Interviewing
What to Decide in the Thirty Seconds Before You Record
Use the thirty seconds to choose, not to draft. Pick one example, decide what its result was, and say that result to yourself first. In mock recorded interviews rated by people, preparation time helped, and the gain came from actively planning the answer rather than from settling nerves. Check the screen for whether you get more than one take, because on at least one platform a limited countdown means exactly one. Then talk for as long as the question deserves and stop.
The takeMost advice about recorded interviews spends this window on posture, framing and a slow breath. That advice is not wrong, it is just late: lighting is a problem you solve before you open the link, and once a question is on screen the only thing that can still change the recording is what you decide to say. Treat the countdown as the last editorial decision you get, and the ten minutes beforehand as the technical one.
Where Olive fits
Open a role and see what the work shows
An Olive assessment is not an interview and there is no camera in it: you work a real assignment with an AI assistant on your own clock, think aloud or type, and a person writes or edits six findings on what happened. You get the same report the employer reads, free.
Assess your shortlistWhat Is the Prep Timer Actually For?
It is for planning the answer. Basch and colleagues tested the preparation time given before each question in asynchronous video interviews. The measured gain in performance ran through active response preparation, and not through lower strain or through impression management 1. These were mock interviews rated by people rather than a scoring model, and the published abstract reports no effect sizes.
Settling your nerves is fine, and it is not what the window was measured to be worth.
A second set of experiments points the same way. Lukacik and Bourdage ran three mock asynchronous video interview studies covering six design features, with 206, 230 and 297 participants. Longer preparation time and the chance to re-record raised interview performance ratings. Faking warnings, and being told that a human or an automated system would evaluate the answers, changed nothing in behaviour, reactions or performance 2. No job was at stake in any of the three, and the ratings came from the researchers' own raters rather than from a commercial scoring model.
So the window has a job, and the job is a decision. Thirty seconds is not enough time to write an answer. It is enough to choose which of your prepared examples answers the question on screen, and to settle what happened at the end of it.
All of that assumes you arrived with examples, which is what an hour of preparation before you open the link is for.
Pick the Ending First
Decide what the result was before you decide how the story starts. The recording clock takes the end of an answer, never the beginning, so the part most likely to get cut is the part carrying the point. Know your last line and everything before it can be shortened live without losing what the listener needed.
Three things fit in the window, in this order.
- the example you will use, chosen against the question actually asked rather than the one you rehearsed
- what happened at the end of it, in one sentence you could say out loud right now
- the single piece of context the listener needs first for that ending to make sense
The order matters because the ending is the part least likely to be rehearsed. Openings are easy to rehearse. A specific result, with the number or the decision that made it a result, is what carries the answer and what a script leaves vague.
The shape alone stopped carrying an answer some time ago. A four-beat structure costs nothing to produce now, which is why rehearsed STAR answers land flat even when the structure is perfect.
One thing not to do with the window: memorise a first sentence. A sentence written thirty seconds ago and then reproduced word for word tends to come out stiff, because you are retrieving it rather than saying it. Hold the ending and the shape of the story, and let the words arrive as you speak.
The countdown is where the decision happens and the recording is only the delivery.
How Long Should the Answer Actually Run?
Long enough to be clear and relevant, and at two minutes the evidence gives no reason to stop well short of the limit. Orji and Bangerter gave 185 participants in a mock one-way interview a response limit of one, three or five minutes. Performance followed a curve, highest under the three-minute limit and lower at one or five, and the ratings that moved were clarity and relevance while judged competence stayed stable 3.
That is a result about the setting an employer chose, not about how long you talk inside it. Speaking longer went with lower anxiety and higher performance, but only up to the plateau, and the authors say offering more than three minutes did not improve outcomes.
So use the two minutes, without padding them.
Clarity and relevance are the things to aim the countdown at. Relevance means answering the question that was asked rather than the nearest one you prepared. Clarity means a listener can tell on one pass what you did and what came of it.
The people on the other side are reading about this round as well. The employer-side case for keeping a one-way video screen is reach and scheduling, and the argument that scripted answers no longer separate candidates on substance is written for hiring teams. It tells them a recording is at best a first filter and that the decision belongs where a second question can be asked.
Check What the Countdown Means on Your Platform
The same countdown means different things on different products, which is why the instructions screen matters. On Spark Hire, an employer sets think time per question between 30 seconds and 5 minutes. Limiting it also limits the candidate to one take, so the answer has to be right on the first attempt 4. On HireVue, preparation time is a per-question setting of zero to five minutes or unlimited, and retries are a separate switch the employer sets 5.
