Mock-interview prep has two failure modes: too little and too much. Candidates who skip mocks entirely arrive on interview day with answers that have never been tested out loud — a common reason strong-on-paper candidates underperform. Candidates who over-prep (12+ mocks, 60+ hours of study) often arrive stale and over-rehearsed, with rote answers that have lost their texture.

There's a working middle. This post is what it looks like.

Why mocks specifically, not just more study

Solo study and content review max out faster than candidates think. Past a certain point — usually around hour 10-15 of solo prep — you're not learning anything new; you're just re-reading material you already know. The next bottleneck is execution: can you actually deliver the answer in 90 seconds, out loud, when an interviewer is looking at you?

The only way to test execution is to perform it. That's what mocks do. Anything you can rehearse in your head feels easy; everything you have to perform out loud is harder than you expect. The gap between mental and verbal performance is what mocks expose and fix.

The three-week structure

The weekly progression is the working shape. The total time investment lands around 30-35 hours across three weeks, which is enough for most senior interviews without crowding out the rest of your life.

The mistake most candidates make is front-loading — spending week 1 on mocks before they have material to work with. Mocks in week 1 are usually low-quality: the candidate hasn't drafted answers, hasn't done company research, and is essentially using the mock to figure out what to say. This wastes the friend's time and yields no useful signal.

The opposite mistake is back-loading — saving all the mocks for the last 3 days. By that point, you don't have time to absorb the feedback. Mistakes you surface in your last mock 36 hours before the interview will still be in your performance.

The schedule, week by week

Week 1: build the base. Most of your time is content. Re-read the JD. Research the company — recent funding, product launches, leadership departures or arrivals, engineering blog if relevant. Skim Glassdoor for interview-question reports. Draft 5-7 stories using the STAR framework (situation, task, action, result) covering: a hard project, a conflict, a failure, a leadership moment, a customer-impact moment, and one role-specific story. See behavioral-interview-star-framework for the structure.

The solo answer-drafting is the work most candidates skip. Writing out answers — actually typing them — surfaces the gaps that mental rehearsal hides. You'll discover which stories don't have a clear "I did X" sentence, which results aren't quantified, and which transitions are clunky.

Week 2: introduce mocks. Two mocks. Pick a friend or peer who's willing to ask questions and stay in character. The first mock will be uncomfortable. That's the point — discomfort is information.

After each mock, write down the 2-3 things that broke down. Not "I was nervous." Specifically: "I went over time on the 'tell me about yourself' answer." "I couldn't recall the metric for the X project." "I rambled when asked about the failure story." Each one is a target for the next week.

Week 3: simulate the loop. Three to four mocks. At least one should be with someone who knows the role — a peer who's interviewed for similar positions, a friend who's a hiring manager, or a paid mock from a platform like Exponent or Karat. Stack two mocks back-to-back at least once to simulate fatigue. The interview day is exhausting; you should know what your performance looks like in hour three.

Drill the specific weak areas. If your "why this company" answer was weak in week 2, write three versions and rehearse the best one. If your behavioral stories were vague, find the specific numbers and add them.

Day before: rest. This is where candidates most reliably hurt themselves. Don't cram. Re-read your notes for 30 minutes in the morning. Light exercise. Eat normally. Sleep early. The marginal hour of prep on day-before is worth less than the marginal hour of rest.

How many mocks actually matter

The honest number from interview-prep platforms and candidate self-reports: 5-7 mocks is roughly the sweet spot before diminishing returns set in. Below 3, candidates consistently report being surprised by the difficulty of answering out loud. Above 10, candidates report feeling stale and rehearsed.

The 5-7 number assumes variety. Seven mocks with the same person and the same question bank teach less than four mocks with four different interviewers. Variety surfaces different weaknesses. Your spouse won't probe the way an engineering manager will; an engineering manager won't probe the way a recruiter will. Both kinds of probing matter.

Where to find mocks

A few channels that work:

Friends and peers. The cheapest option. Quality varies, but a willing friend who takes the role seriously can run a competent behavioral mock. Use them for week 2, not week 3.

Paid platforms. Exponent, Karat, Interviewing.io, Pramp. Quality varies but the better platforms (Exponent for behavioral/PM, Karat for engineering) pair you with real industry interviewers. $100-$250 per mock; budget for 2-3 in week 3.

Former colleagues. A peer or manager from a previous role who's not currently at your target company. Low cost, high signal — they know the conversational norms of the role.

LinkedIn ask. Reach out to 1-2 people in roles similar to your target with a polite ask: "I'm prepping for a final round at [Company] and would value a 30-minute mock if you're open. I'm specifically working on [area]." Success rate is roughly 30%; the people who say yes often give exceptional feedback.

What to do after each mock

The mock isn't useful unless you process it. A 15-minute post-mock review:

1. Three things that worked. What landed cleanly. What's a story you can keep.
2. Three things that broke down. Specific moments where you stumbled or the answer felt weak.
3. One thing to drill before next mock. Pick the highest-leverage fix from #2 and rehearse it 3-5 times before the next mock.

The candidates who improve fastest are the ones who treat each mock as an experiment and the post-review as the lab notebook.

For the interview-day basics, see phone-screen-what-recruiters-evaluate. For the specific 90-second opening, see tell-me-about-yourself-90-seconds. For the questions-to-ask-the-interviewer portion, see questions-to-ask-the-interviewer.

What this isn't

A few clarifications:

• It's not a memorization exercise. Don't memorize answers word-for-word. Memorize the structure of each answer and the 1-2 anchor sentences.
• It's not equal across roles. Engineering and product interviews need more time on technical reps; sales and customer-success roles need more on situational and stakeholder questions. Adjust the mix.
• It's not a substitute for company research. No amount of mock-prep makes up for not knowing what the company does or who the team is.

The short version: 30-35 hours across three weeks, 5-7 mocks in the back half, rest the day before. Most candidates either skip mocks entirely or over-stuff their schedule. The middle is where the actual prep happens.