Design a survey that customers want to complete
Surveys are a fantastic tool for customer research. You can get lots of information in a short time from a wide range of people, in a form that's easy to analyse.
You can get both quantitative and qualitative feedback in a survey:
Quantitative information is specific and measurable. Things like ratings on a scale, yes/no, age, or salary bracket.
Qualitative information is descriptive and subjective - things like free-form comments.
There are some things to be wary of. If you don’t ask the right questions, you can end up knowing what problems people have, but not why they have them. And if you don’t plan your distribution, you can end up with a skewed sample of responses, or no responses at all.
Improve your chances of getting the data you need by:
- Designing a survey that is easy to complete
- Getting it in front of people interested in your topic
Here are my tips to help you do just that.
Set clear goals
What questions are you trying to answer with your questionnaire? Are you looking to understand your current customers, or explore new markets? If you’re surveying existing customers, are you measuring customer satisfaction with your existing service? Or gauging potential interest in a new offering?
Write down your goals before you start to draft your questions. Avoid tackling too many goals in one survey. It’s better to do several short surveys on different topics, that one giant questionnaire that takes half an hour to fill in.
Match the survey to the audience
Once you know what you’re trying to achieve, you can work out who you want to fill in your survey.
If it’s your own customers, that’s easy - assuming you have a list of your customers, that is. If you don’t have any customers yet, or you're trying to research a different group of people, then who are you looking for? What are they interested in and why would they want to complete your survey?
Jot down ideas about where they hang out, too. Maybe in physical locations, or on social media, or on online forums? You’ll need this information when you come to distribute your survey.
Start on the right foot
Start to design your survey on paper; use a Google Doc or sticky notes, whatever works for you. Don’t start in the survey tool straight away. You’re going to do a lot of drafting and revising, so pick a tool that makes it easy to do that.
Begin with a nice introduction. Who are you, what are you asking, and why should they answer? Lots of people like to give their opinion. They also like to feel they’re having an influence on a product or service they care about. Persuasive introductions play to that desire to feel valued and influential.
Use ‘I’ and ‘we’ and ‘you, and stick to plain English. No jargon! You should also include any special instructions for filling out the survey. It’s good to tell people how long the survey will take to complete. You should also reassure respondents that the information they give will be safe in your hands.
Some people like to offer incentives - like entry into a prize draw, or a discount voucher. If you do that, make sure you add a mechanism for claiming the incentive at the end of the survey.
Let them give their opinion
As you start to draft your survey, balance quantitive questions like yes/no and rating scales, with qualitative feedback in the form of ‘other’ options and comment boxes.
For most questions where there's a list of options, you will need an ‘other’. There will always be an option you haven’t thought of. Use comment fields too, through the survey and again at the end. Don’t worry too much about comment boxes making your survey take too long to complete. People don’t have to use them, but if they want to say something and there's no comment box available, it can be frustrating.
You’ll want to include some demographic questions at some point: age, gender, location, income. It helps you to sort and group responses. Make sure you phrase these questions appropriately. Male and female aren’t the only genders. Age groups need to be inclusive. I once filled in a survey for a music app, which grouped age thus: ‘under 18, 18-24, 25-34, 35 and over’. They may as well have said “old people needn’t bother filling this in”.
Be sensitive to the fact that some people don’t like giving this data at all. Allow them to opt out of the demographics with a ‘rather not say’. Otherwise they might opt out of your survey entirely.
I generally put demographic questions at the end of a survey. When someone has got that far they tend to feel they’ve committed so much time they may as well finish it off. Start with the intriguing questions about the topic itself to hook their interest.
If you want to collect emails for follow-up interviews then you should do that at the end, too. And don’t forget your incentive, if you offered one.
How long should a survey be?
Test your survey at this point for logical flow and for time to complete, on as many people as you can. There’s no right answer for how long a survey should be. If you want to take more than 5-10 minutes of people’s time you might need to offer an incentive. But if it’s the right audience, and a powerful incentive, then you can get 20-30 minutes out of people.
Time to complete isn’t just about the number of questions. The type of questions, the amount of thinking required, and the survey software all play a part. If you are asking about what people have done, like “Do you shop at any of these stores?”, that is usually quick to complete. If you want people to give reasons or opinions, or speculate about what they might do in a given situation, that takes longer.
If your survey is a follow-up to an earlier transaction, keep the survey in proportion. Don’t ask someone to do a 10 minute survey after an online purchase that took only 2 minutes.
Choose the right survey tool
If you’re planning to distribute your survey in a physical location, you'll need to print it. Formatting it in Google Docs or another word processor will be fine. Make sure you leave enough space for people to write their comments.
If you want to distribute the questionnaire by email or on social media, you’ll need an online tool. This also has the advantage of delivering the results electronically. Most tools have built-in reports and data export facilities.
The right tool for you will depend on your budget, the kind of questions you want to use, and whether you need branching. For example, “Question 5: Do you eat cheese? If no, jump to question 10.” Some tools handle this better than others.
Also consider if you are likely to get a lot of respondents using smartphones or tablets. If your customers typically visit your website on tablets, they’ll probably want to fill in your survey on a tablet too. Some software works better on mobile devices. If in doubt, go with a mobile friendly option.
A nice feature of some survey tools is audience analytics. They show you how many responses you got, how long they took, and what device the person used.
Available features change all the time. That's why I’m not giving a specific list of which tools to use and when. Take advantage of free trials to test out the different options.
Test, test, and test again
Proofread your survey before you put it in the tool, and again afterwards. Get someone to check the survey in the tool against the draft - it’s easy to leave out entire questions.
Do final timing checks before you send out the link. Get lots of people to test on phones, tablets and desktop browsers. Once you’ve sent the survey you don’t want to change it.
Promote your customer survey in the right places
This isn’t a one-shot thing. You’re going to need to email, tweet, and post your survey everywhere you can to get maximum interest.
Draft a whole heap of messages in one go, changing the wording each time. Make sure you mention the topic and the kind of person who would be interested. If you’re offering an incentive, remember to include it in your messages. Ask people to pass the link on to their friends.
Use a scheduling tool like Buffer to space out tweets over the days and weeks, so you can set and forget your promotion. You can also use hashtags to attract attention from special interest communities. Try to get retweets from people who have a large following that matches your target audience.
You can also use specialist forums to promote your survey. Some won’t let you post links at all, and some ask you to be a member for a while before you post anything, so plan this part in advance if you can.
If you’re handing your survey out in a physical location, be thoughtful, not annoying. If you want people to complete the survey there and then, have lots of pens. In some cases, it might be easier to hand out cards with a link to the survey online, rather than the survey itself.
Save yourself some time with a swipe file
If you use surveys regularly, you'll find yourself asking the same questions over and over. Save time with a 'swipe file', a document of reusable text to copy and adapt for each new survey.
And to save you even more time, here's one I made earlier. (Just a download, no signup required.)