By the Wavefield Research team · Published Oct 1, 2026
Employee survey questions come in five kinds, matched to moments in the employee lifecycle: onboarding check-ins, the engagement census, lightweight pulses, manager feedback, and the exit survey. The 42 questions below are organized that way — by when they fire — because a good question asked at the wrong moment produces nothing.
The vendor banks ranking for this search list 70 to 275 questions in one undifferentiated pile. The working version is smaller and sequenced — and it comes with the two disciplines the piles skip: the anonymity floor that makes honest answers rational, and the rule that you never ask about what you will not act on.
Fires automatically at day 30 and day 90 for every new hire
The 30-day survey predicts first-year attrition better than any engagement census — new hires still compare you to the job they left, and the expectation-versus-reality question is the single highest-signal item.
Once or twice a year, to every employee
Several of these descend from the tradition Gallup's Q12 instrument established — short, concrete items about daily experience rather than abstract loyalty. The full program design — census-versus-pulse cadence, trust mechanics, benchmarks — is its own discipline. Full program guide: employee engagement surveys.
Monthly or quarterly, 3–6 questions, often to a rotating sample
Pulses track change, not levels: the reading that matters is this month against last month, so keep wording identical between waves. The commonest pulse failure is overuse — survey fatigue shows up as silently collapsing response rates.
Annually or semi-annually, reported only above the anonymity floor
These are the questions where anonymity math bites hardest: a team of four cannot receive a manager report without identifying its critics. Set the reporting floor (five is standard), state it in the survey, and honor it without exception.
Within the final two weeks, or just after departure
Departing employees are the highest-signal, lowest-politeness source a company has — but only inside a two-week window while reasons are specific. Exit data earns its keep in aggregate: one exit interview is an anecdote, twenty coded exit surveys are a retention diagnosis.
The engagement set above stands in a long measurement tradition — Gallup's Q12 is the canonical twelve-item instrument, refined over decades, and its lesson transfers to any engagement questionnaire for employees: concrete items about daily experience (“do you have the materials and equipment”) outpredict abstract loyalty items, because people can answer them without deciding what the answer says about them.
Anonymity is arithmetic, not a checkbox. An employee answering a manager-feedback question runs a private calculation: can this answer be traced to me? On a team of four, the honest answer is yes regardless of what the invitation promises. The working standard is confidentiality with a stated reporting floor — no results shown for any group smaller than five — stated in the survey itself and honored even when an executive asks for the smaller cut. Breaking the floor once is permanent: the next survey's open ends arrive empty.
Never ask about what you will not act on. Every question implies a promise. Asking about compensation satisfaction the year after a pay freeze, with no intention of revisiting pay, doesn't produce data — it produces cynicism with a timestamp. The pre-fielding test: for each item, name who will read the result and what they could plausibly do about it. This is also the honest length control, since it cuts the nice-to-know questions that fatigue people into silence.
Satisfaction and engagement are different measures. Employee satisfaction survey questions ask whether people are content — with pay, benefits, workload, environment. Engagement questions ask whether people are invested — meaning, discretionary effort, advocacy. The distinction matters because the two move independently: a well-paid team can be satisfied and checked out, and a mission-driven team can be engaged while furious about the office move. A good census measures both and reads them side by side; a survey that blends them into one “happiness” score cannot tell leadership which lever to pull.
The format mechanics are the same as any survey — consistent scale design, neutral stems per the bad-question catalogue, and open ends that follow the rules in qualitative survey questions. Response-rate expectations run higher than any other audience — 50 to 70 percent and up — and the honest benchmarks are in survey response rates.
Four building blocks: a small set of outcome questions (recommend-as-a-place-to-work, intent to stay), experience questions about daily work (clarity, tools, recognition, manager support), one or two open-ended questions for what the closed items missed, and nothing you are unwilling to act on. The biggest inclusion mistake is asking about things leadership has already decided not to change — every such question spends trust.
The reliable core: Do you have what you need to do your job well? Does your manager give you feedback that helps you improve? What is the one thing that would most improve your experience here? Is anything slowing you down we could fix this week? Specific, actionable questions outperform abstract satisfaction items — 'what slowed you down' produces a fixable list, 'are you satisfied' produces a number.
An engagement census runs 20 to 40 items at most once or twice a year; a pulse runs 3 to 6; onboarding and exit surveys run 8 to 12. Length is a trust signal in employee research: a 70-question annual survey tells people their time is cheap, and the data shows it — completion quality collapses in the back half of long instruments.
Confidential with a stated reporting floor is the honest standard: responses are identifiable to the system (so results can be cut by team and tracked over time) but never reported for groups smaller than the floor, usually five. True anonymity sounds better but prevents the team-level reporting that makes results actionable. What destroys trust is not the model you choose — it is claiming one and operating the other.
The structured instrument an engagement program fields: typically 20 to 40 fixed items covering meaning, clarity, recognition, development, manager relationship, and confidence in direction, plus an outcome measure like likelihood to recommend the company. The questionnaire is the easy half; the program around it — cadence, anonymity floor, and visible action on results — determines whether anyone answers honestly next year.
Related: employee engagement surveys · survey question examples · survey response rates · open-ended survey questions
Brief the agent and it drafts the instrument; results report with base-size discipline, so small teams stay protected automatically. From $99 per study.
Start a study