A one way video interview lets candidates record answers to fixed questions on their own schedule, while recruiters review them later. It works best as a short, structured screening step, not as a substitute for a conversation. Use job-related questions, a shared scorecard, an alternative channel, clear data notices, and a fast human follow-up.

What is a one way video interview?

A one way video interview is an asynchronous interview. The employer records or writes the questions in advance. Each candidate opens a link, receives the same prompts, and records an answer without a recruiter present. The hiring team watches the responses later.

This is different from a live video call. A live call is synchronous and allows follow-up questions. It is also different from an AI-led voice interview, which can ask spoken follow-ups without requiring a camera. The defining feature is not AI. It is the delay between the candidate’s response and the recruiter’s review.

The format can create a consistent first-round screen across time zones or large applicant pools. It also asks the candidate to perform in an unusually empty social setting. Nobody nods, clarifies a vague question, or reacts to an answer. That trade-off should drive the design.

How does an asynchronous video interview actually work?

A credible process begins before the camera turns on. The invitation should name the purpose, expected effort, deadline, device requirements, retake policy, reviewers, next step, and a contact for technical or accessibility support. Candidates should be able to see this information before deciding whether to participate.

1. Define the evidence. Translate the role requirements into observable behaviors. Build the interview guide and scoring rubric before recording questions.
2. Invite the candidate. Explain why this format is being used and offer a meaningful alternative where needed.
3. Run a practice prompt. Let the candidate test the camera, microphone, lighting, and upload without being assessed.
4. Present the same core questions. State preparation time, answer limit, and permitted retakes before each response.
5. Store and route the recordings. Limit access to named reviewers and connect results to the ATS without creating uncontrolled copies.
6. Score independently. Reviewers score evidence against anchored criteria before discussing candidates.
7. Close the loop. Tell candidates when a human decision will arrive and provide a channel for their questions.

A good workflow never treats completion as proof of motivation. A person may decline because of disability, bandwidth, privacy, caregiving, cultural expectations, or an aversion to being recorded. Those factors do not establish job performance.

How many questions should a one-way video interview have?

There is no universal, research-backed optimum. The right number depends on the complexity of the evidence and the burden of each answer. A practical pilot starts with four scored questions plus an unscored practice prompt. Add a question only when it measures something that the application and other questions cannot establish.

Question count alone is a weak measure of burden. Four prompts with long preparation windows, multiple retakes, and multi-part answers may feel heavier than six concise factual prompts. Show the expected total time in the invitation, then measure the actual median time and exits by stage.

Avoid “Tell us about yourself” unless you can say what evidence it should produce. Avoid compound prompts that ask for a conflict, a failure, a lesson, and a future plan in one recording. A candidate cannot request clarification, so each prompt must stand on its own.

How should recruiters evaluate recorded answers fairly?

Standardized questions do not automatically create a fair interview. Fairness comes from job analysis, consistent administration, anchored scoring, reviewer training, and documented decisions. Video adds visual information that may be irrelevant to the role. That makes discipline more important.

Use one scorecard for every candidate in the same role. Define what weak, acceptable, and strong evidence looks like. Ask reviewers to record evidence before a rating. Do not let an overall impression replace item-level scores. If the team uses AI to summarize or rank responses, validate the output separately and keep accountable human review. The wider risks are covered in this guide to AI interview bias and fairness.

• Score answer content before discussing presence or style.
• Do not infer personality from gaze, facial movement, accent, room, clothing, or recording quality.
• Use two reviewers for borderline cases when the decision has high impact.
• Audit rating gaps by question, reviewer, device failures, and candidate groups where lawful.
• Recalibrate when reviewers interpret an anchor differently.

Retakes are a design choice, not a moral test. No retakes may capture stress and technical accidents. Unlimited retakes can reward available time and editing skill. A limited, disclosed retake policy is often easier to defend. Whatever the rule, apply it consistently and never hide it until the candidate starts.

Why do candidates drop out of asynchronous video interviews?

There is no trustworthy universal completion benchmark. Vendors, roles, labor markets, invitation timing, and candidate seniority differ too much. Report your own funnel instead of borrowing a headline percentage.

Measure invited candidates, link opens, consent or notice views, practice starts, first scored answer starts, submissions, technical failures, alternative requests, and withdrawals. Completion rate is submitted interviews divided by valid invitations. Also compare completion with progression and acceptance. A high completion rate can coexist with a poor experience if candidates feel they have no choice.

The format has a real acceptance problem. A 2025 study spanning one field study and three controlled experiments found lower psychological comfort and lower job-acceptance intentions after asynchronous video interviews than after face-to-face interviews. Earlier research also found that explaining standardization and scheduling flexibility can improve perceived fairness and usability. Explanation helps, but it does not turn a recording into a relationship.

