TO SURVEY, OR NOT TO SURVEY? EXISTENTIAL QUESTIONS ASIDE… HERE’S HOW TO CREATE A QUESTIONNAIRE:

You’ve heard that creating surveys can be a useful method for gathering feedback. You are motivated and ready to go. Great! But, wait… how to get started? Well, this article is hopefully going to shed some light on the issue.

I cannot tell you the types of questions to ask, because I cannot know what it is that you are looking for. I can however build a helpful, general, guide to give you a push in the direction of survey creativity.

There are two umbrella steps to this process: First, background construction of the project, and second, elaboration and design of the survey.

Background construction:

You would not go to a championship game without a game-plan, or strategy, well you would not create a survey without one either.

What is the question?

Your first step in building up your background or strategy is to determine your goals, set out the issues you are striving to address, and ask the questions you want answered. You need to determine precisely what you are looking for.

Unfortunately, questions like: “How can I make more money?” or “How can I improve my business?” are not going to cut it. They may be the overarching issues you have in mind, but you have to be looking for more pointed answers to more focused questions: “What is the likelihood of purchase at a specific price point?”

If you already have a set of questions that were sent for you to answer from the powers above that be, then great! This step is done.

If you need to come up with your own, then try getting a few people together in the office and write down all the questions you might want answers to. When you have an exhaustive list, strike out the ones that are very obviously off-topic (“Where is your co-worker Harry going on vacation next week?” is not a valid question unless you are creating a survey about your co-worker Harry). Now cut the list of questions in half. Yup, all those questions you spent hours trying to come up with… lose half of them (keeping the more important half, obviously). Now comes the fun part: do it again. Keep halving the list until you have a maximum of 8 to 10 topics you are looking to answer or address. This step does not mean that you will only have eight questions in your survey. In order to get a comprehensive answer to one of these questions you will need several survey questions around them. Also, a survey always includes qualifying questions and such in addition to your main topics (the gender and age of your survey-taker are probably not going to answer your main issues, but they are still imperative parts of the survey, for example).

The thing that is important to remember is that in order to get meaningful results you need to concentrate on asking a few important topics well, rather than trying to answer too many questions and doing it superficially.

Equally as important to keep in mind is that by having pointed questions and knowing what you are looking for means that you will be better able to create a survey that is best adapted to getting those answers.

Elaboration and design of the survey

Composing your questions:

Asking a question is harder than it seems. There are many ways to get it wrong, and unfortunately asking the wrong questions or even the right questions in an incorrect way will compromise your results. The most important thing to remember is that you need to make sure you are asking the question you mean to be asking. I know, it sounds weird and cryptic… It is not rocket science, but you are going to have to put a lot of thought into your question formulation. To that effect, here are some things to look out for:

1. Ask only one question per question. It’s an obvious one but, trust me, a lot of surveys get this wrong. Multiple-part questions will only lead to confusion and will muddle your results.

2. Make your questions neutral. Try not to unduly influence your survey-takers by asking a question loaded.

3. Avoid using words that have several meanings or several connotations. You cannot control for survey-takers perception of ‘loaded words’ and they may also lead to confusion. In this same vein, it is important to take into consideration words that may have a different meaning for different generations.

4. Avoid language that is too familiar or colloquial. This type of language will alienate part of your sample audience. (One exception to this is if you have a non-representative group that you have specifically chosen, who you know will speak in this way, or respond to that type of language. For example, a ‘trendy’ clothing website whose clientele consists almost entirely of 18-25 year-olds, might want to use familiar language to create a rapport with its survey-takers.)

5. Use simple vocabulary. Careful! Beware of condescendingly simple language. Your survey-takers will abandon the survey if they think you are belittling their intelligence.

It is really important that you read and re-read your questions. Test them on a co-worker. Re-read and re-write them. Test them on another co-worker, edit, adjust…. You get the idea. To make sure your audience understands your questions and their meaning you need to have someone else look at them as well (someone who you trust will tell you if you are doing it wrong). You know what you mean, but will your survey-takers? Check you work and be critical of it.

