A one-way video interview alternative, in writing

If you are comparing async video tools, it is worth separating two markets first, because they are usually reviewed as one and they are not.

At the top are the enterprise platforms: HireVue, VidCruiter and the like. Annual contracts, procurement, implementation measured in months, assessment libraries, compliance certifications. Then there is the self-serve tier, Spark Hire, Willo, Hireflix and a long tail behind them: published pricing, sign up on a Tuesday, send a link the same afternoon.

If you are reading this page you are almost certainly shopping in the second tier, and the thing you are actually buying is not video. It is a screening gate: a way to look at sixty candidates without running sixty calls. Video is one way to build that gate. This page is about what it costs and what the written version does differently.

What one-way video genuinely gets right

Skipping this part would make the rest of the page dishonest.

It removes the calendar. This is the whole reason the category exists, and it works. No scheduling, no reschedules, no timezone arithmetic, no half-hour blocks eaten by someone who was never going to pass.

You see and hear the person. For roles where the job is talking, that is not a nice-to-have, it is the thing being assessed. Warmth, pace, whether they can hold a thought under mild pressure, how they sound explaining something to a person who does not already understand it.

It is hard to outsource. Somebody has to sit in front of the camera. That was always a weaker guarantee than it sounded, and it has weakened further, but it is not nothing.

Structure comes free. Everyone answers the same questions in the same order, which beats three phone screens run by three people who each improvised.

So the format is not a mistake. It is a trade, and the trade has a price tag that shows up in two specific places.

Where one-way video stops working

Candidates say no, and the good ones say no first. Recording yourself answering questions into a webcam is an unusual thing to ask of a stranger, and people feel it. Some need a quiet room, decent light and no flatmate walking past. Some hate how they look on camera. Some are at work and cannot exactly step out to film a job application. And a candidate with options does not have to put up with any of it: they have three other processes running that ask for a phone call. You will never see the declines as declines. They arrive as an invite that was opened and never completed.

Reviewing is still real time. This is the part the category tends not to say out loud. Async moved the scheduling problem, not the watching problem. Five questions at two minutes each is ten minutes per candidate, and sixty candidates is a working day of sitting through recordings, most of which will not lead anywhere. So in practice nobody watches all of them. People skim the first thirty seconds and move on, which reintroduces exactly the shallow first-impression filter the structured process was supposed to remove.

It is difficult to compare and impossible to re-read. Two recordings watched an hour apart are compared from memory. There is no scanning back to what someone said about handling a difficult client; you scrub a timeline looking for it. A week later, when the hiring manager asks why you passed on candidate forty, the answer is somewhere in a video file.

It puts appearance in front of judgement, and that is a compliance surface. A video shows age, ethnicity, disability, home, accent and whether the flat has a bookshelf, before it shows a single relevant thing about the work. That is a lot of protected information delivered ahead of the answer. The industry has been retreating from automated analysis of faces and voices for years, and regulators have become steadily less tolerant of inferring what a person is like from how they look while speaking. If you are hiring in the EU, a screening step built on recorded faces is a heavier thing to defend than a screening step built on text.

The anti-cheat argument has expired. For a while, a face on camera was proof that the person answering was the person applying. Video and voice can be generated now, and the tooling is not exotic. Anyone selling you video as an authenticity guarantee is selling last year's guarantee.

One-way video, a phone screen, or a written screen

The last row is deliberate. No screening format can prove authorship any more, and the pages that claim otherwise are the ones to distrust.

What changes when the screen is written

The asynchronous part stays. Candidates get a private link and answer whenever suits them, on any device, with no account and nothing to install. What goes away is the camera, the quiet room and the recording.

More people finish. Typing an answer on a phone at lunch is a smaller ask than filming yourself, and the difference shows up most among the candidates who have other options.

Reviewing stops scaling with headcount. Every set of answers comes back already read and scored against the competencies you attached to each question, with a note explaining what earned each level. You read a report, not a recording, and candidate one hundred and eighty gets the same attention as candidate three.

The score is inspectable. The model reads the answers, but the final number is computed by a fixed rule in code from the per-answer assessments, and every report shows the working. That matters the first time a hiring manager disagrees with a rejection: you can point at the reasoning instead of at a number.

The answers stay readable. Text can be scanned, compared side by side, quoted in a hand-off note and re-read a week later without a scrubbing bar.

Each report says whether the answers name anything checkable. Everyone writes with an assistant now, which is exactly why video was supposed to be the safer format. It is not. What can honestly be read is whether an answer contains things that could be false: a named system, a decision with a cost attached, a specific outcome. Fluent text with nothing specific in it gets flagged as something to ask about, not as an accusation.

You can read a full sample report, questions and answers included, without signing up.

When video is the better choice

Genuinely, and this is not a courtesy paragraph.

If the role is talking, screen by talking. Phone sales, front-line support, customer success, anything client-facing where the first impression is the product. You are hiring for delivery, and delivery does not survive translation into text.

If pronunciation or spoken fluency in a specific language is a real requirement, a written answer cannot tell you. Test the thing you need.

If you are hiring for presenting or teaching, watch someone present.

And if your volume is low, four applications for one role, none of this matters much. Call them.

Moving a video screen to writing

The questions transfer directly. That is most of the work already done.

1. Take the questions from your video script. They are usually already competency-shaped, because someone wrote them to be answered in two minutes each.
2. Attach a competency to each one, so the scoring means something later. A question you cannot map to a competency was there to see how someone looked while answering.
3. Write the private brief: the role, the team, what a strong answer looks like in your context. Candidates never see it.
4. Send it to everyone you would have invited to record, plus the ones you would have cut first. The step is cheap now.
5. Keep a short live call for the shortlist. Written screening is not trying to replace the conversation, only to decide who gets one.

Questions recruiters ask

Is this just a form?

The candidate's side looks like one, deliberately, because that is the part that should stay easy. What is different is everything after submission: the answers come back read and scored, not stacked in a spreadsheet. The longer comparison is on the Google Forms page.

Do candidates need an account or an install?

No. A private link, and they type. No camera, no upload, no software.

Will they answer with an assistant?

Some will, and no tool can tell that text from typed text. That is true of video now too. What the report tells you instead is whether the answers named anything a person could check at interview, which is the question that still has an honest answer.

How long does it take a candidate?

Three to five questions, a few minutes each. Longer than a form, shorter than filming, and it can be done on a phone without finding a quiet room.

Can I keep using my video tool as well?

Yes, and for some roles that is the right setup: written screen first to decide who is worth an hour, video or a live call after. They answer different questions.

What happens to candidate data?

Answers are processed on EU infrastructure, kept only for a period tied to your hiring process, and deleted automatically afterwards. A data processing agreement under Art. 28 GDPR is published and forms part of the Terms. There are no recordings of anyone's face to store, share or delete.

Which one you need

If the job is talking, screen by talking. If the job is thinking, and the recording was only ever a way to avoid sixty phone calls, then the camera was never the point, and it is costing you candidates who had other options.

The deeper version of what each format actually measures is in written pre-screening or one-way video. Pricing, including the free tier, is on the main page, and the volume argument is on the high-volume roles page.