Typical reasons for exit are qualitative rather than mysterious:

• The candidate does not know who will watch the recording or how long it will be stored.
• The step appears before any human contact or useful information about the role.
• The expected effort is hidden until the link opens.
• The camera fails, upload stalls, or mobile use is awkward.
• The questions feel generic, invasive, or impossible to clarify.
• The candidate cannot ask whether the role, schedule, or compensation fits.

Fix the invitation before changing the reminder cadence. Name a recruiter. Show the total effort. Explain the decision logic. Offer accessibility support and a non-video path. Give a response date. Then ask non-completers one optional, neutral question about why they stopped. Do not label them unmotivated.

When is voice better than video?

Use video only when visual demonstration is genuinely job-related. Examples might include presenting a physical portfolio or explaining work on a visible artifact. “We want to see personality” is not a sufficiently precise criterion. Personality is not a camera angle, and visual polish can contaminate judgments about capability.

Voice is often the better screening channel when the evidence is verbal and the role does not require on-camera performance. It reduces preparation around appearance and background. It works more naturally on a phone. A conversational voice system can also clarify responses and ask relevant follow-ups. It still needs structure, disclosure, accessibility, security, and human oversight.

The practical choice is not “modern video versus old telephone.” Compare the evidence needed, candidate burden, accessibility, interaction, and review method. This voice, video, and phone screening comparison provides a format-level decision. Sprad’s voice interview workflow is one option for teams that need structured spoken screening without a camera; live phone screens and specialist video platforms remain valid alternatives.

What legal checks apply in Germany, Austria, and Switzerland?

Recorded interviews contain personal data and often reveal more than the answer itself. Faces, voices, surroundings, health indicators, religion, and family circumstances may become visible or inferable. Do not collect that extra information merely because a platform can store it. Legal review must cover the complete workflow, including hosting, subprocessors, transfers, retention, deletion, ATS copies, access logs, AI analysis, and the decision made from the result.

Across DACH, give candidates a plain-language notice before recording. State the controller, purposes, legal basis, recipients, storage period or criteria, transfers, rights, and whether automated analysis affects progression. Separate optional consent from necessary processing. Set a deletion event, not “keep until someone remembers.” Restrict downloads and prevent reviewers from forwarding clips through chat or email.

Do not use emotion recognition, facial analysis, or voice characteristics as shortcuts to suitability. Check the EU AI Act, GDPR, national employment law, anti-discrimination rules, and collective participation for the specific deployment. This article is an operational checklist, not legal advice.

How should a team pilot the format before rollout?

Run a narrow pilot for one role family. Keep the existing screen as a comparison or offer candidates a genuine choice. Define success before launch. Time saved by recruiters is only one outcome.

• Candidate outcome: completion, alternative requests, technical failures, complaints, and post-step feedback.
• Selection outcome: reviewer agreement, progression, false-negative review, and job-relevant evidence per question.
• Operational outcome: review time, response time, support load, ATS errors, and deletion compliance.
• Fairness outcome: group differences where measurement is lawful and sample sizes are meaningful.

Set stop conditions. Pause if recordings are inaccessible, candidates cannot obtain an alternative, reviewers score irrelevant visual cues, or deletion does not work. Read candidate comments rather than averaging them into a satisfaction score. Research on what candidates think about AI interviews can help frame that review without assuming all candidates want the same thing.

The best outcome may be to retain the structure but change the channel. A scorecard and consistent prompts can improve a live phone screen or an asynchronous voice interview too. The pilot should test the evidence process, not defend the original tool choice.

Frequently asked questions

Is a one way video interview the same as a live video interview?

No. A one-way interview is recorded and reviewed later. A live video interview happens in real time and allows both sides to ask follow-up questions.

Should candidates be allowed to rerecord answers?

Usually, a limited and clearly disclosed retake policy balances technical recovery with consistent administration. Test the rule in your pilot and provide accommodation when needed.

Can AI reject candidates from recorded video?

Solely automated rejection can trigger significant data-protection and AI-governance obligations. Avoid facial or emotion analysis, validate job relevance, explain the process, and retain meaningful human review.

What is a good completion rate?

There is no universal benchmark worth copying. Track valid invitations through submission, segment technical failures and alternative requests, and compare the result with candidate feedback and hiring outcomes.

When should employers avoid asynchronous video?

Avoid it when visual evidence is irrelevant, the applicant pool is scarce or senior, candidates need a two-way conversation, accessibility cannot be assured, or a lower-burden voice or live screen can collect the same evidence.