Target your audience:

The next steps help you to structure your future survey. Here are the choices you will have to make:

1. Choose your sample population carefully. This is a crucial step, the number of people and the types of people you choose to send your survey to will strongly affect your response rate and the quality of your data.

a) Know who you want information from. “Simple,” you say, “my clients,” well unfortunately it is not so simple. If your category is indeed your clients at large then your survey will have to be tailored to a very wide variety of people and be full of general questions. It is also more difficult (but of course possible) to get responses on this type of population, because creating incentive for them to answer them is challenging. If you target an enthusiastic group of people who love a product, they’ll probably be more than willing to tell you everything about the product. However, this type of group does not tell you anything about the needs of future clients, only what ‘true believers’ already think. There are many types of groups you can ask questions to choose wisely.

b)Depending on the type of people you are questioning and about what, then you will want to use different types of publishing techniques.

2. Choose how you will publish your online survey.

a) Embedded onto a website. This can be useful if you are trying to gain feedback on a specific website or services provided there. If you are embedding it onto a Facebook page or Twitter feed, you have the potential to use social networking to your advantage. The disadvantage here is that you have no way of controlling who is answering the survey (though if you ask good filtering questions this should not hinder your results), or, worse, no one answers it at all because it was not shared or re-tweeted.

b) Sent out via email list. This can be a good way to contact clients already on a mailing list to get feedback. They may be more willing to answer a survey about services rendered for example. A disadvantage is that people may not want to answer your survey anyway.

Remember that responses to your surveys will not be 100%; they probably will not even be close. In fact, there may not even be 5% response rates. This response rate depends on a huge variety of factors including your choice of sample population and how you contact them, so do not get discouraged if you get low response rates. Just remember, you have to send out a lot more surveys than the amount of answers you would like to get back.

Tips for a better survey experience

Finally, after all this background work and question formulation you have to put it all together. Here are some tips that will increase your survey completions by creating a better survey-taking experience:

1. Use filter questions to avoid survey-takers having to take questions that do not concern them. Filtering questions are also indispensable for good analysis.

2. Make sure the order of the questions ‘flow’ nicely.

3. Fit questions into sub-sections. Put questions with similar themes or goals on the same page or the same section. Then organize these sub-sections logically so that they follow nicely from one to the next.

4. Do not make the survey too long. Think critically about whom you are asking questions to and what about, in order to judge how much time they would spend on your survey before abandoning.

5. Use your opening page to tell the survey-taker what your survey is about, why you are collecting this information, what types of questions they will be asked, and how long it will take them. (If it’s going to take more than ten minutes, then do not write ‘this will only take a few minutes of your time.’ Be specific.)

6. Also on your opening page, tell the survey-taker if they will remain anonymous. If they will be anonymous, saying so will increase their chances of answering honestly. (This is especially important if you are sending a survey to your employees to ask questions about their satisfaction, for example.) If the survey-takers will not remain anonymous it is REQUIRED that you say so. Be aware that this is an ethical issue and that it could become a legal issue for you if someone is unhappy with how you use their information.

7. Answers to a question should be exhaustive and mutually exclusive. A.k.a. in a multiple choice question make sure all possible options are available (but without repeats), when in doubt add an ‘other’ category with the possibility of filling in a blank. Careful not to make the survey-taker feel trapped into answering a specific answer choice.

8. Try to limit open-answer questions as they make data analysis difficult. Use these types of questions sparingly and as a method for gathering additional details about a previously asked question.

Finally, remember that a good-looking survey is more likely to keep survey-takers engaged therefore lowering drop-out rates. It should be aesthetically pleasing, but also practical. Try to put no more than four or five questions per page. Avoid any formats that will make the survey-taker feel overwhelmed.

Get to it!

Surveys are an important market research tool and can be extremely useful for your business. To be sure that you are gaining all the benefits from this tool you should carefully construct your survey and make sure your survey-takers will have a pleasant experience while taking it. This post is by no means an exhaustive list of all the things to look out for when creating a survey, but it is a good place to start. The more surveys you create the better technique you will acquire, so get to it!

Happy Survey Creating!