What HireVue interviews look like (and what small businesses can learn)

Search “HireVue interview” and almost everything you’ll find is written for the other side of the camera: coaching candidates to rehearse, beat the timer, and game the format. There’s a whole cottage industry teaching people to outperform your screening step.

That should tell you two things if you’re thinking about one-way interviews for your own hiring. First, the format is common enough at big employers that candidates treat preparing for it as a standard job-search skill. Second, candidates approach it with enough suspicion that they’d rather study workarounds than just answer the questions.

Both facts matter, because the format itself is genuinely useful. HireVue built the enterprise version of the one-way interview, and enterprises adopted it for reasons that apply even more strongly to a small business with no recruiter. The trick is separating what the format does well from what the enterprise implementation does to candidates, and from where the enterprise version is headed next. This post walks through all three, and what I’d copy or avoid if I were designing a one-way interview for a 5 to 100 person company.

The mechanics: what a HireVue interview looks like

Strip away the branding and the format is simple. A candidate gets a link and completes the interview alone, on their own time, before a deadline.

The standard shape, per HireVue’s own candidate guidance:

• 5 to 8 questions, fixed in advance, identical for every candidate. HireVue’s FAQ puts it plainly: “Most video-based assessments include between five and eight questions.”
• About 30 seconds to review each question, then up to 3 minutes to record the answer.
• Under 20 to 30 minutes to complete, though HireVue tells candidates to set aside up to 45 minutes as a buffer.
• One practice question you can redo as many times as you want before anything is scored. HireVue’s own guide says candidates can “do several of these practices until you feel comfortable to continue.” Real questions don’t get that same unlimited do-over.

The questions themselves are unremarkable on purpose. Openers like “tell me about yourself” and “why this role.” A block of behavioral prompts (“tell me about a time you faced a challenging situation,” “describe a time you failed”). A few day-to-day operating questions about prioritization and organization. Logistics like salary expectations at the end. If you’ve run a first-round phone screen, you’ve asked most of them.

What’s different is what happens after. HireVue’s own FAQ describes the assessment as evaluating “both what you know about performing the job (‘hard skills’ or competencies) and what recruiters call ‘soft’ competencies, like communication skills, conscientiousness, problem-solving skills, team orientation, and initiative.” At enterprise scale, responses are transcribed and scored with AI assistance against those criteria, and reviewers work from the scores rather than watching every minute of every recording. HireVue frames the whole thing as a replacement for “the typical 20-minute screening phone call with a recruiter,” which is exactly the piece that lets one team screen candidates by the thousands.

What the enterprise format gets right

It’s fashionable to dunk on this format. Before doing any of that, it’s worth being honest about why serious companies with serious legal exposure standardized on it.

Every candidate answers the same questions. This is the big one. In unstructured phone screens, you drift. You chat longer with people you like, ask different follow-ups, and end up comparing your memory of ten different conversations. Fixed questions in a fixed order make answers directly comparable, which is both fairer to candidates and more defensible if a decision is ever challenged. Schmidt and Hunter’s 1998 meta-analysis put the predictive validity of structured interviews at around 0.51 for job performance, versus about 0.38 for unstructured ones. That gap, roughly 30% more predictive, is the difference between an interview that adds real signal and one that mostly adds noise, and this format enforces the structure mechanically.

Nobody schedules anything. The first conversation happens without calendars. For a candidate working a job, answering at 9pm beats sneaking out for a phone screen. For the employer, thirty first interviews happen in parallel instead of eating a week.

The evidence persists. A phone screen exists only in your notes. A recorded answer can be reviewed, compared side by side with another candidate’s answer to the same question, and shown to a co-owner or manager before a decision.

Rehearsal isn’t actually a bug. Candidates who prepare are showing you prepared communication, which is what most jobs reward anyway. A format everyone can prepare for is more level than one that rewards whoever happens to click with you on the phone.

If you hire without a recruiter, notice that these advantages aren’t enterprise advantages. Comparability, zero scheduling, and reviewable evidence are worth proportionally more when screening happens in the gaps between running a business.

Where it earns the bad reputation

So why does the format have such a hostile fan base? Mostly because of how it gets deployed at scale, not because of what it is.

