What is a pulse survey?

What makes a survey a pulse

Three features, all of them about repetition. The questions are fixed, so this month's figure can be read against last month's. The schedule is fixed, so people know when it is coming and results arrive at a predictable rhythm. And the survey is short enough to answer in a couple of minutes, because nobody will answer a twenty minute survey every month. Qualtrics describes a pulse as a short set of questions sent to employees on a regular basis to continuously gather feedback and measure sentiment, and notes that pulses let an organisation plot trends over time and link improvements back to actions. That last part is the real purpose. An engagement survey tells an organisation where it stands. A pulse tells it whether the thing it changed in March made any difference by May.

Choosing questions and cadence

Start from the decisions the results will feed. If managers review workload and staffing every month, a monthly pulse with statements about those topics makes sense; if the leadership team reviews culture twice a year, a monthly pulse on culture produces numbers nobody uses. Qualtrics suggests about 10 to 15 questions for a monthly pulse and 15 to 20 for a quarterly one, and a smaller set is often better still. Gallup's Q12 engagement survey is built on 12 core items and recommends measuring every six months after an initial benchmark, which shows how much can be covered in a short, stable instrument. Keep one team question for splitting results, a handful of statements on one shared scale, and an optional open question. Test the draft on a phone, since many staff will answer on one during a break.

Anonymity in small teams

Pulse surveys ask people how their manager and workload are treating them, so honest answers depend on believing that nobody will be singled out. The team question is where that belief can break. A team of four with results shown by team is not anonymous in any meaningful sense, especially once a workload answer or an open comment narrows it further. Set a minimum group size before results for a team are shown, merge small teams into a larger unit, and say both things in the invitation. If a manager might want to follow up on concerns, ask permission in the survey itself, as a separate question, rather than assuming it. Promising anonymity and then quoting a comment that everyone recognises does more damage to the next round than any low score.

Fatigue comes from inaction, not frequency

People tire of pulse surveys when answering changes nothing. The fix is less about sending fewer surveys and more about closing the loop: share the results within days, say what will be done about the one or two issues that stood out, and report back on it before the next round. If nothing can change between rounds, the cadence is too fast. A useful rule is that the interval should be at least as long as the time it takes to act on a result. Watch the response rate from round to round as closely as the scores, because a falling rate is the earliest sign that people have stopped believing the survey leads anywhere.

Running a pulse as a form

A conversational form moves through the questions one screen at a time, which suits a two minute check on a phone. Publish once and leave the published version alone for as long as the questions stay the same; each response keeps the version number it was submitted against, so a later wording change is visible in the export. A stored response carries no IP address, browser details or account, so a public link with no identifying questions is anonymous in practice. The Submitted At column in the CSV splits one month from the next.

Questions people ask