Opinions About Decisions, Not Feelings
Two different conversations hide under the phrase employee feedback. One asks how people feel at work — satisfied, drained, motivated. The other asks what people think about the company’s actual choices: the benefits package, the expense process, the tools they’re issued, the rules about where work happens. Feelings change slowly and belong to the annual census. Opinions about decisions are actionable immediately, which is what makes this survey the operations team’s favorite instrument.
Run it twice a year. The output is not a sentiment score — it is a ranked list of things the company does badly, straight from the people the company does them to.
What Gets Graded
Five domains cover how an organization actually operates:
- Benefits and perks. A rating item per benefit — health cover, allowances, leave policy — because one blended score hides which benefit earns its cost.
- Tools and systems. The blunt multiple choice question: which single system wastes most of your week? Results here are procurement arguments won in advance.
- Processes. Expenses, approvals, onboarding paperwork, meeting norms — each rated on a simple opinion scale.
- Where and when work happens. Hybrid rules, core hours, office setup — opinions that never move a satisfaction score but decide resignations at the margin.
- Internal communication. Do people learn things on time, from the right source, in the right tone?
Then one open long text field: if you could change one operating decision tomorrow, which one and why. This single field outperforms every rating in the survey — it is where the fix list actually gets written.
From Results to Repairs
The survey’s product is the ranked fix list: domains sorted by score, each with verbatim suggestions attached. Leadership picks the top three, names owners, and — the step that decides everything — publishes the list back. Six months later, the follow-up edition can recall the original answers and ask the pointed questions: did the expense process actually improve? Employees who watched their suggestions vanish rate every future survey accordingly.
Segmentation Without Exposure
Department and tenure banding makes results useful; identifiers make them dishonest. Collect segment metadata only, set a minimum group size — five is the common floor — below which results fold into the parent group, and state the rule on the first screen. The response archive then becomes the running record of what was asked, what was promised, and what got fixed. Sibling instruments live in the HR surveys collection, with the wider people-ops toolkit in the human-resources category.
Frequently Asked Questions
How is an opinion survey different from a satisfaction survey?
Satisfaction measures the person; opinion measures the organization. A satisfied employee can still think the expense process is a joke — and that opinion is the fixable part. Mature programs run both and read them together.
Twice a year — why not monthly?
Operating decisions change on quarterly cycles, not monthly ones. More frequent opinion surveys mostly re-measure the same opinions, and each repetition costs candor.
A suggestion targets one specific team. Now what?
Route it to that team’s lead privately, not into the all-hands results. Some truths need a smaller room, and protecting that distinction is what keeps people writing specific suggestions instead of safe ones.
Should open answers be published verbatim?
Rarely. Publish themes and counts; reserve verbatims for the small group working the fix list. Anonymity survives on the promise that raw text stays put.