Do your employee engagement survey questions and scales measure up?

To ensure your employee engagement survey questions and scales measure up: Relevance: Ensure questions address key engagement drivers (e.g., recognition, growth). Clarity: Use clear, specific language to avoid ambiguity. Validity: Align questions with organizational goals and employee experiences. Reliability: Use consistent scales and methods for accurate comparisons. Actionability: Design questions to provide insights that can drive actionable improvements. Regularly review and update your survey based on feedback and evolving needs.

To eliminate potentially ambiguous, unecessary or unclear survey questions, run the questions across a group of people from various teams - the survey team, HR, communications, and possibly legal depending on the topic.

I'm going to swear... ... But AI may help here! Load your questions into your AI platform to understand if the questions make sense, or suggest improvements. Avoid utilising AI to form the question set, unless you're confident you've programmed the platform to support your goals. If you're reading this, you must be tempted.. I mean this is a response to an AI article ??

It's useful to have definitions for terms like "manager" and "team" so people are clear who they should be thinking about. You also need to consider question order. For example, if you ask several questions about the same topic one after the other or in a laundry list you can get a "halo effect" (where someone answers all the questions based on how they answered the first one).

Lack of Relevance: Ensure that survey questions are relevant to the organization's specific goals and challenges. Generic questions won't provide actionable insights.

If your survey provider has the functionality to, then I recommend adding points of clarity to questions or before a set of questions E.g. under a question about leadership teams add a clarifying point "When thinking about leadership teams we want you to focus on the following..." E.g. Before asking questions relating to Inclusion put a page in to say "When answering these questions please focus on how you feel rather than on you view of how your colleagues might feel"

A good test is to ask yourself what will you do with the results to the question (good or bad). If it is difficult to define an action that will be taken to either maintain or improve things based on the question, then it is probably not adding much value. Distinguish between 'outcome' questions and 'input' questions. Outcome questions are things like 'I would recommend the organisation as a great place to work' - they tend to score highly when everything else is also positive and require many actions to have an impact. Input questions are more specific behavioural questions that, when combined, will ultimately result in engagement. Input questions are questions such as "I receive frequent praise when i do good work".

Ideally you will use questions that have been validated by academic research. At a minimum, ask questions that you can use to help inform you of future actions, not just a "this would be interesting to know". Think about if, and how, you will use the answers to the questions before you use those questions.

Overuse of Scales: Limit the use of extensive scales (e.g., 1-10) as they can be overwhelming. Sometimes, a simple "agree-disagree" scale suffices.

Try to avoid gimmicks or “survey theatre”. Will people understand the meaning of smileys or a flower going from wilting to full bloom, in relation to the question and, importantly, will it make easy-to-take survey on any device? Anchor the points on the scale to text, such as strongly disagree to strongly agree, rather than numbers (and turn the points into a numerical score). Don’t have too many points on your scale, to ensure that it remains meaningful. Do people really know how to differentiate between 6 and 7 on a 10 point scale? Limiting the points and anchoring to text will, in combination, make for more realistic responses. I would generally also recommend enabling genuinely neutral responses by using an odd number of points.

Keep the scale consistent throughout the survey. Many surveys use different scales for different parts of the survey (eg the NHS National Staff Survey). The problem with this is that people get used to the survey scale and don't always notice when it changes part way through. Especially where the positive responses switch sides throughout the survey!

Don't overthink it - most of the scales provide the same outcome. However, if you want to avoid people sitting on the fence, remove the fence. Go for a 4-point scale removing 'neither agree or disagree' - a tip that helps with recruiting to nudge for more decisive responses.

It's important to consider how you will report results when you design survey questions and scales. For example, numeric scales that produce a mean score are tricky for managers to interpret. While a Likert scale and a percentage score is much easier to understand and communicate. A manager can translate a favourable score of 75% into "three-quarters of my team think this".

In my experience there is not enough time given to question outcome and managing different stakeholders is important …the best employee surveys are actionable not just a % point in a report.

Don't overthink it. There's no 100% fullproof approach. Where possible, add comment, and be explicit in what you're looking for with comment (i.e. 'why I responded this way' Vs 'what do I understand from the question')

Surveys get you the quick and dirty, but alone are not going to solve your engagement mission. If you are using them, make them non-anonymous and the responses to start conversations or gather material to improve engagement.

Avoid using jargon or complex language in survey questions. Keep them simple and concise to ensure clarity for all employees. Each question should have a clear purpose and focus. Avoid ambiguous or vague statements that can lead to misinterpretation.Ensure that survey questions are relevant to the organisation's specific goals and challenges. Limit the use of extensive scales as they can be overwhelming. Sometimes, a simple "agree-disagree" scale suffices.Don't rely solely on quantitative data. Include open-ended questions to gather qualitative insights and suggestions.This can reveal disparities that need attention.Also ensureclear plan in place to act on survey results. Employees need to see that their feedback leads to positive changes.

Follow-up questions can be very powerful. For example, if someone is negative on a question about feeling respected, you might ask for more information and examples. From a respondent's point of view, surveys can be frustrating unless you're given space to explain your feedback. And from an analysis point of view, having a list of examples provides a rich pool of insights. It's like running an "instant focus group".

Great point from Nick Lyn, using question routing or branching to get extra context is useful I would add that when designing a survey response mechanism avoid non inclusive solutions like Sliders and Graphics and stick to a scale using radio buttons or likert style selections

Consider adding open-ended text boxes in each section of the survey to collect qualitative data and stories that can illustrate additional information from employees. These text-based additional answers can be more informative than the quantitative results sometimes.