HireVue Interview Guide (2026): Format, Scoring, and a 5-Day Prep Plan

A HireVue interview is a one-way, asynchronous screen your employer sends before a real interviewer ever gets on the phone. You receive a link, you get 3 to 8 questions on the screen, and you record spoken answers against a timer, usually with 30 seconds of prep time and 1 to 3 minutes to answer. There is no interviewer nodding back. Most employers cap you at zero retakes on the real questions; a minority allow one re-record on the practice question only. The entire session runs 20 to 30 minutes.

The recording gets evaluated later against a structured rubric. On some employers it is human-only. On others an AI first-pass scores it against a competency model, and a human reviews the flagged candidates. Either way the winner is the same: the candidate whose answers are already structured and timed before the recording starts. That is the muscle to build, and the rest of this page is how to build it fast.

HireVue is heaviest in early-career pipelines: summer analyst programs, graduate rotations, and high-volume operator hiring where the pool exceeds recruiter capacity. If you are interviewing at any of the below and the role sits in an early-career or high-volume band, expect a HireVue link:

• Investment banking + finance rotations: JPMorgan, Goldman Sachs, Morgan Stanley, Citi, Bank of America, Capital One, Barclays, Deutsche Bank
• Consulting: Bain, BCG, Deloitte, EY, Oliver Wyman
• Consumer + FMCG graduate schemes: Unilever (games first, then video), Nestle, Mondelez, PepsiCo
• Tech + engineering early-career: IBM, Microsoft, Amazon (some ops roles), Boeing, Mercedes-Benz
• Retail + hospitality volume hiring: Tesco, Hilton, McDonald's, Delta
• Energy + logistics: Shell, BP

Employers configure HireVue differently, so a Goldman summer analyst screen and a Delta ramp-agent screen use the same product with very different question banks and evaluation weights. Do not assume the tips one candidate gives you online map to your specific screen. Match your prep to the role, not to HireVue in the abstract.

Most third-party guides still claim HireVue analyzes your facial expressions. It does not, and has not since January 2021. HireVue removed facial-expression analysis after the EPIC FTC complaint and ahead of the Illinois Artificial Intelligence Video Interview Act (see SHRM coverage and Fortune reporting). Advice built on the old model (smile more, look confident on camera) is not wrong, but it is not what wins the score.

What actually gets scored in 2026, per HireVue's own candidate FAQ and the current competency-model documentation:

• Verbal content mapped to competencies. Communication, problem-solving, conscientiousness, teamwork, and role-specific competencies configured by the employer. The score reads what you actually said and how it maps to the competency the question was designed to probe.
• Response structure. Answers that have a specific situation, action, and outcome score higher than answers that speak in generalities. STAR structure carries the same weight it does in a live interview.
• Delivery signals in speech. Pacing, filler-word density, and clarity of enunciation. Not tone-of-voice sentiment; the actual speech metrics.
• Integrity signals. Tab-switching during the recording, obvious use of AI-generated audio, and answers that cluster suspiciously close to other candidates' answers all get flagged. Do not read a scripted answer off a second monitor; the platform tracks focus.

The single sentence to internalize: the system rewards evidence. A vague answer that ends in "I learned a lot from that experience" scores below a specific answer that names the decision you made and the outcome that followed, even if the specific answer is less polished.

Most candidates prepare for the video interview and get surprised by the game-based assessment their employer bundled in. HireVue actually ships four separate assessment types, and any given screen can combine them:

1. Video Interview. The default: 3 to 8 spoken-answer questions on video. This is what most people mean by "HireVue."
2. Game-Based Assessments. Short cognitive and behavioral games. The core set is around nine games: Numerosity, Flashback, Digitspan, Pathfinder, Puzzle Picture, Shapedance, Singularity, PortraitXT, and E-Motions. Each targets a specific cognitive skill (working memory, attention, spatial reasoning, emotion recognition) and typically runs 20 to 60 seconds. There is no way to "study" them; the useful prep is warm-up puzzles the night before so your first plays are not your slowest. Unilever, Boston Consulting Group, and JPMorgan all lean on these.
3. Coding Challenges (CodeVue). For technical roles: 1 to 3 problems in Java, Python, Perl, PHP, or Haskell. Behaves like a shorter LeetCode or CodeSignal session with a timer. Prep the way you would for those platforms, not the way you would for a behavioral interview.
4. Virtual Job Tryout. Role-simulation scenarios. Customer service candidates handle a mock chat; sales candidates work through an objection; analyst candidates read a scenario brief and answer questions. The tryout is graded on judgment, not speed.

Ask the recruiter (or check the invite email) which formats are in your screen before you build a prep plan. If the invite mentions "games and video," do not skip the games.

The specific questions vary by employer, but the type mix is consistent. Expect a blend of:

• Motivational (opens the interview). "Tell me about yourself." "Why this company?" "Why this role, now?" The screen for whether you read the posting.
• Behavioral (usually the middle two or three questions). "Tell me about a time you dealt with a difficult teammate." "Describe a time you missed a deadline." "Walk me through a project you led." Structured with the STAR method.
• Situational (closes). "You inherit a project two weeks behind schedule. What do you do first?" "A customer is upset about a policy you cannot change. Walk me through your first two minutes." Answered with a specific approach and one honest trade-off.

