One-way interview questions to ask candidates

Ask three to five questions in a one-way video interview. Map each one to a single trait you care about, like written clarity, judgment, or how someone handles an objection. Phrase them so a rehearsed script falls flat. Keep one or two identical for everyone so you can compare answers cleanly across the pool.

That is the short answer. The rest of this page is the reasoning behind it. A one-way interview is mostly its questions, since the platform records video either way. What separates a screen that predicts fit from one that wastes everyone’s time is the question set. If you want the full setup, scoring, and candidate-experience playbook, how to run an asynchronous interview covers the rest.

The short version

Pick three to five questions. Tie each one to a single trait you are screening for, like written clarity, judgment, or how someone handles an objection. Write questions that demand a specific example or a decision, so candidates cannot read the same generic paragraph. Keep one or two questions identical across every applicant so you can compare cleanly, and swap in one or two that are specific to the role. Then score every answer against the same short rubric.

Start from the trait, not the question

Most question lists you will find online are organized backward. They give you fifty questions and leave you to guess which ones matter. Work the other way. Before you write anything, name the two or three things this stage has to prove.

For a customer-facing role that might be written and spoken clarity, empathy, and composure under pressure. For a sales role it might be discovery instinct and how someone reacts to an objection. For an early-career hire it might be coachability and how they think through a problem they have not seen before.

You are not making a final decision at this stage. You are deciding who is worth a live conversation. So screen for the signals that separate “worth a call” from “not yet,” and leave the rest for later rounds. Every question you keep should point at one named trait. If a question does not map to anything on your list, cut it.

Keep it to three to five questions

Three to five is a common range among people who run these well, and there is a plausible reason behind it. Each question can cost the candidate far more than the recording time suggests. One candidate writing on Reddit put it bluntly: “On average it takes something like 60 to 90 minutes of work per 1-minute of video.” Their account: clean the room, pick the outfit, prepare talking points, rehearse, then re-record until it feels right. That cost is invisible to you and can be very real to them. Treat the number as one person’s estimate, not a measured average. The direction is what matters.

Three to five questions, at roughly sixty to ninety seconds of recording each, respects that and still gives you enough to compare. Push past five or six and the risk is that fewer people finish. And the ones who walk away may be the candidates juggling a current job and a family, not the least qualified. The Reddit complaint above points the same way: the candidate flagged people “who have families and children” and no quiet place to record after work. A shorter interview is not a weaker screen. It is often a fairer one. For more on length, see how long should a one-way video interview be.

Write questions that kill the script

The most common complaint from people who do not like one-way interviews is that they produce hollow answers. A recruiter laying out the case against them wrote that “you’ll just get scripted answers as it won’t be an organic conversation.” That is a real failure mode, and it is almost always a question-design problem, not a format problem.

Generic questions invite generic answers. Specific questions force a real one. The fix is to ask for something only that person could say.

• Weak: “Are you a good communicator?”

• Strong: “Tell me about a time you had to explain something technical to someone who did not have the background. What did you do, and how did you know it landed?”

• Weak: “How do you handle pressure?”

• Strong: “Walk me through the last time a customer was angry and you were the one who had to fix it. What did you say first?”

• Weak: “Are you a problem solver?”

• Strong: “Here is a situation you would actually hit in this role. [Describe it in two sentences.] How would you approach it, and what would you check first?”

The strong versions share three traits. They ask for a specific example or a concrete decision. They cannot be answered by reading a rehearsed paragraph that would fit any company. And they reveal thinking, not just history. Scenario prompts are especially good at this, because a candidate cannot pre-write the answer to a problem they are seeing for the first time.

A note on AI-assisted answers, since it is on many hiring teams’ minds in 2026. You will not stop candidates from using AI to prepare, and some already run the job description through a chatbot to predict the questions. What you can do is ask questions specific enough that a generic AI answer reads as exactly that. A prompt rooted in a real scenario from your team, or one that asks for a particular moment in the candidate’s own experience, is much harder to fake convincingly than “describe your greatest strength.”

Anchor a couple of questions so you can compare

There is a tension in question selection. Role-specific questions surface the most relevant signal, but if every candidate answers different questions you have nothing to line up side by side. The way out is to split your set.

Keep one or two anchor questions identical for everyone applying to a given role. A motivation question (“why this role, specifically”) and a communication question (“explain something complex simply”) work as anchors for almost any job. Those give you a consistent column to compare across the whole pool.

Then add one or two questions that are specific to the work. This is where the by-role question banks earn their place. Each one gives you a starting set of questions that the role actually demands, written to surface the trait that matters:

• Sales and customer success
• Nursing and healthcare
• Software engineer
• Customer service and call center
• Retail and hospitality
• Finance, teaching, virtual assistant, and graduates

Take a couple of role-specific prompts from the relevant bank, pair them with your anchors, and you have a set that is both comparable and relevant.

Order them so people warm up

Sequence matters more than it looks. Open with the lowest-stakes question, usually motivation or a short “tell me about yourself,” so the first thing a nervous candidate records is one they can answer without thinking hard. Put your sharpest scenario question second or third, once they have settled. Try not to end on the hardest one. A stumble on the final answer is the freshest in your mind when you stop watching, and it can color how you score the rest.

If your tool lets you record yourself asking the questions on video, do it. One recruiter who runs one-way interviews well described recording themselves asking each question and pairing it with an email explaining what to expect and how answers get evaluated. It costs you a few minutes and it changes how the whole thing feels to the person on the other end. More on that in how to run a one-way interview candidates don’t hate.

Decide what a good answer looks like before you watch any

Choosing the questions is only half the job. The other half is deciding, in advance, what a strong answer to each one sounds like. Write a one-line description of a good answer per question, or score each on a one-to-four scale against the trait it maps to. This is what keeps a large pool fair, because the rubric, not your mood at 4pm, decides who advances. It also gives you a defensible record of why one candidate moved forward and another did not.

This matters most because of a complaint that shows up again and again from candidates: nobody seems to watch. One person who finished four recorded interviews and heard nothing for two months asked the obvious question, “what’s the point?” If you are going to ask people to record answers, the deal is that you watch them and score them against a clear standard, and you do it quickly. The full approach is in how to score async interviews.

Preempt the questions candidates will ask you

Candidates judge your question selection more than you might think. A widely shared Reddit post listed the four questions thoughtful applicants ask before agreeing to a one-way interview: what are you really looking for, are you assessing things you could not get from a phone screen, how many of these do you actually watch, and is this an efficient use of everyone’s time.

Good question selection answers all four before they are asked. Questions that clearly map to the role show you know what you are looking for. Questions that surface real thinking prove you are getting something a resume and a phone screen would not. Keeping it to three to five, and reviewing fast, shows you respect their time. The strength of your questions is the clearest signal a candidate gets about whether your process is serious or lazy.

Put it together

1. Name two or three traits this stage must prove.
2. Choose three to five questions total, each mapped to one trait.
3. Make one or two of them anchors you ask everyone, for clean comparison.
4. Pull one or two role-specific prompts from the by-role banks.
5. Rewrite any generic question into one that demands a specific example or decision.
6. Order them easy-to-hard, and never end on the hardest.
7. Write what a good answer looks like for each before you watch a single video.

Get the questions right and the rest of the one-way interview gets easier: answers are comparable, scripts fall flat, and the candidates worth a call rise to the top. When you are ready to build the full process around these questions, start with how to run an asynchronous interview.