It feels like a void. At a large employer, a candidate records twenty minutes of answers and frequently hears nothing back. No name, no timeline, sometimes no rejection. The format removes the human from the first interview, and the process around it often fails to put one back anywhere.

The scoring has been genuinely opaque, and HireVue’s own history shows why that’s dangerous. For years, HireVue scored facial expressions alongside word choice and speech patterns. In January 2021, after a third-party algorithmic audit and mounting criticism from privacy researchers, HireVue announced it was dropping facial-expression analysis entirely. The company’s own numbers showed why: facial data contributed only about 0.25% to the model’s accuracy in most cases, and around 4% even for customer-facing roles, nowhere near enough to justify the bias risk or the backlash. CEO Kevin Parker put it plainly: “When you put that in the context of the concerns people were having [about potential bias], it wasn’t worth the incremental value.” That’s not a hypothetical warning. It’s the market leader building a scoring signal, running it for years, then discovering its own audit couldn’t justify keeping it. Whatever tooling you use, “the AI decided” is not a position you want to be in with candidates or with the law.

The ask is often disproportionate. Eight questions at up to three minutes each is a lot to ask before any human contact, especially for hourly roles. And that effort only pays off if someone actually reviews it. Ask for twenty minutes of a candidate’s time and let the response sit unwatched, and you’ve wasted your own time right along with theirs. When the effort demanded exceeds the attention it gets in return, completion drops and resentment fills the gap.

Generic questions invite generic answers. When every large company asks the same rehearsed prompts, candidates supply the same rehearsed responses, increasingly drafted by ChatGPT. A question list that hasn’t been tailored to the actual job measures preparation more than suitability.

Notice that none of these are inherent to one-way interviews. They’re consequences of running the format at maximum scale with minimum human contact, plus one company’s real, documented mistake with an opaque scoring signal. A small business doesn’t have to inherit any of them.

Where HireVue is headed next

The reputation problem above is the one-way interview at its most conservative: fixed questions, a timer, a recording, a human (eventually) reviewing it. It’s worth knowing where the format goes from here, because HireVue itself just showed its hand.

In March 2026, HireVue acquired the technology behind Hireguide, a startup that had built a voice-based AI interviewer. Instead of a candidate recording an answer to a fixed prompt, Hireguide’s agent holds a live, conversational interview on its own and asks follow-up questions in real time. HireVue’s CEO described the move as going “from simple automation to a smarter way to qualify candidate capability at every touchpoint.” Hireguide’s founder called it bringing “agentic-led hiring into a new era.”

Read that plainly: the next step for the enterprise leader in this category isn’t better scoring of recorded answers. It’s an AI agent that conducts the interview itself, no human asking the questions at all.

That’s worth pausing on, because HireVue’s own research shows the trust hasn’t caught up to the ambition. The company’s 2026 Global AI in Hiring Report, a survey of more than 3,100 hiring managers, found that only 41% fully trust AI in their own hiring process. That’s the vendor selling the AI, reporting on what its own buyers actually believe.

A small business doesn’t need to chase that trajectory to get what makes the one-way format work. You can keep a human writing the questions, a human skimming the answers, and a human making the call, and still get the same comparability and time savings HireVue built its business on in the first place. Enterprise screening is heading toward less human involvement at every step. Yours doesn’t have to follow it to get the benefit.

What to copy for your own one-way interview

If you’re designing a one-way interview for your own hiring, take the structure and leave the scale defaults, including the newest one.

Copy the fixed question set. Same questions, same order, every candidate. Write them from your role’s actual must-haves, not from a generic list. Three to five questions is enough for a first screen, and behavior-based prompts (“tell me about a time you dealt with an angry customer”) outperform hypotheticals. If you need a starting point, here’s a full set of video interview questions organized by what each one tests, or a broader list of interview questions worth asking candidates at any stage of the process.

Give people real think time, and let them warm up first. Thirty-plus seconds of prep before the clock starts doesn’t dilute the signal. It removes the gotcha factor, which is what candidates hate most about the format. A free-form practice question, the one part of HireVue’s own process candidates can redo as many times as they need, costs you nothing and does the same job for your version.

Copy the timer, loosely. Caps keep answers comparable and review fast. Two to three minutes per answer is plenty. A first screen that takes more than 15 minutes of a candidate’s time is taking too much.

