Culture Snapshot

Measure the parts of culture that measure well, read the spread instead of the average, check your stated values against what people actually do when it costs something, and pick the one or two gaps you will act on before you survey again.

Psychological safety is strong research (Edmondson, 1999; meta-analysis: Frazier et al., 2017). Engagement has strong to moderate support (Kahn, 1990; Harter, Schmidt & Hayes, 2002). Inclusion theory is well developed (Shore et al., 2011), but the evidence on specific inclusion programs is mixed. Lencioni's vulnerability language is weak or contested (Lencioni, 2002), and eNPS is largely untested. All score bands and thresholds on this page are rules of thumb.

Use it when:

• You want a quarterly read on how the team is doing, beyond the work itself.
• The team has been through a change: a reorganization, new members, a new manager, fast growth.
• Something feels off (people quieter in meetings, tension, a resignation) and you want evidence before you act.

Skip it when:

• You need a fuller diagnosis of team effectiveness: use 4.13 Team Health Scan.
• You already know the gap and want to design new team habits: go to 4.4 Ways-of-Working Agreement.

Culture seems impossible to pin down, yet some of it measures well. Psychological safety, a shared sense that the team will not punish questions, mistakes or disagreement, has the strongest record of the factors here, starting with Edmondson's (1999) field study. Engagement, putting real energy and attention into the work, goes back to Kahn (1990), and unit-level engagement relates to business outcomes in meta-analysis (Harter, Schmidt & Hayes, 2002). Beyond that the ground is softer. Inclusion has solid theory (Shore et al., 2011), but programs meant to improve it show uneven effects. The vulnerability talk common in popular writing comes from Lencioni's (2002) business fable, a story rather than a study.

So the page weights evidence accordingly: the safety statements carry most weight, and the spread of answers matters more than the mean, since a split team can hide inside a decent average. The limit: a survey shows where to look, not why; the why comes from talking to people.

Ask each statement on a 1 to 5 scale (strongly disagree to strongly agree). Record the average and the share who answered 4 or 5.

Strong answer: (illustrative numbers) "Feedback: average 3.6, but only about half at 4 or 5; the rest cluster at 2. Two groups, not one lukewarm team."

The engagement item replaces eNPS ("would you recommend this team"), which is largely untested. Engagement relates to performance in meta-analyses (Harter, Schmidt & Hayes, 2002; Christian, Garza & Slaughter, 2011). For a validated measure, use a published engagement scale rather than one item.

Illustrative example: the roles, the company and all numbers are made up.

The warehouse manager at a freight forwarder in Norway runs the snapshot with a team of sixteen after a merger of two shifts.

Section 1: most items average near 4. Feedback averages 3.6 with only half at 4 or 5: answers split between the old day shift and the old night shift.

Section 2: safety score 3.4, moderate to low. The lowest statement is about asking for help.

Section 3: "safety first" was lived when a picker stopped the line over a damaged rack in the peak week. It was broken when a near miss went unreported because the shift was behind.

Section 4: one-to-ones dropped from 15 to 11 held, all on the night shift.

Section 5: two actions. Handover opens with what went wrong and what was learned (shift leads, from next week). The night-shift lead gets cover for one-to-ones (manager, by month end). The next pulse looks at Feedback, Learning and the help statement.