A climate survey tells you what to fix
Engagement scores tell you how people feel. A workplace climate survey tells you which conditions made them feel that way — workload, autonomy, leadership, recognition, safety, belonging — and which team is carrying the worst of it. This is how to design one, how often to run it, and how to keep the answers honest enough to be worth reading. If you employ people in Mexico, it also has to line up with NOM-035.
Climate is the weather
your team actually works in
Organisational climate is the shared perception of what it is like to work somewhere: how work is assigned, how decisions get made, how much control a person has over their own day, whether speaking up costs anything. It is not culture — culture is the slower thing underneath, built from values and history. Climate changes in weeks, which is exactly why it is worth measuring often.
Why the distinction matters
An engagement score of 62 is a number without an instruction attached. It tells you something is wrong; it does not tell you whether the cause is a manager, a rota, a promotion cycle, or a tool that breaks twice a week.
Climate items are written to be actionable. "I have enough time to do my work well" points at staffing and planning. "I can raise a concern without it being held against me" points at a specific manager relationship. Both are fixable. "How engaged are you?" is not.
This is also why climate data ages badly. A restructure, a peak season or a single departure can move a team's climate inside a month — and an annual study will not notice until the following spring.
Climate, engagement, culture
Climate — the current conditions. Measurable, volatile, and the thing a manager can change this quarter.
Engagement — the outcome those conditions produce. Commitment, effort, intent to stay.
Culture — the values and habits underneath both. Real, slow, and not something a survey moves.
A programme that measures only engagement is measuring the symptom and hoping the cause is obvious.
| Survey type | Measures | Typical length | Best cadence | What it is good for |
|---|---|---|---|---|
| Climate | Working conditions, by dimension | 15–40 items | Rotating pulse, plus one deep study a year | Deciding what to change |
| Engagement | Commitment and intent to stay | 10–30 items | Quarterly or annual | Tracking whether it worked |
| Pulse | A slice of either, repeated | 3–5 items | Weekly to monthly | Noticing a change in time to act |
| Psychosocial risk (NOM-035, ISO 45003) | Hazards a regulator recognises | 46 or 72 items | Set by the standard | Evidence an inspector accepts |
Six dimensions,
and one honest question each
Most climate instruments collapse into the same six areas. You do not need forty items to cover them — you need one well-written item per dimension, rotated so nobody answers the same question twice in a row. Here is the shortest version that still works, with a sample item you can use as written.
Demand and pace
The single strongest predictor of burnout, and the one managers most consistently underestimate on their own teams.
"I have enough time to do my work to a standard I am proud of."
Control over the work
High demand with low control is the classic strain combination. Either one alone is survivable; together they are not.
"I have a real say in how I do my job day to day."
The manager relationship
Climate scores cluster by manager far more tightly than by department, which is why the segment matters as much as the item.
"My manager gives me the context I need to do my job well."
Effort and reward
Effort–reward imbalance predicts exit intention months before anyone updates a profile on a job board.
"The effort I put in is noticed by the people it should be noticed by."
Psychological safety
The dimension that gates every other answer: a team that cannot say a hard thing will not say it on a survey either.
"I can raise a problem here without it being held against me."
Relationships and fit
Weak on its own, decisive in combination — it is what keeps people through a bad quarter, and its absence is what makes a good offer easy to accept.
"I feel like I belong on this team."
Five steps, and the fifth
is the one everyone skips
A climate survey is not a document, it is a loop. The measuring part is the easy half; the part that decides whether anyone answers the next round is what happens after the results land.
Set the segments first
Decide how you will slice the data before you send anything. Team and manager beat department almost every time.
Write for action
Every item should point at something a named person could change within a quarter. Cut the rest.
Run it short and often
Three to five items every two to four weeks, rotating dimensions. Response rates hold; annual studies do not.
Protect the answers
No identity link, and no segment shown below five respondents. Say so in the invitation, in plain words.
Report back, then act
Share what you found and what you will change, within two weeks. Silence after a round is what kills the next one.
What Bloomder automates
- Sending, a single reminder 48 hours after launch, and closing each round.
- Rotating question sets, so the same item does not repeat back to back.
- Scoring by team and by manager, with the five-respondent floor enforced in the data layer.
- A one-click AI read of the open text — themes, burnout risk by team, and what to do about it.
- Templates written by an organisational psychologist, not recycled from a generic form builder.
