Requesting an assessment
A typical cycle runs about two weeks from request to briefed results. Most organizations survey on a fixed annual cadence and again within 90 days of a change of command, when climate moves fastest.
| Step | Who | Timing |
|---|---|---|
| Register the organization and define the roster boundary | Commander, director, or designated administrator | Day 1 |
| Receive the survey invitation details and announcement template | Administrator | Day 1–2 |
| Survey window open | All assigned personnel | Day 3–16 |
| Aggregation, screening, and suppression checks | defense-survey.com | Day 17–18 |
| Report delivered | Commander or director | Day 19 |
| Results briefed to the organization | Commander or director | Within 30 days of delivery |
Response rate is a leadership signal in itself
A window that closes at 22 percent participation tells you something before you read a single score. The strongest predictor of participation is whether people saw anything change after the last survey.
Reading your report
The unit report is organized so the first page answers the question a leader actually has: where are we weak, and is it getting worse?
Factor scorecard
All eight factors with the mean, favorable percentage, band, and change since the prior cycle. Read the change column first — direction matters more than level.
Item-level detail
Every scored item with its response distribution. This is where a mid-range factor score resolves into a specific fixable problem.
Subgroup breakouts
Factor scores by grade band, tenure, and supervisory status where group sizes permit. Gaps between junior and senior perceptions are the most common finding.
Screened comments
Written responses grouped by theme with identifying detail removed. Comments explain the numbers; they should not be read as a vote count.
Three common misreadings
- Treating the unit mean as the whole storyA 3.6 built from half the unit at 5 and half at 2 is a very different organization than one where everyone answered 3 or 4. Check the distribution.
- Chasing the lowest single itemOne low item inside an otherwise healthy factor is usually noise. Act on factors and on items that are low in several subgroups at once.
- Reading comments as representativePeople with strong views write comments. Use them to understand the mechanism behind a score, not to estimate how widespread a view is.
Building an action plan
The measurable difference between organizations that improve and those that do not is rarely the quality of the analysis. It is whether the people who took the survey ever heard about the results.
-
Brief the results back, including the bad ones
Within 30 days, in person. Show the actual scores. A briefing that presents only favorable factors costs you the next survey's response rate.
-
Pick two priorities, not eight
Choose the lowest-banded factor and one that is trending down. Everything else goes on a watch list.
-
Name an owner and a date for each
An action without a name attached is a statement of intent. Put both in writing and publish them to the unit.
-
Separate what you can change from what you cannot
Manning levels may be outside your authority. Say so plainly rather than staying silent — unexplained inaction reads as indifference.
-
Re-measure in 90 days
A short pulse on the two priority factors shows whether the intervention moved anything, while there is still time to change course.
Access and confidentiality boundaries
Reports are released to the commander or director of the surveyed organization and to their next higher echelon. Raw individual response records are never released to any person in the chain of command, including by request, and are not retained in a form that could be linked back to a respondent.
Start an assessment for your organization
Registration takes about ten minutes. You will need your roster boundary and a preferred survey window.
Request an assessment