It seems like the week for surveys. I’ve done lots of them this week; some general market research type surveys and several customer satisfaction surveys. Some were crafted by expensive consultants, some were slapped together by the business owner. They were on different subjects and asked different questions, but they did have one thing in common — they all sucked!
Surveys are a great way to get inside the heads of your customers, clients, prospects, and target market.
And while there are several wrong ways to do surveys, there’s really only one right way: the way that gets your survey completed accurately.
And this is what these surveys failed to do — gather accurate information from as many people as possible. If you don’t have accurate information, your survey results are useless.
How can your survey gather inaccurate information?
Your survey can actually gather inaccurate information when you require a respondent to answer a question and then not provide enough variety in your answer choices to reflect the reality of all of your respondents.
Let me give you a couple of examples to illustrate:
If you are a restaurant and ask a question like this:
Please rate the condition of the restrooms on your visit.
- Outstanding! Everything was clean and the toilet paper was full.
- Great! There were a few pieces of paper on the floor, but it looked like the restroom had been cleaned recently.
- Poor. There was paper all over the floor and/or the toilet paper needed to be filled.
- Awful. The most disgusting restroom I have ever encountered. In fact, it was so dirty that I didn’t even use it.
If this person didn’t actually use the restroom, and you make this question required, you force them to randomly choose one of your options. Even if they randomly give you an Outstanding rating, this isn’t an accurate reflection of the state of your bathroom on that day.
Let’s look at another example.
Based on name recognition alone, which of these plumbers would you consider hiring the next time you need plumbing work:
- Bob’s Plumbing
- The Guy Down the Street Plumbing
- Quick and Cheap Plumbing
- Big Name Plumbing
Although I’ve changed the names of the plumbing companies here, this is an example taken from an actual survey that I completed.
Like many others who saw this survey, much like you now, I didn’t recognize all the names listed. However, answering the question was required so I had to choose one even though I had no basis for comparison.
This is a problem that I see often in surveys and it stems from the underlying assumptions the survey writer has about their subject. In the above question, the survey writer assumed two things:
- The names of the plumbers in their question would be recognized by all respondents.
- People use name recognition when choosing a plumber.
When we are close to a subject, it is easy to assume that everyone else has the same understanding that we do about it. Imagine talking to a programmer about computers and you get the idea. Simply becoming aware of this bias can help you take steps to overcome it. But when it comes to creating your surveys, its best to take it one step further.
Give Your Survey a Critical Look
Before sending out any survey, give it a critical look first. You can train yourself to do this, or you can recruit someone to be an impartial set of eyes for you. When you are first getting started with surveys — or when the results really matter — its best to have someone help you who you trust to give you honest and critical feedback.
When giving your survey a critical look, ask yourself the following questions:
- Have I written each question clearly? Is it obvious what I want to know about?
- Is there enough variety in the available answer choices to cover everyone?
- For required questions, have I allowed for the possibility that someone cannot answer my question?
In our two examples above, the survey writer would have gotten more accurate results by simply adding an option for those who cannot answer the question as written. For the restroom question, adding “Not applicable” or “didn’t visit the restroom.” For the question about plumbers, adding “not enough information” or even simply “other” with the option to provide more information would suffice.