Candidate experience in volume hiring: measure, then fix
Candidates rate our AI interview 3.64 out of 5. That’s from 1,443 ratings collected across our customers’ hiring pipelines in India between March and August 2026, and it’s our own measurement, not a survey someone else ran. It isn’t a number we’d put on a billboard. It’s a number we use, because it sits next to two others from the same period: 39.4 percent of self-booked interviews were scheduled outside 9-to-6, with 9 PM the most popular hour, and 21 percent of booked interviews never started.
Put those three together and you get a fairly honest picture of candidate experience in volume hiring. Candidates will meet you more than halfway when you let them choose the time. A meaningful share still don’t show up. And the ones who finish rate it as decent, not delightful. Every team that hires at volume has a version of this picture. Most don’t measure it, so they can’t tell which part to fix.
What is candidate experience?
Candidate experience is how applicants perceive every step of your hiring process, from the job post through the interview to the decision and what happens after. It is shaped mostly by speed, clarity and fairness, and in volume hiring it is decided mainly by how you treat the candidates you reject, because they are the majority.
That last clause is the one I’d underline. If you hire 50 people from 2,000 applicants, 1,950 people’s experience of your company is the rejection path. The CandE benchmark program, now run by Survale, builds its candidate experience score with 80 percent of the weight on candidates who weren’t hired, for exactly that reason.
Why it matters more in high-volume hiring
Three reasons, and only one of them is about employer brand.
Your candidates are often your customers. In retail, banking, telecom, food delivery and edtech, the people applying for frontline roles are often the same people who buy from you. A careless rejection to 1,950 people is a customer communication.
Volume roles re-hire constantly. Support, sales and operations roles open again in three months. A candidate who had a decent experience and narrowly missed is your best applicant next quarter. A candidate who was ghosted won’t apply again.
Resentment is measurable and rising. Survale’s 2025 research, based on more than 66,000 candidate surveys, recorded candidate resentment at record highs: 15 percent in North America and 17 percent in EMEA. The top negative theme in candidate comments was lack of communication, ghosting and no follow-up. The same research found that 56 percent of candidates rejected during screening or interviews received no feedback at all.
None of that is unique to the US or Europe. In Indian campus and volume hiring, the scale is larger and the silence is often longer. We wrote separately about the campus applicants you never reply to, which is the most common candidate experience failure we see.
The six moments that decide candidate experience
| Moment | What the candidate wants | What usually goes wrong |
|---|---|---|
| Applying | A short form and a clear job | Long forms, vague job descriptions, no confirmation |
| Waiting after applying | To know they’re in a queue and roughly how long it is | Weeks of silence |
| Getting invited | A clear invite, a choice of time, no app to install | A fixed slot during work or college hours, or a link that doesn’t work on a phone |
| The interview | Relevant questions, a fair hearing, knowing how it works | Questions unrelated to the job, an interviewer who’s late or distracted, no explanation of the format |
| The decision | A decision on the date promised, ideally with a reason | No decision at all |
| After | Respect, and a sense that applying again is worth it | A no-reply email, or nothing |
Notice how much of the table is about time and clarity rather than warmth. Candidates in volume hiring rarely expect a personal touch. They expect not to be left guessing.
What to measure: six candidate experience metrics
You don’t need a new tool to start. You need six numbers, tracked by stage:
- Completion rate per stage. Of candidates invited to a stage, how many finish it? A sharp drop at one stage points at that stage.
- Days in each stage. Especially the wait between applying and the first contact, and between the final round and the decision.
- No-start rate. Of booked interviews, how many never start? Ours is 21 percent; if yours is far higher, look at reminders and at how far ahead the slot was booked. Our guide to reducing AI interview no-shows covers the fixes.
- Withdrawal reasons. Ask every candidate who withdraws one question: why? Most will tell you.
- Survey score. A simple 1-5 rating, by stage and by outcome (hired versus rejected).
- Feedback given. What share of rejected candidates got any reason at all?
Report them monthly, by role family. A single company-wide number hides everything; the support role and the analyst role will have very different problems.
A five-question candidate experience survey
Send it within a day of the candidate’s last contact with you, to everyone who reached an interview, hired or not:
- Overall, how would you rate your experience with our hiring process? (1-5)
- Did you know what to expect at each step? (Yes / Partly / No)
- Did the interview feel fair and relevant to the job? (Yes / Partly / No)
- Did you hear back when we said you would? (Yes / Later / Not at all)
- What is one thing we should change? (open text)
Five questions is enough. Response rates fall quickly after that, and question 5 will tell you more than ten rating scales. Read every answer to question 5 for the first month; after that, tag them into themes and count.
One caution: survey only after the decision, or rejected candidates will rate the process by the outcome they’re hoping for, and you’ll learn nothing about the process.
What our own numbers taught us
Each of the three numbers from the opening says something specific.
The 39.4 percent outside 9-to-6 says that a fixed daytime interview slot is itself a candidate experience problem. Students are in class, and people with jobs are at work. When candidates can book their own interview slot from an email link, a large share choose the evening, and 9 PM was the single most popular hour. We wrote about this in after-hours screening.
The 21 percent no-start rate says that booking isn’t attendance. Some of it will be candidates who took another offer, and no reminder fixes that. Some of it is forgetting, which reminders can help with. HireQwik sends email reminders before a candidate’s booking deadline, and WhatsApp reminders can be switched on for a job.
The 3.64 out of 5 says the interview itself is acceptable to most candidates but not yet something they’d praise. The things we’ve worked on around it are mostly about clarity: telling candidates exactly what will happen before they start, keeping the interview to 10-20 minutes, and asking questions drawn from the job and their own résumé so the conversation feels relevant. Before an interview starts, every candidate sees this notice: “This interview is monitored with video recording. Your camera must remain on throughout the interview. Switching tabs or turning off your camera will trigger warnings and may lead to automatic termination.” It’s blunt on purpose. A candidate who knows the rules in advance is less likely to feel ambushed by them.
