5 tips to make your next survey work harder
05.10.18 Brianna Ripp
One of the best ways to get into the minds of your customers is to ask their opinion via surveys. Surveys are awesome tools to help improve a product or service, measure customer satisfaction, deliver actionable insights and improve customer retention.
I recently attended CXFusion; four glorious days of learning about fostering a customer-first culture. Here are some tidbits to help make the most of your surveys.
Design surveys with your customer in mind
In today’s day and age, we do business from the palm of our hand. And the same can be said for taking surveys. A survey showed, 60% of B2C respondents used mobile to complete surveys, with 54% of that being from a smartphone.
For B2B it’s a slightly different story with only 10% completing surveys on a mobile device. Due to the nature of the B2B world, it’s more likely they’ll provide you feedback via a desktop, so keep this in mind when laying out your survey.
Surveys not designed with the user in mind can:
- Lower response rates, impacting the sample size that informs your action.
- Skew results if respondents are not able to see all options in a question.
- Lead to fewer completions over time if the design hiccups go unchanged.
When creating a survey, there are five areas to consider: design, question presentation, survey length, testing and the email invite.
Tip 1: Whenever possible, use responsive and adaptive design to enhance the experience.
Tip 2: Consider your use of images and backgrounds.
Heavy use of images can give the impression the survey will take longer to complete. In some cases, it could be perceived to be up to 10 minutes longer than what it actually takes.
A lot of images can also slow load time. We’ve all been there – frustrated when our phone took 3 seconds to load instead of the usual 0.5 seconds. Keep your respondents happy – your survey results depend on it.
Tip 1: Don’t underestimate the value of a prompt.
For open-ended questions, add a voice-to-text prompt ahead of the question; this has been shown to double the response length, giving you greater insights.
Tip 2: Keep it horizontal.
For scale questions, keep the orientation the same as you would on desktop. If a scale doesn’t fit or reduces readability, see ‘Tip 1’ and add a prompt to turn their phone horizontal.
Fun fact: The impact of length on completion rate varies by device category – for non-mobile devices it’s around 8% decline for every additional 10 minutes in length and 20% for mobile device users.1
If you’ve ever taken a survey or provided feedback, you know from first-hand experience that the shorter the survey, the better. With shorter surveys come high completion rates.
If you must have a longer survey, consider chunking out the survey based on the device a person is using. If they are on a desktop, serve them the entire survey; if a mobile, offer up a shortened version that will deliver you the most insights.
Honestly, this should be rule #2, right behind ‘design with the audience’ in mind.
I cannot stress enough the importance of testing. Test your survey on all browsers for desktop or mobile (Safari, Chrome, Explorer, etc.). If you’re designing with mobile in mind, also test your survey on all mobile devices.
In my opinion, this overlooked often, yet makes all the difference.
Once you have a survey, you’re going to need an email to go with it. The purpose of this email is to hook and entice the user to start the survey.
Tip 1: The CTA belongs above the fold and immediately visible. If your customers must scroll, they won’t get to it.
Tip 2: Test an embedded question. Instead of having “Start”, fold in a question to the email that starts the survey. (E.g., “Do you currently own a potbelly pig?”)
In this ever-evolving, customer-driven world we’re living in as marketers, it’s important to take the time to view everything from your customer’s’ lens. I hope you’ve picked up a couple of quick tips that will help you put your customer at the heart of your surveys.
And if you ever want to nerd out over customer-centric approaches, let’s connect!
1MaritzCX study on survey length