Pulse Surveys People Actually Answer: Questions and Cadence

Most small companies run pulse surveys one of two ways. Either there's a single sprawling annual engagement survey that lands in a slide deck and never gets mentioned again — or someone discovers a survey tool and starts firing off vague monthly check-ins that nobody answers by month three.

Both failures share a root cause: the survey was built as a measurement exercise instead of a decision-making one. If you can't name the decision a question will inform, that question doesn't belong in the survey.

This is a build-your-own guide. Pick a cadence, steal the questions, run the loop-closing ritual, and your second round will be easier than your first.

Why pulse surveys stall out on small teams

Three things kill participation, and they compound.

Nothing visibly changed last time. People are pragmatic. If round one produced no observable difference, round two is a chore with no payoff. Response rate is less a measure of engagement than a measure of whether you've earned another answer.

Anonymity feels theoretical. In a company of thirty-odd people, everyone knows there are only two people in Finance. If someone writes a pointed comment about workload, they assume you'll work out who wrote it. Sometimes they're right. That fear doesn't produce blank surveys — it produces bland ones, which is worse, because bland data looks like healthy data.

The questions are too abstract to act on. "I feel valued at this company" is a fine sentiment and a terrible input. An average score on that tells you nothing about what to do Monday.

Decide what you're measuring first

Before writing a single question, pick two or three things you actually want to learn this quarter. Common ones:

• Engagement / intent to stay — the broad temperature check
• Manager effectiveness — do people get coaching, clarity, and time?
• Workload and sustainability — is the pace survivable past next quarter?
• Role clarity and goals — does everyone know what winning looks like?
• Belonging and team trust — can people disagree without cost?
• Enablement — do they have the tools, access, and decisions they need?

Two to three themes per pulse. Everything else waits. If you're already tracking employee engagement metrics elsewhere, don't duplicate them here — a pulse is for the things you can only get by asking.

Choosing a cadence that fits your size

Cadence should match how fast you can respond, not how often you're curious.

Quarterly is the sensible default for 10-200 people. Four data points a year, enough room to ship a change between rounds, and enough distance that the survey stays a small event rather than background noise.

Monthly works only if you have a dedicated owner. A monthly pulse means twelve loop-closing updates a year. If nobody's job includes that, you'll drift into surveying without answering, which trains people to ignore you.

Event-triggered pulses are underrated. Short, targeted, easy to act on:

• End of a new hire's first month: did it match the pitch?
• Two or three weeks after a reorg or leadership change
• Right after a review cycle closes — was the process fair and useful?
• After a big launch or a crunch period ends

A reasonable annual rhythm: two or three broad quarterly pulses plus a handful of event-triggered micro-surveys.

The 3-5 question rule

Keep every pulse to three to five questions, and make one of them open text.

The scaled questions give you a trend line. The open text tells you why the line moved. In practice, the open box does most of the work — it's where you learn that the drop in "I have what I need to do my job" is really about a broken staging environment, not a management problem.

A clean four-question pulse takes a minute or two to answer. Short enough that people finish it inside the notification.

A reusable pulse survey question bank

Rotate two or three of these per round, keeping one or two fixed as your trend anchors. Unless noted, use a 5-point agree/disagree scale.

Workload and sustainability

• My workload over the past month has been sustainable.
• I can usually disconnect from work outside my normal hours.
• When priorities compete, I know which one to drop.
• I have enough uninterrupted time to do focused work.

Role clarity and goals

• I'm clear on what's expected of me this quarter.
• I understand how my work connects to the company's goals.
• I know what "great" looks like in my role, not just "done."
• Our team's priorities have been stable enough to make progress.

Manager support

• My manager gives me feedback I can act on.
• My manager removes obstacles when I raise them.
• Our 1:1s are a good use of my time.
• I could bring my manager a problem before it became a crisis.

Growth and career

• I've learned something meaningful in the last three months.
• I understand what progression looks like from where I sit.
• I've had a conversation about my development in the past quarter.

Recognition

• Good work on my team gets noticed.
• I've received meaningful recognition in the past month.
• Recognition here goes to the right people for the right reasons.

Tools and blockers

• I have the tools and access I need to do my job well.
• Decisions that affect my work get made quickly enough.
• What's the single biggest thing slowing you down right now? (open text)

Team trust

• I can raise a concern on my team without it being held against me.
• My team handles disagreement well.
• I trust the information I get from leadership.

Open-text anchors (pick one per round)

• What's one thing we should start, stop, or keep doing?
• What's the biggest thing getting in your way this month?
• If you could change one thing about how we work, what would it be?

Writing questions that produce action

Use a 5-point agree/disagree scale for most things. Fast to answer, easy to read. Save the 0-10 scale for eNPS, which conventionally uses it.

Kill double-barreled questions. "My manager gives me clear direction and useful feedback" measures two things and tells you neither. Split it.

Strip the leading language. "How much do you enjoy our new flexible work policy?" already assumes the answer.

Use eNPS as an anchor, not the whole survey. One likelihood-to-recommend question makes a fine trend line across quarters, but the number alone won't tell you what to fix. Our guide to eNPS benchmarks covers how to read it without over-interpreting.

Never reword a trend question. Change "My workload is sustainable" to "My workload has been manageable" and you've started a new trend line and thrown away the old one.

Anonymity on a team where everyone knows everyone

Promise only what you can deliver.