Candidate experience when an AI runs the first round
This deserves its own section, because candidates have strong views. Survale’s 2025 research found that complaints about AI screening and impersonal rejections were among the top three negative themes in candidate comments. At the same time, it found one-way recorded video interviews matched live interviews in perceived fairness for the first time. Candidates don’t object to technology as such. They object to being processed without being told how, and rejected without a person ever looking.
Greenhouse’s 2026 Candidate AI Interview Report, a survey of 2,950 job seekers in five countries, puts numbers on both. Among its 1,200 US respondents, 70 percent said they were never clearly told upfront that AI would be evaluating them, and 38 percent wanted assurance that a human reviews the AI’s evaluation before any decision is made. Both are cheap to fix.
So, if you use AI in the first round:
- Say so in the invite, and say what it is: a short voice interview, roughly how long, what it covers. Our guide to preparing candidates for an AI interview has wording you can borrow.
- Give a human contact. A named recruiter email for problems, and actually answer it.
- Keep a person on the decision. In HireQwik the AI only recommends; the final call is HR’s.
- Close the loop. Rejecting a candidate in HireQwik doesn’t send them an email by itself, so send one. Our post on the candidate rejection email has templates by stage.
- Expect nerves and plan for them. Some candidates find talking to an AI stranger than talking to a person, at least the first time. We’ve covered what helps in nervous candidates in AI interviews.
Candidate experience in campus drives
Campus hiring in India has its own version of every moment in the table above, and two extra parties: the placement cell and the student’s batchmates.
The placement cell is part of the experience. Students hear about your drive through the placement officer, and they judge you partly by how that office talks about you afterwards. A drive where results reach the college on the promised day, with a clear list of who moves forward, earns you a better slot next season. A drive where the college has to chase you for a week does the opposite. We’ve written about sending a placement cell results report that answers the questions placement officers actually ask.
Batchmates compare notes in real time. When 300 students from one college take your interview in the same week, every inconsistency is noticed within hours: one student was asked about their final-year project and another wasn’t, one heard back and their friend didn’t. Consistency matters more here than anywhere else. The same core questions for everyone in the role, the same decision date for everyone in the batch.
Students can’t interview at 11 AM on a weekday. They’re in class or in exams. This is where self-booked evening slots matter most, and it’s one reason the 9 PM figure above isn’t a curiosity. If your campus process only offers interviews during college hours, you’re asking students to skip lectures to talk to you.
The offer-to-joining gap is long. Campus offers often go out months before joining. Candidate experience doesn’t stop at the offer: a short check-in every few weeks, and a clear joining date, keeps students from accepting the next offer that comes along. For drives that span many colleges, shortlisting across colleges on one consistent bar is the fairest way to keep the experience the same for everyone.
Fixes ranked by effort
| Fix | Effort | What it changes |
|---|---|---|
| Tell every candidate the date they’ll hear back, and keep it | Very low | Removes the worst part of waiting |
| Send a rejection to every candidate who interviewed | Low | Ends silent rejections |
| Let candidates choose their own interview time | Low to medium | Evening and weekend availability, fewer reschedules |
| Add reminders before the interview, on email and WhatsApp | Low | Fewer no-starts from forgetting |
| Explain the format in the invite (length, type, what’s checked) | Very low | Fewer surprises, fewer abandoned interviews |
| Cut one interview round | Medium | Faster decisions, fewer withdrawals; see our post on interview process stages |
| Give a one-line reason with each rejection | Medium | Turns a rejection into something useful |
| Survey candidates and act on question 5 | Medium | Tells you what to fix next |
Start at the top. The first two rows cost almost nothing and fix the failures candidates complain about most.
Five candidate experience mistakes
- Surveying only hired candidates. They’re the happiest and the least informative. Survey everyone who reached an interview.
- Treating it as a branding project. Candidate experience is an operations problem: speed, clarity, follow-through. Fix the process and the brand follows.
- One number for the whole company. Measure by role family and stage, or you won’t know where to look.
- Fixing the interview and ignoring the wait. The waiting is usually worse than the interview.
- Promising feedback you won’t give. If you can’t give every rejected candidate a reason, don’t promise one. Promise a decision by a date instead, and keep that promise.
The quickest way to judge our candidate side is to be the candidate: sit the live AI interview yourself and see whether you’d rate it above 3.64. If you’d rather see it from the recruiter’s chair, book a demo and we’ll walk from invite to result.
Frequently asked questions
How do you measure candidate experience?
Combine behaviour and opinion. Behaviour: completion rate at each stage, days spent waiting in each stage, the share of booked interviews that never start, and withdrawal reasons. Opinion: a short survey sent to every candidate who reached an interview, including the ones you rejected, with a rating question and one open question. Track both by stage so you can see where the experience breaks.
What questions should a candidate experience survey ask?
Keep it to five: how would you rate your experience overall (1-5), did you know what to expect at each step, did the interview feel fair and relevant to the job, how quickly did you hear back compared with what you were told, and what is one thing we should change. Send it within a day of the candidate's last contact, to hired and rejected candidates alike.
Does AI screening hurt candidate experience?
It can, if candidates are not told how it works, cannot reach a person, or are rejected with no explanation. Done well, it can help: candidates pick their own time, including evenings, and get a consistent interview. In large candidate surveys, complaints about AI screening are among the top negative themes, so disclosure, a human contact and a person making the final decision matter more than the technology.
See your own candidates screened
Book a 30-minute demo. Bring a live JD and we'll screen against it, then start with a pilot on your own candidates before committing to anything.
Existing customer? Sign in