early reps, a coach session late if budget allows, a peer round if you want a human in the loop for free. How many rounds to run and how to space them is its own question, covered with the rest of the method in the mock interviews and practice answers.

The feedback phase is where a mock interview pays for itself, so it helps to know what competent feedback contains. Expect three layers, in descending order of value: structure, whether each answer had a point, a concrete example, and a result, or wandered; content, whether the example you chose actually demonstrated the competency the question targeted; and delivery, pace, filler words, confidence, the surface layer that is easiest to notice and least decisive.

Convert feedback into a list of 1 to 3 repairs, not a rewrite of everything. The highest-yield move after a session is re-answering just the flagged questions the next day, which turns the diagnosis into an actual repetition. In SubcueAI's mock interview the scoring arrives with the session record, so the flagged answers are reviewable rather than half-remembered; the session history also shows the credits each round consumed.

And if the feedback is uniformly positive, change something: a harder question set, a stricter interviewer, a less familiar story. A mock interview that stops finding problems has stopped being preparation; the goal before the real conversation, where the desktop app takes over for permitted live contexts, is to have already met your weaknesses in private.