Surveys are powerful ways of gaining great insight into your customers. Well-managed surveys can help you discover who your customers are, how they behave (to some extent), and, most importantly, what they think and why they do the things they do. In the digital space, there are a number of ways that you can collect customer feedback using surveys either on the site itself or by e-mail, if you have the relevant details. The cost of survey data collection can be quite low, and there are many different survey technologies out there to choose from. But what makes a successful survey? Here are some thoughts.
KISS – Keep It Simple…
If you’re running surveys on the site itself, then aim for maximum simplicity. This will help with your response rates and with the quality of the data you collect. Try and be as efficient as possible in your data collection and always ask yourself when you’re designing the questionnaire what you will do with the data that you’re collecting. Do you really need to know?
Think about the effort involved in answering your surveys. Asking people to write in answers to open-ended questions can give you great data, but it can also put people off. The questions should be optional and limited.
Make It Engaging
Try and make the survey as engaging as possible. You don’t need flashing lights and bells, but think about how it looks visually, how it reads, and whether or not it’s interesting. If it isn’t relevant, then people won’t fill it in. So don’t focus just on the data you want to collect, but also think about the experience of filling in the service itself. Is it something you would want to do?
Don’t Ask Everyone
Typically, don’t invite everyone who visits the site to take part in your survey. Most survey technologies allow you to select a sampling rate, which determines how often the survey invitation will appear. With surveys, you need a certain amount of data for the responses to be reliable. I usually aim for a minimum of 500 respondents for the amount of data that I want to analyze. If you want 500 respondents, you’ll need to figure out how many people you’ll need to ask to get that many responses and then compare that against the number of people that visit the website to get your sampling rate. It will take some trial and error to start off with.
Test, Test, Test
Always test your survey before going live. There’s nothing worse than finding out afterwards that something didn’t work properly or you weren’t getting the kind of data that you hoped. By then, it’s too late. You can test it by doing it yourself, by getting colleagues to fill it in, or by making it live with a very low sampling rate. Look at the data you get back. Does it make sense? If you’ve put some routing in where only certain types of people answer particular questions, make sure that’s working in the way you expected. Once you’re happy, then make it live or increase the sampling rate.
Be Prepared for Feedback
You’re asking for feedback and you will get it, but maybe not in the form that you expected. Customers can see surveys as an opportunity to communicate rather than just to give feedback. Often if you have questions in your survey where respondents can write something down (such as “Can you tell us how we can improve the experience you had today?”), then they will tell you. That might involve a customer service issue, such as they tried to make a booking and failed and they want you to call them back.
Many times I’ve seen those sorts of comments from surveys, where the respondent leaves their name, address, and telephone details expecting the company to get back in touch with them. If nothing happens, that’s not good, so you should have a system that means you are regularly looking out for that kind of customer service issue.
Survey data is great, but it needs to be handled with care. All data has bias and error in it and survey data is no exception. Certain types of people are more likely to fill in a survey than others and so you need to take that into account when you’re looking at the results. Wherever possible, compare and contrast your results either over time or among different segments, such as gender, age, frequency of use, and so on.
Surveys are a great source of insight, but make sure that you make the most of them by getting good quality inputs.
Emily Ma, product director of Tencent’s advertising platform products department, was a keynote speaker at ClickZ Live Shanghai where she discussed the ... read more
In today's multichannel world how can marketers use data to ensure the experience a customer receives is relevant to them?
The terms that customers type into your site search function can help you to gain an understanding of user behaviour and can be used to optimise ... read more