Pulse Surveys vs eNPS vs Health Checks: Which Do You Need?

"Engagement software" is one shelf holding four different instruments, and most buying guides never say so. An annual survey, a pulse tool, an eNPS score and a team health check answer different questions at different speeds; the expensive mistake is buying the wrong instrument, not the wrong logo. So: instruments first, vendors second.

What each instrument detects, and how fast

The annual engagement survey is the deep one: 40 to 80 questions, once or twice a year, sliced by department and tenure, benchmarked against industry. It's the only instrument that reliably surfaces structural problems (pay, career paths, confidence in leadership), and its weakness is latency: by the time the report lands, the people who flagged the problem have had months to start interviewing. Good as an audit, useless as an early warning.

The pulse survey trades depth for frequency: a few questions every week or two, often rotating through a question bank. Detection speed: weeks. The failure mode is fatigue: response rates sag once answering feels like a chore, and rotating questions make trends slippery: this month's number isn't measuring last month's thing.

eNPS is one fixed question (how likely are you to recommend this place to work, 0–10) rolled into one number from −100 to +100. It's comparable over time, legible to a board, and says nothing about why. Detection speed: whatever cadence you keep, at a resolution of one number. We've written a full guide to the math and how to run it honestly.

The team health check has everyone on a team rate a stable set of dimensions (workload, clarity, support), read only in aggregate. Detection speed: weeks, and it tells you which team and which part of work is slipping. The unit of measurement matches the unit of action: a team can rebalance its own on-call load; nobody can action a company-wide 6.8.

Our opinion: under about 150 people, the annual survey is usually the wrong first purchase. You lack the sample size its statistics assume, and you can walk the floor faster than it reports.

Whichever you buy, the graveyard is the same shape

Every one of these instruments dies the same death: measure, present, file, repeat. Scores dip in round two, participation dips in round three, and the program is quietly retired as "survey fatigue." The fatigue isn't from being asked; it's from answering honestly, twice, and watching nothing change.

The fix is boring: every round ends with one or two visible changes, each with an owner and a date, before you ask again. If you won't spend three times as long acting on results as collecting them, buy a cheaper instrument, or none. No vendor can sell you out of this; a few make the loop easier to keep.

Culture Amp is priced like the research platform it is

Culture Amp calls itself "The Employee Experience Platform" and sells three modules (Engage, Perform, Develop) plus people-science consulting. It's the research-grade option: benchmarks, driver analysis, a validated question bank. Pricing is sales-gated. As of July 2026 the pricing page publishes no numbers at all, just "depends on the number of employees, product chosen, and service tier," billed annually, quote required.

Verdict: pick Culture Amp if someone's actual job is people analytics and they'll use the benchmarks you're paying for. Skip it if running surveys is somebody's part-time duty; most of what you'd pay for goes unused.

Officevibe is no longer the cheap pulse tool you remember

For years, Officevibe was the reflex recommendation for lightweight pulse surveys. That product has moved. As of July 2026, officevibe.com redirects to Workleap, now an AI-first HR platform priced like one: Standard $4,999 per year, Pro $11,999, Enterprise custom above 250 employees, engagement surveys folded in alongside performance tools, plus a "Manager Agent" add-on at $99 per manager per month.

Do the per-head arithmetic before assuming it's still the budget pick. For a 25-person company, $4,999 a year is about $16.70 per person per month, against the $4 engagement modules elsewhere on this list; at 200 people the same fee is about $2 a head, genuinely good. Verdict: strong at 100 to 250 people if you want one flat-fee platform; no longer the answer for a 20-person team that just wants a pulse.

15Five and Lattice sell performance; engagement rides along

Both are performance-management products with engagement attached, and their price lists say so. As of July 2026, 15Five sells Engage at $4 per user per month, Perform at $11 and Total Platform at $16, billed annually; Lattice sells Engagement at $4 per seat next to Performance at $10 and a Foundations bundle at $13, with a $4,000 annual minimum on every contract.

The $4 engagement tiers look like the category's bargain, and at 100 people they are. Mind the floor: Lattice's $4,000 minimum means a 30-person team buying Engagement alone still pays about $11 per seat per month. Verdict for both: buy when performance reviews are the real purchase and engagement data should live next to review data. Buying either for surveys alone is renting a stadium to use the parking lot.

SquadBear puts the check inside the HR tool (yes, it's ours)

We make SquadBear, so read this entry accordingly. It's an HR system (leave, time, records) with team health checks and eNPS built in, not a standalone analytics platform. You get per-team checks with anonymity by destruction (closing an anonymous run deletes the person-to-answer link rather than hiding it), trends with drop alerts, actions with owners on each run, and an AI assistant that operates the whole loop:

"Start an anonymous Pulse 5 check for the Platform team, and when it closes, draft one action from the lowest dimension."

Pricing: free up to 5 people, then $4 per active person per month, health checks included, not a paid module. What you don't get: an annual-survey engine, industry benchmarks, driver analysis. Pick Culture Amp if you have an actual people-analytics team. Pick 15Five or Lattice if performance tooling is the purchase. Pick us if the person who approves leave is also the one worrying about morale. Start free or poke the live demo.

AI comment summaries: useful until they quote someone

The category's 2026 feature is AI summaries of open-text comments; Workleap leads its homepage with agents, Lattice bundles an AI agent into every plan, 15Five ships AI-assisted reviews. The useful version is real: nobody reads 300 comments, so themes-with-counts across a large group turns dead text into signal.

The line: a summary that quotes verbatim, or slices below a group size where a writing style or a specific incident identifies its author, has un-anonymized your survey. "Summarize what engineering's detractors wrote" against eight respondents is a deanonymization query with extra steps. And per-person sentiment inference ("who sounds disengaged?") isn't summarizing; it's monitoring, whatever the feature page calls it. Our rules: themes over quotes, a hard floor on group size, no individual-level inference, and know whether raw text leaves your instance.

The shortcut

• Under ~30 people, no HR function: don't buy a standalone engagement platform. Quarterly eNPS plus a monthly team check, inside a system you already pay for, beats a dashboard nobody owns. The $4,000 minimums tell you whose customer you aren't.
• 30–150, first people hire: pulse or health checks on a cadence you'll keep: 15Five or Lattice if you're buying performance management anyway, Workleap if a flat fee suits your size, SquadBear if you want it inside HR ops.
• 150+, with (or hiring) people analytics: Culture Amp, and actually use the benchmarks.
• Any size, never closed a feedback loop before: start with the cheapest instrument you'll act on, prove one loop end to end, then upgrade the tooling.

Related: burnout tracking without the Big Brother, on why aggregated team signals beat monitoring individuals.