How to Prepare for Timed Video Essays and Virtual Admissions Prompts

By Muntasir Minhaz • Published May 16, 2026 • Education Planning

Table of Contents

TL;DR
What Admissions Committees Score in a Video Response
The Main Platforms You Will Encounter
Standard Format and Length You Should Plan For
How to Structure a 60 to 90 Second Answer
The AI Detection Question
Equipment and Environment Setup
Final Proofreading Checklist Before You Submit
Sources

TL;DR

Most timed video prompts give you 10 to 60 seconds to think and 60 to 120 seconds to record one take, with no edits or retakes.

The two dominant platforms are Kira Talent (used by INSEAD, Yale SOM, Kellogg, Berkeley Haas, Oxford Saïd, Rotman, Imperial) and HireVue (used for Goldman Sachs campus recruiting). Brown caps its optional undergraduate video introduction at 90 seconds. MIT Sloan's Video Essay 2 gives 10 seconds of prep and 60 seconds of response. Goldman Sachs told 2026 EMEA private investing academy interns to stop using ChatGPT in HireVue interviews, as Fortune reported in June 2025.

How to Prepare for Timed Video Essays and Virtual Admissions Prompts

Timed video essays are the part of the application where you sit in front of a webcam, read a prompt you have never seen, and get under a minute to answer it on the first take. Schools and recruiters use them to test the parts of you a polished written essay hides: how you speak, how you think under a clock, and whether you sound like the person who wrote the rest of the file. This guide covers the formats you will face, what assessors score, and how to prepare without sounding rehearsed.

What Admissions Committees Score in a Video Response

Reviewers watch for four things at once: clarity of speech, structure of the answer, authenticity of voice, and fit with the program's stated values. MIT Sloan's admissions FAQ describes Video Essay 2 as a randomly generated open-ended question designed to help the committee get to know the applicant and see how they express themselves, and notes the question does not require prior preparation. Kira Talent is marketed to schools as a way to evaluate communication, leadership potential, and cultural fit beyond written applications, with admissions committees watching pre-recorded answers after submission.

Goldman Sachs's HireVue runs the same logic for hiring. Spokesperson Jennifer Zuccarelli told Fortune , as reported on 11 June 2025: "This language is consistent with what we send to any of our campus applicants across all positions.

We want to hear from our applicants in their own voice." One sentence is the rubric. Anything sounding like a chatbot, a memorised script, or a parent's words gets marked down.

Common reasons applicants get rejected at the video stage

Reading visibly from a script or second screen, breaking eye contact with the lens
Generic answers with no named project, course, person, or date
Filler heavy delivery, where 30 seconds of a 60-second answer carries no content
Answers contradicting the written application, like a stated weakness no other essay mentions
Poor audio, low light, or interruptions signalling the applicant did not test the setup
Over-rehearsed openers matching advice videos on YouTube word for word

The Main Platforms You Will Encounter

The format details below come from the platforms' own documentation and the schools using them. Always confirm timings on the school's portal before recording, since each program configures Kira and HireVue differently.

Kira Talent

Kira is an on-demand video and written interview platform. Schools using Kira, per GMAC's Kira Talent guide , include INSEAD, Oxford Saïd, Berkeley Haas, Rotman, Imperial, Kellogg, and Yale SOM. EDHEC also uses Kira, per its official prep page .

The format is fixed prep time followed by fixed recording time, with one take per question and no retakes. Typical timing is 30 to 60 seconds to prepare and 60 to 90 seconds to answer, though each school configures its own values.

Kira's practice mode lets you record unlimited dummy answers and replay them to check audio, video, and lighting before you start the real assessment. Use it.

HireVue

HireVue is the dominant platform for graduate employer interviews, including Goldman Sachs's campus recruiting. The standard Goldman Sachs format reported by candidates: five behavioural questions and one role-specific technical question, with 30 seconds of prep and a two-minute response per question. HireVue removed facial analysis from its assessments in March 2020 after SHRM reported CEO Kevin Parker's statement: "Over time the visual components contributed less to the assessment to the point where there was so much public concern to everything related to AI that it wasn't worth the concern it was causing people." The platform now scores audio, transcript content, and structure.

Brown University Video Introduction

Brown's Video Introduction page confirms the video is optional, capped at 90 seconds, and pre-recorded outside any timer. Supported formats are .3g2, .3gp, .avi, .m2v, .m4v, .mkv, .mpeg, .mpg, .mp4, .mxf, .webm, and .wmv, with a 5 GB file size limit.

Brown warns videos are not reviewable after upload and resubmission is not possible unless there is a technical issue. Preview the file before you click submit.

MIT Sloan Video Essay 2

MIT Sloan asks every MBA applicant a single, randomly generated open-ended question inside the application portal. You get 10 seconds to think and 60 seconds to respond, with no retakes and no edits, per the MIT Sloan MBA admissions FAQ , Ellin Lolis Consulting , and GMAT Club .

Reported prompts include "Tell us about a time a classmate or colleague wasn't contributing to a group project. What did you do?" and "Tell us about a time you took a risk. What did you learn from it?"

InitialView

InitialView is a third-party interview service aimed at international and non-traditional applicants. It pairs a 12 to 15 minute unscripted live conversation with a 20-minute proctored writing sample.