What it does not do
- Write your action plan. The AI proposes; a manager still decides and still has to show up.
- Replace a conversation. A survey finds the room; someone still has to walk into it.
- Substitute for the statutory instruments where a regulator names one — see NOM-035 and ISO 45003.
- Identify a respondent. Structurally, not as a policy — there is no individual record to produce.
Four ways a climate survey
quietly stops working
Segments too small to be safe
A four-person team gets its results published, one member is obviously the outlier, and the whole organisation learns that anonymity here is a figure of speech. Every score after that is a polite fiction. The fix is a hard floor enforced in the software, not a guideline in the survey policy.
Asking about things nobody will change
Items about pay bands, office locations or the strategy are fine in principle and corrosive in practice if the answer is always "noted". Ask about what is actually on the table this quarter, and say plainly which things are not.
An instrument long enough to be dreaded
A forty-item study run once a year gets a strong first response rate and a weaker one every year after. Length is a tax on honesty: people who are busy — the ones you most need to hear from — satisfice, and the data quietly turns into noise.
Results with no reply
The single most reliable way to end a survey programme is to run a round and say nothing. Reporting back what you heard, including the uncomfortable part, is not a communications nicety; it is the mechanism that buys the next round's response rate.
If you employ people in Mexico,
climate is also a legal obligation
NOM-035-STPS-2018 obliges Mexican employers to identify, analyse and prevent psychosocial risk at work, and to document that they did. It names its own instruments — Guia de Referencia I, II or III, depending on headcount — and sets the cycle. A general climate survey overlaps with them, and does not replace them.
Where the two meet
The dimensions above are close cousins of the NOM-035 domains: workload, control, leadership and relationships all appear in both. Running a climate pulse gives you the early signal; the statutory questionnaire gives you the record an inspector will ask for.
Doing them in one platform is the practical difference. Two vendors means two response-rate problems, two anonymity stories to defend, and two sets of numbers that will not reconcile in front of an auditor.
What Bloomder ships for it
- The official Guia de Referencia II and III questionnaires, in Spanish, item for item.
- Distribution on the cycle your headcount requires, with a timestamped record of every send.
- Aggregate risk scoring with colour-coded thresholds by area.
- PDF and CSV exports in the shape STPS auditors expect.
Climate survey FAQ
What is a workplace climate survey?
It measures how employees experience the conditions they work under — workload, autonomy, leadership, recognition, safety and relationships — rather than how loyal or satisfied they feel. Climate is the environment; engagement is one of the outcomes that environment produces. A climate survey is diagnostic: it is meant to tell you which conditions to change.
How is it different from an engagement survey?
An engagement survey measures an outcome — commitment, discretionary effort, intent to stay. A climate survey measures the causes of that outcome. Engagement tells you the temperature; climate tells you where the draught is coming from. Most teams need both, and a short recurring pulse survey is the cheapest way to run them together.
How often should it run?
A long annual study answers questions that are eleven months stale by the time anyone acts. The practical pattern is a short pulse of three to five items every two to four weeks, rotating dimensions, plus one deeper study each year. Bloomder automates the recurring half — sending, one reminder 48 hours after launch, and closing the round.
How do you keep it anonymous?
Anonymity is a mechanism, not a sentence in the invitation email. Bloomder issues a single-use token that is invalidated on submission, stores answers with no identity link, and hides any segment until at least five distinct people have responded — a floor no administrator can lower. Because no individual record exists, none can be produced later. The full method is documented on /security/.
Does it satisfy NOM-035 in Mexico?
Not on its own. NOM-035-STPS-2018 requires specific instruments on a defined cycle, with documented action plans and records available for inspection. A general climate survey overlaps with those instruments but does not replace them. Bloomder ships the official questionnaires alongside its climate templates, so one programme covers both.
How many people do I need before the numbers mean anything?
For a single team, five respondents is the floor Bloomder enforces before it will show a segment at all — below that, results are hidden rather than rounded. For trend reading rather than a snapshot, three completed rounds is roughly where a line becomes a signal instead of noise. The free plan covers up to 25 employees, which is enough to run a real programme before you pay for one.
Run your first climate pulse this week
Templates written by an organisational psychologist, the five-respondent floor on by default, and AI insights in about a minute. Free up to 25 employees, no credit card.
Employing people in Mexico? Start with NOM-035 compliance instead.