Are employee surveys still worth doing?

We’ve been doing employee surveys for several decades, typically as a part of our internal communication audits. We like the hard data that surveys provide, which complement the more nuanced comments we hear in employee focus groups.

But some organizations are questioning the value of surveys, and the reasons range from the mundane to the alarming. They tell us that employees are suffering “survey fatigue” and won’t provide good feedback or even bother to take the survey at all.

We’ve also heard of employees figuring out how to “game” engagement surveys so their scores will be higher than how they’re actually feeling. The reason, they say: If our department scores low, or lower than other departments, “we’ll have to do something,” like attend workshops or other HR initiatives designed to raise spirits.

The biggest concern about surveys is illustrated by a recent Microsoft employee survey. Business Insider reported that employees were questioning why survey results released by the company omitted certain results that made it appear employees had no concerns about their level of satisfaction with the company or its leadership.

One employee posting on a Microsoft internal forum put it this way, Business Insider reported:  “I don’t think they value getting an answer to a question they already know the answer to,” along with a meme of the famous line from the movie “A Few Good Men”: “You can’t handle the truth!”

Despite these problems, or maybe because of them, surveys remain a critical tool for communications measurement, but there are pitfalls. Here’s our guide to avoiding them:

1. Guarantee and protect anonymity. Many employees are highly skeptical of the confidentiality of internal surveys, especially when they are produced by the company itself. This may sound self-serving from a company like ours, which conducts employee surveys, but our clients tell us time and again: having a third party ask the questions makes a difference in the quality of the answers. Trust drives honest responses.

This is made far worse by the recent trend of organizations pre-loading their demographic data into the survey, like department, role and job location. This allows respondents to be identified through elimination, and employees are hyper-aware of this.

For example, if you’re the only manager in a 12-person engineering team, tagging your response with “department: engineering” and “job level: manager” makes you identifiable even without your name attached; small or unique demographic combinations effectively strip away the anonymity the survey promised.

2. Pay attention to margin of error. Margin of error tells you how much the results from your survey sample might differ from what you’d get if literally every employee answered. It’s usually expressed as a plus-or-minus percentage.

For example, if your survey finds that 72% of employees describe themselves as “engaged” with a margin of error of ±4%,” that means the true number for your whole company likely falls somewhere between 68% and 76%.

It’s not a flaw in your survey… it’s just the mathematical price of asking a sample instead of everyone (which is called a census).

3. Don’t rely on favorability ratings alone. Ask about conditions, not just sentiment. Traditional surveys lean heavily on favorability scores that may look good on paper but lack predictive and actionable value, since high marks don’t necessarily mean employees will stay or that conditions are healthy.

Instead of “How satisfied are you with your job?“ ask about the specific conditions that produce satisfaction: “I have the resources I need to do my job effectively” or “My manager gives me feedback that helps me improve.” Conditions are observable and actionable; sentiment is a mood.

4. Ask what drives engagement, not whether it exists. Instead of uncovering what drives engagement, many surveys simply ask whether employees feel engaged, resulting in shallow data with little chance for action.

So rather than “Rate how engaged you are at work,” ask questions tied to known drivers — organization, manager, team and job, as in “My team collaborates effectively to solve problems” or “I understand how my work connects to company goals.”

Each question should map to something leadership can really change.

5. And speaking of change, don’t ask if you don’t plan to act. This is the fastest way to kill trust and future response rates. Don’t ask “Do you have opportunities for career growth?” if there’s no budget or plan to create any.

Collecting feedback without follow-up is worse than not asking at all, and when leaders fail to make changes, employees become skeptical, disengaged and less likely to respond honestly in the future.

Before finalizing a question, ask internally: “If this comes back negative, can we do something about it in the next 6 months?” If not, cut the question or reframe it as intended to set a baseline and say so.

6. Share the results. Employees don’t mind surveys, but they hate when no one tells them what came of it all. They hate it even more if they think leaders are cherry-picking the results, as the Microsoft episode reveals.

The Microsoft “head of employee listening” responded to employees on an internal company messaging board, saying questions omitted from the survey summary “are still being asked and acted on; they just show up in different surveys based on how our listening programs are designed.”

The better practice is to provide a complete summary, so employees can see their input mattered. Write a story about the key findings, quoting both employees and leaders.

And when the organization acts on recommendations from the survey, make a point of reminding employees that the change came from their feedback.

7. Write your survey with the audience in mind. Use these five tips:

a. Keep it short. We look at the time it takes to complete the survey, not the number of questions. Long surveys tank response rates and diminish the quality of the responses.

b. Time it well: Avoid the week of layoffs, a time close to major deadlines or events, or right after other big surveys.

c. Avoid jargon in your survey questions. Plain language gets better data.

d. Don’t ask too many questions per category: Three to five well-crafted questions per theme beats 15 mediocre ones.

e. Don’t ignore open-ended comments: They often contain the most actionable insights.

Jim Ylisela, co-founder of Ragan Consulting Group, learned everything he knows about survey research from measurement consultant Katrina Gill. Interested in a communications audit? Email Jim to set up a free call with him and Katrina.

Follow RCG on LinkedIn and subscribe to our weekly newsletter here.