How to do a mock interview by yourself

A mock interview needs three things a solo candidate can absolutely supply: questions you did not choose, answers produced out loud under time pressure, and an honest review afterwards. Everything else is staging.

The minimal loop looks like this. Collect twenty role-relevant questions and put them somewhere that can surprise you, shuffled cards or a randomizer. Draw one, start a timer, and answer aloud in full sentences as if a person were across the table; ninety seconds is a healthy ceiling for a behavioral answer. Then, before drawing the next card, spend one minute reviewing: did the answer have a beginning, a concrete middle, and a result, or did it orbit?

Do five questions per session and stop. The value is in the production and review, not the volume, and short sessions are what make daily repetition sustainable across an interview week.

Nobody enjoys hearing their own recorded answers, and that discomfort is precisely the signal. This section covers the one-recording review, which delivers most of the feedback a solo format can produce in a few minutes.

Record audio for one full session on your phone. On replay, listen for four things and nothing else: whether the answer started with the point or wandered toward it, whether a concrete example appeared and carried detail, whether the ending landed on a result or trailed off, and the filler-word rate when you were thinking. Mark one improvement per answer, not five; a single fix you actually apply beats a list you avoid.

One pass per week of practice is enough. Recording every session turns rehearsal into performance review and makes people quit; the recording is a diagnostic, not a habit.

The card-and-timer loop has a structural ceiling: it never pushes back. Real interviewers respond to your answer, probe the vague part, and pull the conversation somewhere you did not script. Self-drawn questions cannot do that, and that gap is exactly what an AI interviewer covers.

SubcueAI's mock interview reads your resume and the target job description, asks questions out loud through a named interviewer with a voice, and generates each next question from what you actually said, including the follow-up that digs into the weak spot. Sessions end with scored feedback: what worked, what needs another rep. It runs in the browser with typed or spoken answers, so the practice loop above carries over directly; the difference is that something now answers back.

A practical week, in the spirit of a named example: a frontend engineer with a Thursday onsite runs the card loop solo on Monday to warm up, then two AI mock rounds Tuesday and Wednesday evening focused on her two weakest stories, using the scoring to confirm the tightened versions hold up under follow-ups. Costs sit on the free tier's monthly credits for a normal interview week; the pricing page has the exact numbers, and the broader practice playbook lives in the mock interviews and practice answers.