When analysing results from employee surveys, much attention is focused on deriving a single measure to gauge employee engagement. The irony however, is that these surveys can often in themselves be, well, rather un-engaging.

Something must be done, not least for all those spiritless employees, passively pointing and clicking their way through the next 15 minutes of apathy. So for all those thinking of conducting their next employee survey, why not think…SPAM?

Let us explain.


Employee surveys tend to comprise banks of rating scale questions. These are often presented as grids, which will inevitably lead to some respondents ‘straightlining’ their responses (ie. giving the same answers for each statement), particularly if these grids are numerous and long.

Ask yourself whether information from straightliners is useful and then ask whether there may be a better way to visually present these questions to keep people thinking?

Why not try mixing different question types with different response structures, try inverting scales, maybe add in opportunities for respondents to give free-text opinions.


Remember that your employee survey is likely to be completed by a wide audience. Concepts and phrases that may be commonplace to the researcher may be considered cringeworthy to non-office staff, or even worse, be entirely misunderstood.

Put a ban on industry jargon, buzzwords and technical concepts.

Plain English is key – if your audience feels patronised, then you should expect the potential hijacking of responses.

Try to test out a draft on staff from different roles in your office to see whether the language is pitched appropriately to as wide a group as possible.

It is also important to consider the impact of double- (or even triple-)-barrelled statements. For instance, take “My organisation inspires me to do a good job and achieve its objectives” as an example. Surely, you can miss the objectives, but still do a good job?


You have to sell your survey to staff, and importantly, stress how the results may impact on organisational change.

If employees consider employee surveys to be just another tick-box exercise, then they won’t take it seriously.

This can potentially undermine the value of what can be derived from the responses.

One useful approach is to include a review of key results from previous surveys, and importantly, some specific actions that have were taken as a result. There is so much evidence demonstrating that when people feel their view matters, they consider their responses carefully.


Do you have multiple questions that measure the same concept? I bet you do!

This in itself is not an issue, many models have been built around finding whether multiple questions actually measure some common underlying concept; take measures of happiness for instance or IQ scores. However, when the difference between question statements is semantic, it only serves to bore, or even antagonise the respondent.

Yawn…I’m getting tired just thinking about it!

If you don’t put much effort in to question design, it will be obvious to participants and they won’t put in any effort either.

Employee surveys are not some romantic academic exercise, they should be a real and valid opportunity for employees to have their say, and hopefully, for any organisation to identify strengths and weaknesses to better understand and implement change.

Remember – you are an employee too. Why not ask yourself whether you found it a rewarding experience when filling in your last staff survey?

And if you didn’t, remember, think;





Think SPAM.

Get in touch

If you, or someone you know, are thinking about conducting an employee survey in your organisation, or you simply want some tips, feel free to pick up the phone.

Our advice is free and it might just help your next employee engagement survey become a little more, well, engaging.

Dr Michael Turner

Tel: 01213336006

Email: michael.turner@bmgresearch.co.uk

Website: www.bmgresearch.co.uk

Article written by Alan McConville, Clive McDonnell & Michael Turner – BMG Research

Share this article: