Creating a winning survey with high completion rate, just like writing a good newsletter or making a sales call that closes, is part art, part science. These 7 tips (and a bonus one) will get you to grips with both.
1. Pick the right audience
Narrowing down your target audience will do wonders to your completion rates. Knowing who will receive the survey will help you tailor the message, choose the right incentives, and time the survey just right.
If you have your own panel or customer list, think of your most important customer personas and tailor everything to them. Don’t try to please everyone.
When buying your audience from someone like Ansr.io (or using mobile ads to get people to land on your survey) then narrow down the panel by gender, age, are they a smartphone or feature phone user, then start with one or two screening questions to get only well matched people to answer your "real" questions.
2. Good incentive
People who take your survey give you their most valuable resource - their time and attention, not to mention their opinion. You have to earn it and offering good incentive goes a long way.
Pick a tight group to focus on and coming up with the right incentive becomes easy or at least easier. It doesn’t need to be anything super special or expensive but it needs to match the group and effort involved.
Making beauty products and need women to fill out your survey? Offer gift cards or a discount code.
An alternative is to turn it into a lottery and promise a bigger prize for one person. We’ve seen both models work well in different scenarios.
3. Have a killer message
Bill Drummond of the KLF fame knows all about having a killer message
In our typical use case there are two places where killer copy writing skills can really impact the survey results:
SMS message copy: Most of the times this is how people are introduced to your survey so make it count. Short, relevant, and action oriented text works the best. Answer the "what?", "what’s in it for me?" and "what’s the next step?" and you have a winner.
Survey landing page: Repeat the main points - what’s the survey about, point out the incentive and indicate how long the survey will take. If you think the SMS did good enough job then it might make sense to pass this step and link SMS directly to the first question page.
Same rules apply when sending the survey out via email or promoting it in your social channels.
4. The science bit: AB test your message
Not AB testing your message and your survey? You’re missing out and wasting people’s time and your own money. AB testing on a smaller segment before sending to the whole audience is the easiest way to increase the number of completed surveys.
- Clone your survey into two separate surveys
- Write two messages
- Send both to 10% of the audience
- Wait 2 hours and check which message drove more respondents and completed surveys
- Send winning message to the rest of the audience (80%)
5. Science bit 2: Test your survey
Asking 300 people to take your survey only to find out that your key questions were set up in a way that most people ignored them is not good for business.
This is why you should test first with 50 and check if any unwanted results pop up. A customer recently discovered that their open ended questions were not giving high enough response rate - only a few people bothered to answer their last two open ended questions.
The customer cleverly reworked the question as a multiple choice and the questionnaire was sent to the rest of the audience. Slight loss in answer depth outweighed the fact that most respondents now answered those questions.
6. Time it right
Timing can play an important role in how much attention your survey gets. In general we’re seeing that immediate period after work (6-8pm) is best in the US, UK and continental Europe as people are commuting or have just reached home and are relaxing.
In Africa (and curiously enough also in Greece) the best time is earlier - around 3 in the afternoon.
7. Design it for mobile
The small screen, the fiddly typing - mobile is not your good old computer so design the survey accordingly.
There's also more and more proof emerging from online panel companies that large proportion of people fill out the survey on their mobile device even if the survey was designed for a larger screen.
Avoid grid questions - so common in web-based survey, absolute nightmare on a mobile screen.
Interestingly open ended questions and drop-down lists don't pose such a hurdle for people.
8. Bonus tip - send a reminder!
People are busy. They read your message but then life happens and they promptly forget about you. Give them a gentle nudge a day or two later. Try slightly different time to send the follow-up message. They’ve seen your message once and if you raised any interest are more likely to take action.
Have any questions about survey design for mobile? We're happy to answer, just drop us a line.
