surveys and questionnaires - for staff or customers

It is important to know what your staff and customers think and need in relation to your organization.

Don't guess or assume - or worse tell them. Ask them.

A survey is the common method to discover staff and customer attitudes, needs, desires, problems, etc.

Usually a survey is based on a questionnaire. Market research companies can design and organize staff and customer surveys. So too can good telemarketing agencies. You might prefer to organize a survey internally due to control or costs reasons, in which case it's helpful to follow a sensible process. Even if you use an agency, it's helpful to understand the process.

Here is a quick guide for the process of creating and organizing a staff or customer survey - or some market research - based on a questionnaire. All situations are different, so seek other ideas and adapt your own plans accordingly. Seek help from colleagues and external people where possible in areas that you are not capable or confident.

To develop your questionnaire you first need to identify exactly the data you wish to discover.

Brainstoming this can be a useful start. You should also consult with all interested parties in listing your survey criteria. It's a lot of effort to design and manage a survey, so it's silly to miss something important because the early planning stage was rushed.

Here are the main steps to designing a survey of staff, customers or your market, using a questionnaire:

1. Decide and agree the purpose of the survey. Keep it as simple as you can. There is a temptation to expand surveys into additional sectors and subjects, but this normally dilutes the usefulness of the response and the resulting analysis. It helps to concentrate on the key issues for your essential target group. In this respect, surveying is rather like marketing and selling. If you spread your efforts too wide and thin your results will be wide and thin too.
2. Decide your target respondents or audience or market sector or staff audience. Ensure that your target respondent group is relevant to your survey subject, and satisfy yourself that you can identify and reach the target group via whatever communications and survey method you choose.
3. Decide the level of privacy and anonymity which is appropriate for your survey. Many surveys work better if conducted anonymously. On the other hand, a survey of business customers generally works far better if respondents are known and given the opportunity to express specific views from their own particular standpoint.
4. Decide the minimum response (number of completed questionnaires) that you need for a useful sample. For business customer surveys a minimum of 100 responses is an acceptable number provided respondents represent a suitable cross-section of the relevant target audience or customer base. Consumer surveys tend to require several hundred respondents for very useful results.
5. Organize your survey to allow for the anticipated response rate. For example anticipate a low response rate (between 2% and 10%) if the survey method is passive, such as postal or email or web-based. More proactive methods like telemarketing give a higher response rate (assuming the contact list is reliable you can work on about 20-50% response from the contact list - and be guided by the telemarketing agency if you use one). For general consumer market research surveys via street or door-to-door interviews again consider that most people decilne to take part, and therefore you should build a low response expectation into your planning of numbers and time. The highest response rates are from focus groups (basically a focus group is an arranged meeting for interviews and discussions, usually combined with a questionnaire) which by their nature enable 100% response. Interestingly the other hugely ignored opportunity for very high responding surveys is complaints and grievances from your target group. Think about it... complaints and grievances are an extremely useful source of valuable feedback and views, which ideally should be incorporated into any survey project. It's a terrible waste not to.
6. Decide the survey method(s) - email, internet, telephone, written document, focus group discussions, street surveys, door-to-door, or combination of these - whatever will fit your situation and tagret group best. Consider the reply mechanism if one is required. For example include postage-paid addressed envelopes. Or for internal staff attitude surveys consider tasking someone to encourage and collect replies. Whatever - make it easy for people to respond.
7. Consider incentivising or offering prizes to survey respondents, or even a payment - especially to focus group members. It's very frustrating to put the time and effort into designing and running a survey only to find that you get a response that's too low to be useful. People are very busy and mostly are not prepared to give time in responding to questionnaires, even if it's in their interests to do so. For passive survey methods (for example postal or internal mail) expect response rates to be less than 10%. Sometimes they can be less than 1%. Business customer surveys work well if postal questionnaires are supported by telephone introduction to explain the survey purpose, then followed-up ('chased') by telephone too if necessary.
8. Design the actual questionnaire: List the individual questions/issues. At the earliest possible stage it helps to build the survey onto a spreadsheet - this enables data and structure and scoring, etc., to be organized much easier than in a text editor. Try to create a natural flow or sequence in the questions. Use closed questions (yes/no) where useful, and offer multiple-choice answers, and avoid giving a bias to the questions influenced by your own assumptions, or the CEO's personal views.