Each of those numbers comes from a vendor writing about its own product, which is not an audit, and neither set transfers to a third platform. What transfers is the habit. Look on the screen for the think time, the number of takes and the response limit, because those three numbers change what the countdown is for.
One take and five takes are not the same interview.
Spark Hire's candidate documentation describes the sequence in a way worth knowing in advance. With limited think time you click to view the question, a countdown starts, and recording begins automatically when it ends 6. Nothing asks whether you are ready. The same page says a candidate out of takes who still does not want to submit is offered the option to skip, though it does not say what skipping costs.
If it goes wrong anyway, one platform publishes what happens next. HireVue's explainability statement gives two conditions for an answer to be scored by AI: the candidate consented, and the system detects enough content in the response. An answer that is too short, not given in the allotted time, or not clear enough to the system fails the second condition and is flagged for human review instead 7. The vendor calls this a small proportion of candidates and publishes no number.
Where the answer went is not the same as how it did.
Common questions
What exactly should I do in the thirty seconds?
Choose the example, then decide its ending. Pick the story that answers the question actually on screen, settle what the result was, and pick the one piece of context a listener needs before that result makes sense. Say the ending to yourself once. Do not try to draft or memorise a first sentence, because a sentence you are retrieving comes out stiffer than one you are saying.
Should I write notes during the preparation time?
Only a few words, if anything, and never full sentences. The measured benefit of preparation time in mock recorded interviews came through actively planning the answer rather than through calming down. If you write anything, make it the example and its ending. On Spark Hire, for one, the countdown flips straight into recording when it ends, so anything you write has to be finished before that happens rather than while it does.
Can I get more preparation time or another take?
Both are usually settings the employer chose. HireVue's own statement lists preparation time per question as a setting of zero to five minutes or unlimited, and question retries as a separate yes-or-no switch. Spark Hire lets an employer set think time between 30 seconds and 5 minutes, and limiting it also limits the candidate to a single take. Read the instructions screen before the first question rather than assuming either way, and ask the recruiter if the screen does not say.
What happens if I run out of time in the middle of an answer?
Platforms differ here, and one of them publishes its rule. HireVue says an answer is scored by AI only if the candidate consented and the system detects enough content, and that an answer which is too short or was not given in the allotted time is flagged for human review instead. The vendor calls that a small proportion of candidates and gives no number. Treat it as a safety net rather than a plan, since another platform may simply rate a thin answer low.
Does it change anything if an AI is scoring the answer?
Being told so made no measured difference in the experiment that tested it. In a mock asynchronous interview experiment with 297 university students, being told that a human or an automated system would evaluate the answers did not change behaviour, reactions or performance, and neither did warnings about faking 2. That was a mock interview with no job at stake, and what participants were told about the evaluator was a manipulation rather than a real scoring model. The answer that is clear and specific is the one to give either way.
References
- 1. A good thing takes time: The role of preparation time in asynchronous video interviews doi.org Supports the article's central claim that the preparation window's measured benefit ran through active response preparation, with no mediation through strain or impression management.
- 2. Does Design Matter? The (Limited) Effects of Six Asynchronous Video Interview Design Features on Impression Management, Reactions, and Evaluations doi.org Supports the three-experiment sample sizes, the finding that longer preparation time and re-recording raised performance ratings, and the finding that evaluator type and faking warnings changed nothing.
- 3. Tick Tock, Tick Tock, Time's Up! How Response Time Limits in Asynchronous Video Interviews Influence Interview Anxiety, Impression Management, and Performance doi.org Supports the 185-participant response-limit experiment, the curvilinear result peaking at a three-minute limit, and the finding that clarity and relevance moved while competence ratings stayed stable.
- 4. Set think time limits for One-Way Interview questions support.sparkhire.com Supports the statement that Spark Hire think time is set between 30 seconds and 5 minutes and that limiting it also limits the candidate to one take.
- 5. 2025 Explainability Statement hirevue.com Supports the statement that preparation time per question is an employer setting of zero to five minutes or unlimited and that question retries are a separate switch the employer sets.
- 6. How a Spark Hire one-way video interview works candidates.sparkhire.com Supports the description of the countdown starting when the question is viewed and recording beginning automatically when it ends, and the option to skip a question once takes run out.
- 7. 2025 Explainability Statement hirevue.com Supports the two conditions for AI scoring and the statement that an answer too short, not given in the allotted time, or unclear to the system is flagged for human review instead.
7 sources, numbered by first appearance. How Olive sources claims
General guidance for hiring teams. What works at one company and one volume may not transfer to yours.
Olive assesses how a person works with AI. It does not detect AI-written documents, and it never produces a score, a ranking, or a match percentage for a person. Candidates read the same report the employer reads.