What to do differently

Stay visibly human. This is your structural advantage over any enterprise running this format, agentic AI interviewer included. Record a short intro video yourself. Tell candidates who will watch their answers, and when they’ll hear back. Then actually follow up within a couple of days. A one-way interview from “Sarah, who owns the shop” reads completely differently than one from a Fortune 500 applicant portal, or from an AI voice asking the questions.

Keep the ask proportionate. For hourly and high-volume roles, three questions at 60 to 90 seconds each is a fair first ask. Save the longer format for roles where communication is the job.

Be transparent about scoring, and don’t repeat HireVue’s facial-expression mistake. Tell candidates their answers are reviewed against the role’s requirements, and mean it. Whatever one-way video interview tool you use should show you why a candidate scored the way they did, criterion by criterion, so a human can check the reasoning and own the call, instead of scoring a signal nobody can explain. AI that surfaces evidence is a screening tool. AI that silently decides, or conducts the interview with no one asking the questions on your behalf, is a liability, both legally and to your reputation in a small town or a tight industry.

Don’t make video the whole process. This is where I think the single-feature framing goes wrong entirely. A one-way interview shows you how someone communicates. It says little about credentials or judgment. It works as one layer in a screen, next to a scored resume and, when the role warrants it, an assessment, so weak signal in one layer gets caught by another. Video as the sole gate over-rewards on-camera polish, which matters a lot for some roles and barely at all for others.

The screening layer, sized for a business like yours

Everything in the copy list above is what Truffle is built around, at small business scale instead of enterprise scale. Truffle is an AI screening platform that combines one-way video interviews with resume screening and talent assessments. You write your questions once, candidates record on their own time, and AI transcribes each answer, scores it against your criteria with the reasoning shown, and surfaces the most revealing 30 seconds as a Candidate Short. You review a ranked shortlist in minutes and make the calls yourself.

If you’re comparing tools in this category directly, we keep an honest rundown of HireVue alternatives, including where HireVue remains the right choice (it’s built for enterprise volume and priced like it). For a business hiring a handful of roles at a time, Truffle’s plans start at $49 a month, with a 7-day free trial and 30 credits, no card required.

The format was never the problem

The interesting thing about the anti-HireVue backlash is what it’s actually complaining about. Read the threads and almost nobody objects to recording answers. They object to effort disappearing into a void, to a scoring signal even the company that built it couldn’t defend, to being processed instead of considered.

Those are process failures, and they’re optional. The structure underneath (same questions, own time, evidence you can compare) is the closest thing hiring has to a fairness upgrade, and it’s now cheap enough that a 15-person company can run it better than a 15,000-person one. The enterprises proved the format works, and now they’re proving how much further you can push the automation before your own hiring managers stop trusting it. A small business gets to prove the format can feel human, at exactly the moment the biggest name in it is heading the other way.

If you want to try the structure without inheriting the reputation, or the newest experiment, Truffle’s one-way video interviews come with the transparent, per-question scoring built in from the start, so the follow-up you promise candidates is one you can actually back up.

FAQs about HireVue’s interview format

Does HireVue still score facial expressions?

No. HireVue dropped facial-expression analysis in January 2021 after its own audit found the signal contributed almost nothing to the model’s accuracy. If you’re evaluating any one-way interview tool today, ask what it scores now, not what the category scored five years ago.

Is HireVue built for a small business?

Not really. HireVue prices and implements for enterprise volume, not a handful of open roles at a time. If you’re comparing options sized for a smaller team, we keep an honest HireVue alternatives rundown, including where HireVue is still the right call.

Should I copy HireVue’s actual question list?

Copy the structure, not the questions. HireVue’s prompts are written for corporate roles at enterprise scale, and candidates have spent years learning to rehearse them. Write your own from the role’s actual must-haves. Our video interview questions guide is a starting point organized by what each question tests.

Is HireVue moving away from one-way interviews?

It’s moving toward more AI involvement, not less. Its March 2026 acquisition of Hireguide adds a voice-based AI interviewer that can conduct the conversation itself, rather than just scoring a recorded answer. A small business doesn’t need to follow that trajectory to get the benefits of the format, structure and comparability, not automation for its own sake.