Best survey practices: Characteristics of an outstanding survey
The ABC of an outstanding Survey
It is important to consider how you would want to be addressed as a respondent in a survey. You definitely don't want to be left wondering what the surveyor expects from you, so a clear and concise way of asking is a good starting point for everything. You're also probably interested in what the results will be used for, or what benefits you as a respondent will get from filling out the survey. The invitation message is where this message is conveyed to the respondent.
The actual survey then begins with the customer questions that draw them in. Put yourself in the shoes of the respondent. What is the best way to reach you? The goal is to have a high response rate from the target group as a whole, and low dropout rate mid-survey. Find a way to reach a sufficient number of the key respondent groups that are important to you.
What do you need in terms of data? Are the objectives of the survey clear to you? What are you looking for, a low churn rate, new keys for company growth, more leads, better product market fit?
The survey report is simply a combination of the questions asked and the background information, but it is good to take a moment to think about your information needs when you are designing the survey.
We have collected the best practices that have emerged over the years in customer work in the following section:
Here are some clear points to consider when designing a survey:
- Language. Who are you asking from and how should the questions be asked so that the question is unambiguous to the target group? You know your target group (a certain customer group, staff, line managers, stakeholders in group X, ..) so use this knowledge the same way you use it for other communication purposes. A survey is communication and the same rules apply to it as to your other communication.
- Motivation. Balance the scope of the survey according to the motivation of the respondents to answer the survey. Think of this as a trade-off. The respondent gives you 1-30 minutes of their valuable time and the longer it takes, the clearer it needs to be that there is a benefit for them.
In customer surveys, the number of questions is rarely more than 10, there have even been surveys with only 2 questions. Such a survey only takes a minute or two. On the other hand a staff survey can take half an hour, but then you need to think carefully about how you communicate the improvements from the survey to the staff.
A survey with impactful results from the previous year motivates respondents to answer this year. You can also increase motivation (and the number of questions) with competitions or other rewards, this is surprisingly effective, even though it might seem trivial.
- Question types. Now for the big reveal: if you want a high response rate, make the first question a graphical question, a two-dimensional or a slider question, which makes respondent experience as fluent as possible. Extensive university research has shown that the gamifying features that graphical questions offer make respondents answer more questions than traditional "tick in the box" surveys.
In addition, the response data is much more accurate and usable in the reporting phase. You can also ask other question types, such as background questions and open feedback. However, the survey's appeal to respondents is built on graphics and gaming. And don't worry, we can make this look just like your graphical interface.
- Distribution. How do you reach the target group you have chosen? Does an email invitation or an SMS work, where do you need a wide range of background information about the respondents? Or do you want to make the survey completely open, with the same share link for everyone and possibly ask some background information during the survey? Or is your target group users of a certain website, in which case you want to embed the survey on the site or make a pop-up window?
You can also use a combination of these distribution methods, the important thing is that you take this into account to reach the widest possible range of respondents.
- Leading with knowledge. Is there anything else you would like to take into account when analyzing the results of the survey? All the questions and background information that you have collected so far for the survey, as well as the contact information, are available in the report. Can you predict if there is anything else that would make the data even more useful? If there is nothing to add, you are ready to go ahead with inviting the respondents.
Have you tried AI powered Zeffi survey tool?