Photo Credit: iStock.com/Artist’s laflor
Survey design is easy!
Writing questions is a piece of cake!
I analyzed the data in no time! Said no one ever.
Let’s face it survey design is tough, you know it and we know it. Everyone also knows that surveys are important for business and thus can’t be avoided. We decided to create a guide that will make your survey design efforts a little less painful. This guide is not a run-of-the-mill list of best practices; instead, it takes the opposite approach.
We have seen some horrendous surveys in our time. Endured horrors including endless streams of questions, poor grammar, and vague questions. All this was done in the quest to find common features of bad surveys. We came up with a gigantic list of what makes surveys awful.
Let’s dive in and learn what great survey makers don’t do!
Never hound your customers with a survey every time they use your service. No one wants to fill out surveys frequently. Find a suitable time between surveys, it usually depends on what industry you’re working in. About once a month is the minimum limit, cross it and you will frustrate your customers.
You have to walk the line between getting the information you want but not get too personal. Don’t force the respondent to share what they don’t want to share. Err on the side of caution and focus on the useful rather than personal.
Easy means the customer wont have to make much of an effort to understand the questions. Make it clear and help the respondent in any way you can.
Research shows that customer satisfaction does not correlate with future customer loyalty. Instead, focus on your Net Promoter Score.
Figure out why you need to survey and what you will do with the results. This will help you fine-tune the entire design process.
Surveys can include different types of questions such as open-ended, multiple choice, one-choice, ratings, rankings, etc. Give a little thought to what kind of question will get you the right information. Usually a combination of question types is the best.
This applies to multiple choice and one-choice questions. A random list is confusing to skim, follow a logical order so the respondent can easily pick their answer.
Using biased or leading questions will affect the accuracy of your results. This will eventually make your survey efforts useless. Stay away from judgmental language, neutral terms are the best.
This is one of the leading causes of survey abandonment. It might also cause your respondents to give inaccurate answers on purpose. Redundant questions are the best way to make your respondents lose interest in completing the survey.
Chances are whatever questions you are trying to answer; someone has already done similar research. Utilize this information instead of trying to start from scratch. Once you know the general information, simply customize your survey to make it more focused on your niche.
Today’s generation has a short attention span; long surveys will have high abandonment rates. Keep it short and sweet to minimize respondent fatigue. Another good trick is to display an estimated completion time and the total number of pages in the survey header.
Question choice is the lifeblood of surveys. Whether you are a business or non-profit, these two questions must be a part of the list.
These two simple questions will help you discover the general perception of your business and pinpoint if any resources are being wasted during delivery/implementation.
People often miss the word ‘not’ when skimming through questions. To avoid getting misleading or inaccurate data, phrase questions using positive terms.
Haphazard distribution methods such as placement in newsletters, attaching to utility bills or handing them to random people at grocery stores is a major mistake. It might be convenient , but your results won’t reflect the opinions of the target demographic.
SLOPS or Self Selected opinion polls are highly suspect because people with negative views are more likely to respond. A balanced and accurate view is not possible with SLOPS.
If you have a scale of 1-5, with 1 being worst and 5 being best, use this for all the rating questions. Inconsistent scales will drive away your respondent.
The time chosen to send the survey can affect the response rate. Before sending out the survey think about when respondents are likely to check their email. Also consider how long your survey will take, and try to avoid the bust middle of the week days.
These are questions with the word ‘and’ in them. These will mess up the data collection, it’s better to rephrase them or split them up into two separate questions. Research (Bassili and Scott 1996) shows that double-barreled questions increase clarification requests.
Your clients are just that, clients. Treat them with respect and use formal language. Use courteous and friendly language and don’t get too familiar.
Stick to first-person opinions and views. Don’t ask them what they think of the general perception of your product. This line of inquiry wont get you any insightful data. Its off-putting if the customer has to analyze and reflect before answering.
Sometimes bad examples are a useful way to learn good practices. Bookmark this list of survey don’ts and you will never design a bad survey again. Experts know that survey design is a science and art; you have to work at it even if you know the fundamentals. The great business insight increases your revenue is out there. All you have to do is ask the right questions at the right time from the right people and in the right way.