Pulse survey: meaning and application

A comprehensive questionnaire captures one moment well but says little about what changes in the following weeks. A pulse survey fills that gap with a short questionnaire that you repeat at a fixed rhythm. How do you design such a survey, and when does a shift really mean something?

A pulse survey is a short, repeated measurement with a limited number of questions or statements that you use to track how a topic develops over time.

In Dutch it is called a pulsemeting. A pulse survey is quick to repeat and shows whether opinions shift, but does not explain why. It also carries the limitations of repeated measurement: falling response, fatigue, changing internal standards and small numbers per round. Combine it with in-depth research for that reason.

The term pulse survey comes from the practice of employee research, not from a single original publication. Berthelsen and Muhonen describe pulse surveys as short, frequent questionnaires and note a striking lack of research on this practice. In their interview study, pulse surveys were seen as a complement to a more comprehensive annual assessment, not as a replacement (Berthelsen & Muhonen, 2026).

A methodologically related approach is experience sampling. In the experience sampling method (ESM), participants receive a signal at random moments and immediately complete a short form about what they are doing, thinking and feeling (Csikszentmihalyi & Larson, 1987). Ecological momentary assessment (EMA) is the broader term for repeated real-time measurement of experiences and behaviour in daily life, intended to limit recall bias and track change over time and context (Shiffman, Stone & Hufford, 2008).

A pulse survey is less intensive. It usually runs at a rhythm of weeks rather than hours and often asks about the past period rather than the moment.

What you measure directly is self-report: how people respond to a few statements at a given moment. You infer a development by comparing rounds. Why a score rises or falls is not measured. A pulse survey signals; explaining requires other research. Moreover, you only measure what you ask; an open question at the end captures part of what falls outside the statements.

A pulse survey works with few questions, often one statement per theme. That is a design choice with a price: unlike a multi-item scale, a single statement does not let you estimate reliability in the same way or cover a concept broadly.

In a meta-analysis of overall job satisfaction covering 17 studies and 7,682 people, Wanous, Reichers and Hudy found an average uncorrected correlation of .63 between a single question and a full scale. Their estimate of the minimum reliability of such a single question ranged from .45 to .69, depending on the assumptions (Wanous et al., 1997). That applies to a general judgement and does not automatically transfer to broad or composite concepts. In 2022, Allen, Iliescu and Greiff called for more attention to single-item measures in psychology in a short article (Allen et al., 2022).

Use one statement for a narrow, concrete theme and two or three for a theme with several sides.

1. Start with the decision. Which decisions should the results support, and which development must you see for that?
2. Choose themes and statements. Limit yourself to a few themes and write one concrete statement per theme in the language of the target group.
3. Choose a fixed scale. Use the same response scale and labels in every round, for example a five-point agreement scale.
4. Set the rhythm. Match the frequency to how quickly the situation can change and how quickly you can respond. Do not measure more often than you can report back.
5. Decide whom to ask. Survey everyone or a rotating sample; the latter lowers the burden per person but makes each round smaller and less certain. Agree on a minimum group size for results per group, so that individuals cannot be identified.
6. Plan the feedback. Decide in advance who sees the results, when participants hear what happens next and who acts.
7. Keep the questionnaire stable. Do not change wording or scale midway. If something has to change, document the break in the series.

Frequent measurement has a downside. Porter, Whitcomb and Weitzer describe research in which multiple surveys within one academic year significantly suppressed response rates in later surveys (Porter et al., 2004). That research concerned students, but the principle is also relevant to pulse surveys.

Feedback appears to play a role. In Berthelsen and Muhonen's interviews, participants reported that response stayed high mainly when line managers kept discussing the results and acting on them; otherwise it dropped (Berthelsen & Muhonen, 2026). These are experiences from sixteen qualitative interviews, not a measured effect. In a systematic review, Huebner and Zacher point out that the way an organisation handles results can influence attitudes towards future surveys (Huebner & Zacher, 2021).

Plan the rhythm around the feedback. Fortnightly surveys without fortnightly feedback ask a lot and give little back.

Repeated measurement assumes that a statement measures the same thing in every round. That can go wrong in three ways.

• Response shift. Sprangers and Schwartz describe response shift as a change in the meaning of a self-evaluation due to recalibration of the internal standard, a change in values or a redefinition of the concept (Sprangers & Schwartz, 1999). Their model comes from research on quality of life in illness, but the principle also applies to an implementation: a 3 in week two can mean something different from a 3 in week ten.
• Effects of measuring itself. Repeated questioning can change attitudes, behaviour or answer quality. Warren and Halpern-Manners call this panel conditioning and warn that researchers may then misjudge change between survey waves (Warren & Halpern-Manners, 2012).
• Measurement invariance. Comparisons between groups or moments only make sense if the instrument measures the same thing in each. Vandenberg and Lance describe measurement invariance as a prerequisite for such comparisons and how to test it (Vandenberg & Lance, 2000). With one statement per theme formal testing is hardly possible, so keep wording and context stable.

• Look at the trend, not at one round. A single outlier may be chance; a direction that holds across several rounds is more informative.
• Show the uncertainty. Report respondent numbers and a confidence interval or other measure of spread per round; with small groups, intervals are wide.
• Do not overinterpret small differences. A difference within the uncertainty is not a demonstrable change. Agree on a threshold for action in advance.
• Check who responded. If the composition differs between rounds, a shift in the score may be a shift in the group.
• Read non-response as information. Falling response can signal fatigue or a sense that nothing happens with the answers.

A pulse survey shows that something is shifting; why calls for a deeper look. A drop in one group warrants conversations, profiling research into the considerations behind it or analysis of friction log entries from the same period. Separate articles on both will follow. If you want to know which associations people link to a change, read how to measure associations.

Conversely, in-depth research can yield new statements. Add them as a new theme with its own baseline and leave existing statements unchanged.

This example is fictional. It describes a design; no results are known.

Situation. A university of applied sciences replaces its timetabling system: students get an app, lecturers a new portal for submitting availability.

Decision question. Where does the introduction need adjusting, and does that apply to students and lecturers in the same way?

Available information. Service desk reports show what is said, not how widely it applies.

Suitable approach. Every two weeks, students and lecturers receive five statements on the same five-point scale, plus one open question:

1. I can see my timetable in good time.
2. Timetable changes reach me quickly enough.
3. I know where to go with questions about the timetable.
4. The new system costs me little extra time.
5. I trust that my timetable is correct.

Students are surveyed through a rotating sample, lecturers every round. Results per programme are shown only above an agreed minimum number of respondents. Within a week of each round, a short message reports the main shifts and the follow-up.

Possible interpretation. Suppose the statement about changes scores lower among lecturers for three consecutive rounds but stays stable among students. The team first checks response, composition and uncertainty.

Next step. The team holds short conversations with lecturers and keeps the statements unchanged, so the series stays comparable.

• Too many questions. A pulse survey that keeps growing loses its advantage.
• Changing statements midway. Different wording makes rounds incomparable.
• Measuring without feedback. Response and trust can fall when participants see nothing come back.
• Presenting small shifts as a trend. Without numbers and uncertainty, a few tenths mean little.
• Reading association as cause. A rise after an intervention does not prove that it worked; other things may have changed.
• Overestimating anonymity. In small groups, answers can become traceable.

A pulse survey is a short, repeated measurement that quickly shows how a topic develops. Its strength lies in speed and rhythm, its limits in depth and in the assumptions that repeated measurement requires. If you want to know why something shifts, combine it with in-depth research.