A worked STAR answer, timed to a HireVue window:

Question: "Tell me about a time you disagreed with a decision your manager made."

"Last spring my manager wanted to cut our QA pass from two days to one to hit a release date. [Situation.] I thought we would ship a regression into billing, which had already broken twice that quarter. [Task.] Rather than push back in the meeting, I pulled defect data from the last six releases and showed that every billing bug had been caught on day two of QA, never day one. I proposed keeping the full pass on billing only and compressing everything else. [Action.] He took it. We shipped two days late instead of a week late, and no billing regression. [Result.] What I would do differently is raise it with the data ready rather than reacting in the room."

That answer runs 75 to 90 seconds spoken at a normal pace. It has a specific decision, evidence, an action, and a result, and it closes with the self-reflection that answers the follow-up before it gets asked. That is the shape.

1. Camera at eye level. A laptop on the desk points up your nose. Stack books under it or use an external webcam so the camera is level with your eyes. This one change reads as confidence more than any single content choice.
2. Look at the lens, not the preview. Your eye line only reads as engagement on the recording if you speak toward the camera itself. Cover the preview thumbnail with a sticky note if you have to.
3. Light on your face, not behind you. A window or lamp in front of you, not behind. Backlighting turns your face into a silhouette. A cheap ring light works if daylight is not available.
4. Neutral background. Blank wall or bookshelf. Skip the virtual background: the edge artifacting reads as distracting on a scored recording.
5. Wired mic or earbuds with a mic. Laptop microphones pick up room hum and every keystroke. Even $20 earbuds with an inline mic beat the built-in on transcription accuracy.
6. Run the practice question for real. Do not skip it. Watch the playback of your practice with the sound on. What you hear will not match what you thought you sounded like.
7. Test wifi in the room you will actually record in. Not the kitchen router speed test. The exact chair. Ethernet if you can.
8. Prepare on the assumption that the first take counts. A minority of employers allow one re-record; most allow zero. Rehearse to a timer so your first take is your best take.

Most candidates receive a HireVue link with a 5-to-7-day deadline. Here is the plan that fits that window:

• Day 1: Diagnose. Reread the job description. List the 3 competencies you would probe if you were the hiring manager. Draft one STAR story per competency.
• Day 2: Baseline record. Answer 5 practice questions out loud on video. Watch the recording. Do not edit; just notice what happened. Most people discover the middle of their answer collapses around second 40.
• Day 3: Rewrite. Turn each STAR story into a 75-to-90-second spoken script. Do not memorize word-for-word; internalize the beats.
• Day 4: Reps to time. Record each answer three times, each time to a strict 90-second timer. The third take is where the answer stops sounding recited.
• Day 5: Full run + technical check. Set up the room exactly as it will be on interview day. Run 3 questions back-to-back with a timer. Fix whatever the recording shows: camera height, lighting, background, mic hum.

The free practice tool handles the record + score loop for you: paste a job description, get five questions, answer them out loud, and get written feedback on structure, timing, filler words, and specificity. It runs in the browser, no account, unlimited retakes. Which is the exact thing the real screen does not give you.

If you need extra prep time, an alternative format, or captioning, request it from the recruiter before you start the interview. HireVue supports accommodations through the hiring employer, not through the platform directly, so the ask goes to your recruiter. HireVue also publishes specific guidance for neurodivergent candidates, available in their candidate resources.

In Illinois, the Artificial Intelligence Video Interview Act requires employers using AI to analyze video interviews to disclose it, explain what characteristics are being evaluated, and get your consent before recording. In Maryland, the state AI-interview disclosure law requires similar consent. If you are in either state and you were not disclosed to, you are within your rights to ask the employer to confirm in writing before you record.

On data retention: the employer controls how long your recording is kept, not HireVue. To delete your recording, request it from the company that sent the invite, not from HireVue support.

The preparation that transfers to a HireVue screen is recording yourself answering timed questions out loud, watching the recording, and repeating. Reading through interview-questions lists does not transfer. Rehearsing answers in your head does not transfer. The gap between an answer you have thought about and an answer you have said out loud is larger than most candidates expect.

Doing that loop manually means writing your own question list, timing yourself, and self-scoring, which most people do once and stop. Capstone's free AI coach runs the loop for you: it generates behavioral and situational questions for your target role, times your spoken answers, scores clarity, pacing, structure, and filler words, and quotes the specific moment to fix before the next take. Practicing for an AI-evaluated interview with an AI that scores the same dimensions is the closest available rehearsal for the real format.

Most users sharpen noticeably within 3 to 5 sessions, which fits the typical gap between receiving a HireVue link and its deadline. The 3-day free trial unlocks unlimited practice, saved history, and per-answer video feedback for candidates who want more than a one-off session.