Many managers send their staff a satisfaction survey and get back polite answers that say nothing, or no answers at all. The cause is usually not the staff but the survey: questions that are vague or long, or that make people feel their answer will be traced back to them.
Start with one question: what will you do with the results?
Before writing a single question, choose two or three areas you can actually change: workload, support from managers, or room to learn. Asking about something that will not change whatever the answer teaches the team that the survey is a formality, and fewer people answer next time.
The core questions
- Overall satisfaction with the work, on a 1 to 5 scale.
- Workload: light, about right, heavy, too heavy.
- Support from their direct manager, from 1 to 5.
- What they value most about the work, with several choices: the team, learning, flexibility, pay and benefits.
- How likely they are to recommend the workplace to a friend, from 0 to 10.
- One open question: what single change would make the biggest difference for you?
The recommend question and eNPS
“How likely are you to recommend this workplace to a friend?” applies to staff the Net Promoter Score that Fred Reichheld introduced in Harvard Business Review in 2003; asked of employees it is called eNPS. The arithmetic is simple: a 9 or 10 is a promoter, 0 to 6 a detractor, and the score is the share of promoters minus the share of detractors. Its real value is not the number alone but how it moves from one survey to the next.
Why leave names out?
People answer sensitive questions more candidly when the answer cannot be linked to them; otherwise answers drift towards what seems acceptable. So do not ask for a name or an email. Watch the demographic questions too: department plus age plus years of service can be enough to identify someone in a small team, so keep departments broad and merge small teams.
Keep it short
Seven questions are enough for a regular survey and take two or three minutes. The conversational layout, one question per step, helps keep attention on the current question instead of a long page of fields.
After the responses come in
- Share a summary of the results with the team, including the uncomfortable parts.
- Pick one or two steps only, and say what they are.
- Ask the same questions every three or six months, so you can compare.
- In Formza the results page shows each rating question's average and how the choices split, and you can export the responses to Excel or send them to Google Sheets.
Formza's “Employee satisfaction survey” template comes with these questions in Arabic and English, and asks for no name. Try it first as an employee would see it, then copy it and adjust it to your team.
Sources
- Frederick F. Reichheld — The One Number You Need to Grow, Harvard Business Review, December 2003 (hbr.org).
- Pew Research Center — Writing Survey Questions (pewresearch.org/writing-survey-questions).