9. Then create questions - seek expert help with writing the questions - it's important to get this right. Questions that seem clear to you might be confusing to people far removed from the project. It's crucial to frame the questions objectively and clearly so that they can be quickly and clearly understood by the reader. Clear questions also maximise response rates. Confusion and lack of relevance in questionnaires are big reasons for people not responding. Effective questionnaires must be easily and quickly understood, so test your questions on someone who knows nothing about the situation, even some young teenagers (arguably the most difficult audience of all), to check that your intended meaning is properly and quickly understood.
10. Devise a scoring method and design this into the questionnaire format. Analysis of results is very difficult and time-consuming if you fail to consider this properly. Ideally you must be able to convert answers into numerical data to make analysis quick and reliable, especially if your survey is large. If in doubt seek help from a spreadsheet expert. Finance departments in organizations usually contain such people, who are often delighted to help with survey projects because they are interesting and connected with the customers and/or staff side of the organization. Spreadsheets enable all sorts of clever analysis if you know how to do it, and it helps greatly for good analytical functionality and structure to be built into the design of the spreadsheet from the beginning.
11. Write a suitably appealing supporting explanation of the survey's purpose. Also take care with the questionnaire instructions, and also give some details about the follow-up process. People are more likely to respond if they can see and understand a meaningful purpose and follow-up for the survey. Getting a good response to a survey is always challenging, so the better your supporting explanation then the better your response rate will be. A survey also helps towards positive staff/customer relations exercise - it shows you are interested in their views, so make the most of the opportunity to communicate and explain.
12. Consider and decide about publishing the survey analysis (or a summary), and how best to convey results and follow-up actions to the respondents and other interested parties. This is especially important with surveys of employees. For certain types of market research or attitudinal surveys consider also the PR (Public Relations - publicity) value and opportunities arising from your survey. Subject to rules of privacy and agreement with your respondents, a survey commonly makes excellent press editorial and publicity.
13. Test the survey and method(s) with a small sample of people, preferably representative of the actual target group. Check that the scoring and analysis can be done. This is especially important if the survey is large, expensive, and/or crucial to the organization's strategy and decision-making. The need for testing is one very good reason for planning surveys sufficiently in advance of the deadline for getting the results.
14. If you test the survey, obviously refine the questions and structure and survey methods appropriately.
15. Run the survey. Monitor its operation. Don't wait until the end to discover a problem that you could have fixed at the start. If you use an agency check their progress soon after they start, and again at suitable intervals, depending on the size of the exercise. Again don't wait until the end to discover there was a bloody great big spanner in the works.
16. Chase up the replies using telephone follow-up where necessary. This is another reason for monitoring progress: commonly response levels fail to be as high as planned, in which case the earlier you are able to add some extra impetus the better.
17. Analyse the results and implement follow-up actions as appropriate, which if appropriate must involve giving agreed feedback of results and outcomes to respondents. If you are struggling with the analysis because the format was badly designed, it's still not too late to call in some help from a spreadsheet expert, rather than making a mess. If the data is there in one form or another, a good spreadsheet person can often achieve a minor miracle and save the project, or simply save you several days work.
18. Write up the report fairly and objectively, and circulate it as agreed, especially if it throws up a few nasty surprises, which are actually the most valuable survey results of all.
19. Ensure all specific complaints and matters arising from individual customers are followed up reliably and satisfactorily.
20. Review the survey project overall and incorporate lessons and improvements next time.

Tip - a good way to understand how to structure questionnaires and write survey questions is to see how other organizations do it. Look at the various survey materials which you receive yourself - through your letter-box, in new products that you buy, at airports and stations, in magazines - they are everywhere once you look for them.

See also the notes on designing and managing an employee motivation survey. Essentially this focuses on understanding staff motivational attitudes, but the guidelines also include useful techniques and rules for surveys and questionnaires in general.

The training needs analysis methods are also useful for understanding and designing surveys, and the TNA spreadsheet tools can easily be adapted into more general questionnaires for other purposes.

Another example of a questionnaire is the Multiple Intelligences Test materials - which provide further examples of how to design survey questionnaires.

The personal strengths indicator is another (very basic) example of a simple survey format, which is fine if the survey is small and does not require a lot of statistical analysis.

While analysis and structure are vital in big surveys, ultimately what's most importnat is simply taking the trouble to ask for people's views about important issues, rather than guessing or assuming, or telling people what you think they should be.

Well designed and implemented surveys always produce a positive effect for the organization. People - whether employees or customers - think better of the organization for being asked and consulted, especially if they see you've listened and done your best to react positively to the feedback you've been given.