Whether it is to meet regulatory standards, to gain some form of accreditation or simply to get a better feel for how employees are engaged, organisations are increasingly looking to employee surveys for the answers. However many end up with more questions afterwards or launch knee jerk projects based on spurious data. This is partly because writing any form of employee survey or questionnaire is a skill and interpreting the data, even more so.

Do It Yourself?

While the phrase "Well you would say that wouldn't you?" might apply here, using a specialist company to do this work is often more effective in the end.( My company provides employee surveys, hence the comment) however before you switch off and think this article is biased in that direction, firstly let me promise that the article is designed to help you design your own effective surveys, but it is important to understand why a third party might elicit better results. It really boils down to two things. 1) like so many skills, people who do it a lot, tend to be better at it and 2) outsiders have no history, no hidden agendas, are less prone to bias or influence and employees are far more likely to open up and reveal information, if they know it can't be attributed to them.

Ok that message is over, assuming that you still want to get your own fingers on the pulse of your employees, then here is how to do it.

What is the purpose?

The first step is to decide what the purpose is. If it is to meet a regulatory standard then clearly whatever you design must be fit for that purpose and quite often they already exist and can be copied. If it is to find out how stressed employees are, then it should be designed with that sole purpose and not drift off into other interesting topics. A clear idea of what you want the information for will allow you to challenge the need for each question. Typical needs are employee’s engagement surveys, stress surveys or employer branding surveys.

Management buy in

Having decided what you want to find out, are your senior managers truly going to respond to the employee survey findings however controversial or uncomfortable they are? If not then don't bother for you will simply have raised expectations of change and by not responding, you will have reinforced many negative views about management.

Online or paper?

Now you are ready to think about the best medium. For most organisations an online survey to all employees, followed up by focus groups with at least 5% of the workforce is effective. For employees who do not have access to the internet, you can offer a paper version however setting up some dedicated PCs for that purpose and scheduling employees to go to them, is generally better for both admin and the perception of anonymity. Ensure you have made some provision for employees with reading, mental or physical needs. Similarly don't forget remote workers, their views count too!

Creating an online employee survey is not that difficult particularly since the rise of tools such as Survey Monkey and even Google Docs. The difficulty lies in constructing the survey, writing the questions and attributing scores. Start on paper with the areas you want to cover such as

• Work load

• Relationship with manager

• Role clarity

• Work life balance

• Relationship with colleagues

• Bullying

• Stress

• Remuneration/compensation

• Working conditions

• Development

• Career prospects

• Employer branding

Ensure you are not going too broad, as it will require at least 5 questions per heading to get a meaningful picture and that can lead to a lot of questions and ill thought out responses.

Question Construction

Will your questions all have a positive slant, negative or a mix? For instance;

• I am generally happy with the way my manager behaves towards me

• I am unhappy with the way my manager behaves towards me

There is a school of thought that by making them all positive statements you might be inferring that this is the correct view. However the opposite could equally influence people. A mix though can cause problems. If people get used to the fact that a positive response is to the right of the scoring section they might inadvertently click strongly disagree, when for that particular question they meant strongly agree. It also makes your analysis far more difficult as you will need to assign points to each scoring statement, e.g. strongly agree = 5 points. However in the previous example, it would be worth 0 or maybe 1 point. Some survey systems are clever enough to be able to cope with this but it might be simpler to have all positive statements in one section and negative in the next.

Watch Your Rating Scale

There is nothing more frustrating than a statement that doesn't really fit with the rating scale. So if you are using the same scale all throughout the survey, ensure that every statement can be reflected by that scale.

A typical 5 point Likert scale would be; strongly agree, agree, neither agree or disagree, disagree, strongly disagree. However If the statements are;

• How often does you manager meet to discuss your progress (needs a number)

• Would you recommend working here to your friends or family? (needs a yes or no)

• Colleagues returning from maternity leave are dealt with effectively (probably needs a don't know or not applicable)

Assign points

Make the points for each statement meaningful. If your statement is a negative one e.g.

• I believe bullying is common place here

Assigning 5 points to "Strongly Agree" is only correct if you remember on your charts, analysis and presentation, that in this instance a high score is bad. Also remember that sometimes a zero is required otherwise you would be giving one point for something that is very bad.

Add demographics

if out of 100 staff only 5 say they are bullied but they are all female, their views will get diluted by the average score unless you can filter out the ratings given by female employees. Do the same for different levels and possibly other diversity measures.

Multiple responses

If you are concerned that some people might complete the employee survey multiple times to bias the results, ensure that your system prohibits this.

Free text comments

Give the opportunity to add some Technicolour or comments to scoring, it will help with interpretation and design of focus group questions.

Focus groups.

Never simply do a survey. How would you interpret a very low overall rating for the statement

• My manager discusses my career prospects with me at least once per quarter?

It could be that the managers don't care about careers, or maybe they are deliberately holding on to their staff. Possibly the managers need training in how to have such conversations or of course it could be that the vast majority of respondents are second income earners, have no desire to progress and their manager knows this and therefore it was OK to not have such regular questions. A focus group can pick up these findings and drill deeper. This is the key to avoiding knee jerk reactions that waste time.

Having an external consultant run the employee surveys and focus groups is more likely to get open and honest responses so be careful in feeling that you have got to the bottom of tings if you run them in-house.

To see examples of surveys go to http://protostar-uk.com/EmployeeSurveys.aspx