
One-way vs conversational AI interviews, in short: use a one-way interview to screen a large pool on identical questions. Use a conversational AI interview when you need to hear how a candidate reasons, because only a live interviewer can ask "why?" a second time.
Both formats replace the first-round phone screen. Both run without a recruiter on the call, and both give you a scored report. The difference is what happens after a candidate finishes an answer. In a one-way interview, the next question appears. In a conversational interview, the AI listens and decides what to ask next. Since each format serves a specific purpose, Equip offers both on one platform, so you can pick the right one for each round.
That one difference changes your cost per candidate, the candidate experience, how easy the round is to game, and the kind of signal you get back. This guide covers each format, what research says about how candidates react, and a four-question framework for choosing. If you are new to the topic, start with how AI interviews work.
What is a one-way interview?
A one-way interview gives every candidate the same fixed set of questions to answer on their own time, with no interviewer present. You will also see it called a one-way video interview. Depending on the tool, answers can be recorded video, audio, typed text, or multiple choice.
You and the candidate never need to be online at the same moment. You send a link. The candidate completes it at 11 pm or on a Sunday. The AI scores each answer against the criteria you wrote, and you open a ranked list the next morning. Try a sample one-way interview to see it from the candidate's side.
Where it is strong: scale, comparability, and flexibility. A thousand candidates can take the same interview in the same week, and every one of them answers the same questions in the same order.
Where it is weak: there are no follow-ups. A vague answer stays vague. Candidates can also rehearse, research, or get help between questions, so polished answers do not always mean deep understanding.
What is a conversational AI interview?
A conversational AI interview is a live, spoken conversation between the candidate and an AI interviewer. Instead of a script, you give the interviewer topics, often drafted from the job description. It opens with a question, listens to the answer, and follows up: asking for a specific example, probing a claim that sounded thin, or raising the difficulty when an answer is strong. Try a sample conversational interview to hear how the follow-ups work.
Where it is strong: depth. Follow-ups show whether a candidate understands their own answer. The format is also harder to game, because a memorized or AI-generated response tends to fall apart when the next question builds on it.
Where it is weak: the candidate needs a quiet place, a working microphone, and 15 to 30 uninterrupted minutes. It usually costs more per candidate. And because follow-ups depend on each answer, no two conversations are word-for-word identical, even when the topics and scoring rubric are the same.
One-way vs conversational: side by side
Here is how the two formats compare on the factors that matter most when you design a screening round.
One-way interview
Fixed questions, recorded answers
Answers: video, audio, text or multiple choice
Follow-ups: none
Consistency: identical questions for every candidate
Length: set by your question count
Cheating risk: higher; candidates can rehearse or get help between questions
Best for: first-round screening of large pools
Conversational AI interview
Live voice conversation that adapts
Answers: spoken, sometimes with written or code answers alongside
Follow-ups: based on each answer
Consistency: same topics and rubric, different follow-ups
Length: typically 15 to 30 minutes
Cheating risk: lower; scripted answers break under follow-up questions
Best for: shortlists, second rounds, roles that need reasoning
Tools differ in how far they take each format. On Equip, for example, a one-way interview can hold up to 40 questions and run in around 100 languages, while conversational interviews come in a Quick format (about 15 minutes) and a Comprehensive format (about 30 minutes). The help center comparison lists the full setup details.
Two formats. One platform.
Screen volume with one-way interviews, then go deeper with conversational AI interviews. Pick the format for each round.
Book a demoAI interviews
Every first round
One-way for volume
Conversational for depth
Scored reports
Built-in proctoring
When to use a one-way interview
Choose a one-way interview when breadth matters more than depth. Five situations make it the stronger choice.
1. You have a large pool to screen. A useful threshold: once a role has 30 or more candidates past CV screening, a fixed-question round usually saves more time than it costs. Below that, you can afford to talk to people. This is why one-way rounds are a staple of volume hiring and campus drives.
2. You need a clean comparison. When 300 final-year students answer the same six questions, you can rank them on equal terms. Nobody got an easier follow-up or a friendlier interviewer.
3. The answers you need are knowable. Notice period, shift availability, location, eligibility, and short knowledge checks do not need a conversation. Multiple-choice items handle hard filters; one or two open questions add context.
4. You want to see how someone communicates on camera. For sales, support, and other customer-facing roles, a two-minute video answer shows how a candidate explains things, which a CV cannot.
5. Candidates are spread across time zones or languages. No scheduling means no back-and-forth across regions, and answers in many languages can be scored against one rubric.
For question ideas and setup steps, see this guide on how to conduct one-way video interviews. If you would rather watch and grade answers yourself instead of using AI scoring, check out video interviews.
When to use a conversational AI interview
Choose a conversational interview when you want the closest thing to having a recruiter run the screen. Because the AI follows up on each answer, it feels like a real conversation to the candidate, and it reaches reasoning that a fixed question cannot.
1. The role depends on reasoning. Ask a product analyst how they would investigate a 12% drop in signups, and most strong candidates will recite a sensible framework. The follow-up separates them: "Which of those would you check first, and why?" A one-way interview never gets to ask it.
2. You are evaluating a shortlist. Before a hiring manager spends an hour with someone, a 30-minute conversational round can test depth on the three or four skills that matter most.
3. Rehearsed or AI-assisted answers are a real risk. In remote technical hiring especially, a one-way answer can be drafted with ChatGPT in another window (unless you use Equip's enhanced proctoring). A follow-up that builds on the candidate's own words is much harder to fake.
4. Candidates do best speaking their own language. For frontline and regional roles, a spoken interview in Hindi, Tamil, or Kannada can show what a typed English answer hides. Equip's conversational interviewer runs in 15 languages, including several regional ones; check language coverage for whichever tool you use.
5. You want a phone-screen feel without the phone screen. A short, focused conversational round (around 15 minutes, three to five topics) gives candidates a back-and-forth experience while keeping your recruiters off the calls. For a fuller comparison, read conversational AI interviews vs traditional phone screens.
What research says about candidate reactions
Candidate experience is the most common objection to one-way interviews, and it matters for more than goodwill. A meta-analysis by Hausknecht, Day and Thomas (2004, Personnel Psychology) linked how applicants perceive a selection process to how attractive they find the employer and whether they would recommend it to others. The research on one-way interviews supports some of the concern, and shows the problem is largely about design.
- Reactions start lower. Langer, König and Krause (2017, International Journal of Selection and Assessment) found applicants rated asynchronous digital interviews less favorably than videoconference interviews. Later reviews point to the lack of social presence: there is nobody on the other side.
- Design choices move the needle. Tilston, Krings, Roulin, Bourdage and Fetzer (2024, Human Resource Management) found that giving candidates more preparation time and letting them re-record answers was linked to more positive reactions, while adding more questions was linked to more negative ones.
- Polish is not where your effort pays off. Many teams record themselves asking each question to make a one-way round feel more personal. Niemitz, Rietheimer, König and Langer (2024, International Journal of Selection and Assessment) tested exactly this in a real hiring process and found video questions did no better than plain text. Your time is better spent cutting questions and allowing retakes than filming yourself.
The takeaway: a one-way interview works best when it is short and forgiving. Keep the question count low, give preparation time, and allow retakes on spoken answers (Equip, for instance, lets you allow up to five). If you find yourself writing a long, probing one-way interview, that is a sign you need a conversational round instead.
Research on live AI interviewers is newer, but the early evidence is encouraging. In a field experiment covering about 70,000 applicants for real jobs, economists Brian Jabarian and Luca Henkel found that candidates interviewed by an AI voice agent were 12% more likely to receive a job offer and 18% more likely to start the job, and they stayed in the role longer. When given the choice, 78% of applicants picked the AI interviewer over a human recruiter (Voice AI in Firms: A Natural Field Experiment on Automated Job Interviews). Equip has seen the same shift: as AI interviews become mainstream, candidate willingness to attempt them has risen 50% since launch.
Candidate comfort is no longer the barrier it was. The open question is quality, so judge conversational tools on what you can verify yourself: whether every score comes with a written reason, and whether a recruiter can review and override it.
How to choose: four questions
Run each role through these four questions. If every answer points the same way, you have your format. If they split, use both.
Using both formats in one funnel
For most high-volume roles, the best answer is not one or the other. It is a short one-way round for breadth, followed by a conversational round for depth. You spend the more expensive format only on people who have already cleared the basics.
Here is how that might look for a customer support role with 600 applicants:
600
Applicants
200
Pass CV screening
40
Clear the one-way round
10
Clear the conversational round and meet the hiring manager
One-way round (about 10 minutes): three multiple-choice questions on shift availability, language, and location; one typed answer on why the role appeals; two short video answers on handling an upset customer. It filters for eligibility and baseline communication.
Conversational round (about 15 minutes): three topics on de-escalation, product troubleshooting, and handling a policy the candidate disagrees with. Follow-ups ask what they would say next when the customer pushes back. It filters for judgment under pressure, which a recorded answer cannot show.
Your recruiters review two short reports per finalist instead of making 200 phone calls. This guide to AI interviews shows where both rounds sit relative to skills assessments.
Mistakes to avoid
- Running a deep second round as a one-way interview. Twenty open questions with no follow-ups tire candidates and still miss the reasoning you wanted.
- Using a conversational interview as an eligibility filter. You do not need an adaptive interviewer to ask about notice period.
- Writing vague evaluation criteria. Both formats score against what you tell the AI to look for. "Good communication" produces noisy scores; "explains the fix in steps a non-technical customer could follow" does not.
- Skipping the explanation for candidates. Tell them the round uses AI, how long it takes, and that a person reviews every result.
- Letting the score make the decision. The AI recommends; your team decides, and can override any score.
FAQ
Which format is harder to cheat on?
Conversational. In a one-way interview, a candidate can look up or generate an answer between questions. In a conversational interview, follow-ups build on what the candidate just said, so a read-out answer is easy to spot. Proctoring helps with both formats.
How long should each interview be?
Keep a one-way round short; research links more questions to worse candidate reactions. For conversational rounds, about 15 minutes suits a first screen and about 30 minutes suits a second round or senior role.
Do candidates prefer one format over the other?
Studies show candidates react less favorably to one-way video interviews than to live formats, largely because nobody is on the other side. A conversational interview is closer to talking with a person. Either way, clear instructions, a reasonable length, and a flexible deadline do the most for completion rates.
Can you use both for the same role?
Yes, and for high-volume roles you usually should: one-way for the wide first filter, conversational for the shortlist.
What does each format cost?
One-way interviews are usually cheaper per candidate, which is why they suit the top of the funnel. On Equip, one-way AI interviews start at $1 per candidate, and conversational AI interviews cost a little more per candidate.
Try both before you decide
The fastest way to choose is to sit on the candidate's side of each format. Take a sample one-way interview and a sample conversational interview, then ask which one would have told you more about the last person you hired for the role. That answer is usually your format.