Candidate experience survey questions, by stage of the process

A candidate experience question bank organized by stage lets you check any part of the hiring process on its own, rather than waiting until the very end to find out where it broke down. Each question below is written for one part of the journey — application and scheduling, interview day, communication, and the decision itself — with the rating scale to use and what a low score on that specific question usually points to.

If you want three complete, ready-to-send surveys keyed to hired, rejected, or withdrew, use post-interview candidate survey template instead. This page is for building your own survey from parts, or for checking one stage without waiting for the whole process to finish. For the process audit that usually comes before any of this — response-time promises, interview-day basics, access — see how to improve candidate experience.

Pick one rating scale and keep it

Use a 1-to-5 agreement scale ("strongly disagree" to "strongly agree") for most questions below — it is familiar to candidates and easy to average. Reserve a 0-to-10 scale for a single likelihood-to-recommend question, since that scale is what turns into a candidate NPS score; see candidate NPS for recruiters for the calculation. Do not mix the two scales in one survey — a candidate who just adjusted to a 1-to-5 scale and hits a 0-to-10 question two lines later often answers it like the first scale, which quietly corrupts the data.

The question bank

Copy the rows relevant to what you want to check. Each includes the scale to use and what a low score usually means.

APPLICATION AND SCHEDULING
Q: The application process was straightforward.               [1-5]
   Low score usually means: form too long, unclear requirements,
   or a broken application link.
Q: I was able to schedule my interview without back-and-forth.  [1-5]
   Low score usually means: scheduling tool friction, or too few
   time slots offered.
Q: I knew what to expect before the interview (format, length,
   who I'd meet).                                                [1-5]
   Low score usually means: the confirmation email or briefing
   call skipped key details.

INTERVIEW DAY
Q: The interview started on time.                                [1-5]
   Low score usually means: interviewer scheduling discipline,
   or back-to-back panels running long.
Q: The interviewer(s) seemed prepared and familiar with my
   background.                                                   [1-5]
   Low score usually means: no briefing, or the interviewer opened
   the resume for the first time on the call.
Q: The questions felt relevant to the role I applied for.        [1-5]
   Low score usually means: outdated interview guide, or a panel
   asking questions unrelated to the actual job.
Q: I had a chance to ask my own questions.                       [1-5]
   Low score usually means: interview ran long on the interviewer's
   side, leaving no time for the candidate's questions.

COMMUNICATION
Q: I was told what to expect for next steps and timing.          [1-5]
   Low score usually means: no close-out at the end of each
   interview stage.
Q: Updates came when promised, or I was told about a delay.      [1-5]
   Low score usually means: broken response-time promise; see the
   candidate experience audit for the fix.
Q: However this ended, I was told the outcome directly rather
   than left to guess.                                           [1-5]
   Low score usually means: candidate ghosted; check your rejection
   process for gaps.

THE DECISION
Q: The reason for the outcome was clear enough, even if I
   disagreed with it.                                             [1-5]
   Low score usually means: a generic rejection template with no
   specific, honest reason.
Q: Based on this process alone, how likely are you to recommend
   [Company] to a friend as a place to interview?         [0-10, NPS]
   See candidate-nps-for-recruiters for scoring this one.

OPEN TEXT (use one, not both, to protect response rate)
Q: What is the one thing we should change about this process?
Q: What is the one thing we did well that we should keep doing?

Building a short pulse check from the bank

Not every check needs a full survey. To test whether scheduling friction is a real problem before building a whole survey around it, send just the two scheduling questions above to candidates who scheduled in the last two weeks. A one- or two-question pulse check gets a far higher response rate than an eight-question survey, and it answers a specific question fast enough to act on this month rather than next quarter.

A worked scoring example

An invented batch of 40 responses to the interview-day section, one quarter:

Read across the rows, not just down the average. Here, two mediocre scores tell one story: interviews are running long, which delays the next candidate's start time and eats into the current candidate's chance to ask questions. The fix is a scheduling buffer and a time check at the midpoint of each interview, not four separate initiatives for four separate scores.

Segmenting results before you act on them

An overall average across every candidate hides where the problem actually is. Before treating a score as a company-wide issue, split it by at least one of: department, interview format (panel vs one-on-one), interviewer, or outcome (hired vs rejected). A 3.1 average on "started on time" that is really a 2.0 for one department's panels and a 4.5 everywhere else is a scheduling problem for one team, not a company-wide training issue.

Choosing which stage to survey when you cannot survey everything

Surveying every stage for every candidate produces fatigue and a falling response rate. If you can only run one, prioritize by where you already suspect the biggest drop-off or the loudest informal complaints — a recruiter hearing "nobody told me what happened" repeatedly from candidates is a stronger signal to survey communication first than a hunch that interview day might be running long. Start with the stage where you have the least data and the most anecdotal concern, run it for one full cycle, fix what it finds, then move to the next stage rather than trying to measure everything at once.

Common mistakes

• Asking candidates to explain the rejection decision. "The reason for the outcome was clear enough" is a fair question; "do you think you deserved to be rejected" is not one you should ask.
• Changing question wording between cycles. Breaks the ability to compare a score against last quarter's, which is usually the more useful comparison than any single number on its own.
• Sending the full eight-question survey for a quick check. Use the pulse-check approach above instead; a long survey for a small question depresses response rates for no benefit.
• Reporting an average without the sample size. A 4.5 average from three responses is not the same claim as a 4.5 average from three hundred; always state the count.
• Never closing the loop on what changed. If interview-day scores improve after adding a scheduling buffer, say so in the next report. It is also what earns the response rate for the survey after that.

Timing each stage's check

Sending any of these more than a few days late does not just lower the response rate, it changes the answers you get. A candidate asked about interview-day pacing a week later tends to answer based on their overall impression of the company rather than the specific hour they spent in the room, which is a different, less useful signal.

Where this fits with a full candidate experience program

This question bank is one input into a broader effort, not the whole thing. Pair it with the stage-by-stage audit in how to improve candidate experience, use the complete ready-to-send versions in post-interview candidate survey template when you want the full end-of-process form, and track the single recommend-a-friend question over time using the method in candidate NPS for recruiters if you want one number to report to leadership alongside the detail.