# Candidate experience survey: a starting questionnaire

Metix AI editorial design. Adapt the questions to one hiring stage, pilot them and keep a wording version. This is an operational template, not a validated research instrument. It works with your own approved survey service. This file sends nothing.

## Set up the invitation

Choose one completed stage and a consistent invitation rule. State whether you invite everyone eligible or a defined sample, who is excluded and why, the delay after the stage ends, and the response window. Avoid telling someone the hiring result through a survey invitation.

Configure and test the actual collector, access permissions and retention before writing a privacy promise. Replace every bracket below. An identity-free questionnaire does not establish that the platform collects no identifiers.

## Introduction to adapt

We are reviewing [stage] of our recruiting process. Your feedback is optional. You may skip any question. We will use responses to review this process, not to score your application. [Describe the actual identity and tracking settings, who can read answers, retention, and a contact for questions.] Please avoid names, contact details or personal information in free text. Responses close on [date and timezone].

Only make commitments your team can honor, including keeping feedback outside hiring decisions. If you offer a separate voluntary contact route, explain that it identifies the person who uses it.

## Common items, version v1

Q1. How clear was the information about what would happen next?

1 Not at all clear; 2 Slightly clear; 3 Moderately clear; 4 Mostly clear; 5 Completely clear. Also offer Not applicable and Prefer not to answer. In the worked CSV, `answered` counts only 1-5 selections and `positive` counts 4 or 5. This threshold is an editorial reporting choice, not a benchmark. Keep the wording, options and threshold unchanged when comparing Q1 v1.

Q2. How easy was it to get the practical information you needed for this step?

1 Very difficult; 2 Difficult; 3 Neither easy nor difficult; 4 Easy; 5 Very easy. Also offer Not applicable and Prefer not to answer. Keep this item's counts separate from Q1.

Q3. What, if anything, should we change before the next person reaches this step?

Optional free text. Ask respondents to leave out names, contact details and other personal information. Restrict access to original comments; paraphrase themes before sharing. If a comment describes a serious concern, use your organization's appropriate response route.

## Choose stage-specific items

After an application: How easy was it to submit the information requested? Use Q2's ease scale, with Not applicable and Prefer not to answer.

After a screen: Which practical detail about the role still needs an answer? Optional free text; ask for the topic rather than personal details.

After an interview: How much opportunity did you have to explain work relevant to this role? Use a consistent 1-5 scale from None to Enough, with Not applicable and Prefer not to answer. Pilot the labels before use; keep a separate question ID.

After someone withdraws: Is there anything about our process you would like us to review? Optional free text. Do not require an explanation or infer that the process caused the withdrawal.

Use only items that serve the current decision. A candidate who never had an interview should not be asked to rate the interviewer. Q1-Q3 plus an appropriate stage item are a starting set, not a recommended universal question count.

## Before sending

Assign an owner to the survey and each possible follow-up action. Test the invitation and form as a respondent on a phone. Test a skipped question and confirm what your export records. Keep outcome groups separate without asking candidates to repeat sensitive application details. Review small groups and free text for identification risk. Decide how to summarize changes without exposing a respondent. No invitation or follow-up is sent by this kit.