Once paid, the interview goes to as many participating schools as you choose, including Duke, Vanderbilt, and Boston University. This is the slow, conversational alternative to Kira and HireVue, and the format rewards depth over rehearsed soundbites.

Standard Format and Length You Should Plan For

Most timed prompts fit a tight envelope. Plan inside it before you record.

Component	Typical length
On-screen prompt reading	5 to 10 seconds
Prep window	10 to 60 seconds
Spoken answer	60 to 120 seconds
Word count delivered in 60 seconds	130 to 150 words at conversational pace
Word count delivered in 90 seconds	200 to 230 words
Word count delivered in 120 seconds	270 to 300 words

A 60-second answer holds roughly one full STAR story: one sentence of situation, one of task, two or three of action, one of result. Anything longer overflows. Anything shorter sounds thin.

How to Structure a 60 to 90 Second Answer

GMAC recommends the STAR structure (Situation, Task, Action, Result) for competency-based questions, which fits many Kira prompts. Apply it in this order and ratio.

Hook (5 to 10 seconds): Name the specific moment. For example, "Last March, during my final audit at a Big Four firm, my team lead resigned 48 hours before the client deadline." (Hypothetical example.)
Task (5 to 10 seconds): State your responsibility in the moment.
Action (25 to 40 seconds): Three concrete steps you took, not abstract qualities.
Result (10 to 15 seconds): A number, a measurable outcome, or a specific lesson.
Tie-back (5 seconds): One sentence connecting the story to the question or the program.

Strong example versus weak example, same prompt (hypothetical answers)

Prompt: Tell us about a time you stepped up within a team.

Weak answer (60 seconds, generic): "Teamwork is something I really value. In my last job, I was part of many teams and I always tried to step up whenever there was a challenge.

I think leadership is about being there for your team and helping them succeed. I have always believed that good communication and collaboration are the keys to great results. That is why I think I would fit well at your school, because I love working with diverse people and learning from them."

This answer names no project, no person, no number, no decision. It fits any school and any candidate, which is the problem.

Strong answer (60 seconds, specific): "In November 2024, our Series A diligence stalled because our product manager went on sudden medical leave. I was the junior on the deal team but the only person who had built the pricing model.

I rebuilt the deck in two days, briefed the partner on Friday morning, and walked the investor through unit economics on Monday. We closed at the original valuation, around fourteen million dollars. I learned stepping up means owning the artefact, not the title."

This answer names a month, a deal, a deliverable, a number, and a lesson. The reviewer pictures it.

The AI Detection Question

Schools and employers are now actively scanning for AI-written content. Beginning with the 2025 to 2026 cycle, Virginia Tech moved from two human readers per application to a hybrid model where an AI system and one human reviewer score each file, with a second human stepping in when their scores disagree by more than two points on a 12-point scale, per GradPilot's tracking . In February 2024, Duke stopped assigning numerical ratings to undergraduate application essays, citing the rise of generative AI and admissions consultants, per Inside Higher Ed and the Duke Chronicle .

For video specifically, the giveaways are different from written essays. Reviewers and platforms flag:

Flat, even prosody matching text-to-speech output
Vocabulary inconsistent with the rest of the application
Eye movement consistent with reading from a second screen
Stories with no proper nouns, no dates, and no first-person sensory detail
Polished structure with no hesitation markers, the human equivalent of an AI tell

Goldman Sachs's instruction to 2026 EMEA private investing academy interns to stop using ChatGPT in HireVue interviews, reported by Fortune in June 2025, is the clearest employer-side signal.

The same suspicion exists in MBA admissions even when the rule is not written down. The defence is the same in both cases: use your own examples, your own vocabulary, and your own pace.

How to sound like yourself on camera

Speak the answer aloud while preparing, do not silently mouth a script
Keep one or two filler words like "so" or "right" in the opening, real speech has them
Reference one detail only you would know, a building name, a course code, a colleague's first name
Vary sentence length, alternate short statements with one longer one
Do not start with "Great question" or "I'm passionate about", both are flagged opener clichés

Equipment and Environment Setup

Audio quality matters more than video quality.

A reviewer forgives a grainy webcam. They will not forgive an inaudible answer.

Use wired ethernet or a Wi-Fi 6 connection on the same band as your router
Wear wired earbuds with an inline microphone, or use a USB condenser microphone
Face a window or a soft lamp, never sit with a window behind you
Position the camera at eye level, stack books under the laptop if needed
Test the platform's actual interface, not a different one, the day before
Disable notifications, close every other tab, set the phone to airplane mode
Have a glass of water within reach but out of frame

Final Proofreading Checklist Before You Submit

For pre-recorded videos with playback (Brown, InitialView writing sample, some Kira configurations), run this list before clicking submit.

Audio audible at 50% volume on built-in laptop speakers
Face occupies the middle third of the frame, eyes level with the top third
No background noise, no notifications, no other voices
Answer addresses the literal prompt, not an adjacent question
Total length is within the platform's stated maximum
At least one specific name, date, or number per story
No banned vocabulary from the program's stated culture (Sloan dislikes corporate jargon, Brown dislikes formality)
Clothing matches what you would wear to an in-person admissions visit
File format is on the accepted list, file size under the cap
You watched the full take once before uploading

For live, no-retake formats (MIT Sloan Video Essay 2, Kira on-demand), the checklist runs before you start the assessment, not before you submit. Once you click start, the clock owns the file.