Say this: responses are anonymous, results are reported only in groups of five or more, and nobody — including managers — sees individual submissions. Then hold that line, even when a comment makes someone curious.

Don't say this: "nobody will ever be able to guess who said what." On a nine-person team with one person on parental leave, that isn't credible, and overpromising costs you more trust than honesty does.

A five-response reporting threshold is a sane default. If a team is smaller than that, roll it into a department view. Tell people the threshold in advance — knowing the rule is what makes the promise believable.

Getting a response rate you can trust

Aim high. A pulse that fewer than half your people answer isn't a measurement, it's a sample of whoever's already loudest.

• Send Tuesday or Wednesday morning. Monday's a wall, Friday's a graveyard.
• Use the channel people already live in — Slack or Teams, not a separate email nobody checks.
• Brief managers a couple of days early. Give them the questions, the reporting threshold, and one line to say in standup. A manager who mentions it in person moves participation more than any automated reminder.
• Keep it open five business days. Longer just extends the procrastination window.
• Send exactly two reminders: one midway, one on the morning of the last day. Both should show the current response rate, not scold anyone.

An announcement worth copying:

We're running our quarterly pulse — 4 questions, a couple of minutes. It closes Friday at 5pm. Responses are anonymous and we only report results for groups of 5+. Two weeks after it closes, I'll share what we heard, what we're changing, and what we're not changing and why. Here's what came out of the last round: [one specific thing you shipped].

That last sentence is the highest-leverage line in the message.

Reading results when your sample is 25 people

Most survey advice was written for companies where a single segment has hundreds of responses. It breaks badly at small scale.

Read direction, not decimals. With a couple of dozen responses, one frustrated person can move an average by a few tenths. A small dip is noise. A large drop, sustained over two rounds, is signal.

Look at the distribution, not the mean. An average in the middle could be everyone mildly lukewarm, or half the team delighted and half checked out. Those need opposite responses. Count how many people picked 1 or 2 — that's usually the more useful number.

Segment sparingly. Slicing thirty responses into six teams gives you six meaningless numbers. Segment only where the group clears your reporting threshold and where you'd genuinely act differently based on the answer.

Two data points is a line, not a trend. Wait for three before you tell a story about direction.

The open-text pile: a 30-minute method

You don't need a data team. You need a spreadsheet and a timer.

1. Paste every comment into one column. (7 minutes)
2. Read them all once without tagging anything. Just read. (5 minutes)
3. Go back through and write a one- or two-word theme next to each: staging environment, unclear priorities, hiring pace, manager access. Invent tags as you go. (12 minutes)
4. Sort by tag and count. (3 minutes)
5. Pull three verbatim quotes that best represent the top themes — you'll use them in the update. (3 minutes)

The top two or three themes are almost always obvious by step three. Don't overthink the taxonomy.

Closing the loop in 7 days

This is the part that makes round two work, and it's an operations task more than a communications one.

Within seven days of the survey closing, publish an update with three sections:

Here's what you said. Two or three themes, with a representative quote each. Include the uncomfortable one.

Here's what we're doing. Specific, owned, dated. "Ana is rewriting the on-call rotation by Nov 14" beats "we're looking into workload." One or two commitments you'll actually finish beat six you won't.

Here's what we're not doing, and why. This is the section people remember. If three people asked for a four-day week and it isn't happening, say so plainly. Declining a request with a reason builds more credibility than silence — silence reads as "they ignored it."

Keep a running decision log: theme, date raised, decision, owner, status. That log is why your third survey is easier than your first, and it's where you get the evidence to open the next announcement.

Turn themes into 1:1s and goals, not a slide

A survey theme that lives only in a deck dies in the deck.

If role clarity scored low, that's a goal-setting problem — take it into quarterly planning and make sure every person can name their top two priorities. Our OKR framework guide has a workable structure for that.

If manager support scored low, don't announce a training program. Give managers two or three questions to raise in their next 1:1 — the one-on-one question bank is a good place to pull from — and revisit in the next round. Pair it with a real manager effectiveness framework if the pattern holds across quarters.

If the theme is about feedback quality or fairness, it probably belongs in how you run your review cycle rather than in a standalone initiative.

A first-90-days rollout

Days 1-10: Pick your two themes. Write four questions. Set your reporting threshold. Brief managers.

Day 14: Baseline pulse opens. Five business days, two reminders.

Day 26: Publish the loop-closing update. Start the decision log.

Day 55: Ship at least one visible change from round one.

Day 62: Second pulse. Same two anchor questions, one rotated in. Open the announcement with what changed.

Day 76: Second update.

Day 90: Review the decision log. What's still open? What's the pattern across both rounds?

Common mistakes

• Adding questions every cycle. Surveys grow the way drawers fill. Cap it at five.
• Rewording trend questions. You lose the comparison. Add new questions instead of editing old ones.
• Surveying during crunch. You'll measure the crunch, not the company.
• Letting managers see individual responses. One breach of this ends honest answers permanently.
• Only sharing the good results. People can tell what's missing.

The hard part of pulse surveys was never the questions — it's the follow-through between rounds. LVL Up Performance keeps pulse surveys, eNPS, 1:1s, goals, and recognition in one place, so a theme from Tuesday's survey shows up in next week's 1:1 agenda instead of a forgotten slide. Start free and run your baseline pulse this month.