Designing a survey doesn’t just mean picking the right colors and fonts (although those are important too). Your survey should be created with your goals in mind so that you have actionable data once all the responses are in.
It’s also important to be aware of common pitfalls to avoid, like asking too many questions or overwhelming your respondents with too many answer choices.
Fortunately, our team has run a few surveys and come up with these simple steps to turn your great survey idea into a great survey design.
This article will guide you on start to finish of the design process to help you make sure your survey is a success.
Begin Your Design Process: Organize Brainstorm Notes
As one of the very first steps in your survey creation process you should have done a brainstorming session with your team to identify your high level survey goals.
After the storm has died down and you’ve established your expected survey ROI and your specific survey goals, it’s time to revisit your brainstorm notes.
Identify themes in these notes, and use them to group ideas based on objectives and topics. Keep in mind that there may be some additional refinement that needs to be done at this point.
Translation: you may have to throw out ideas that aren’t relevant.
Once you have some topics, rank them by priority and categorize them based on how their results could be used in your survey reporting.
For example, if you are trying to identify who uses your product, group your product questions together early in the survey before asking demographic questions later.
Turning Ideas Into Questions
Here are some guidelines to keep in mind while you’re turning your brainstorm notes into possible survey questions:
- Keep questions brief, simple, relevant, specific, and direct.
- Do you want to ask qualitative questions (exploratory, usually freeform) or quantitative questions (answers are pre-defined)?
- Make sure your phrasing is relevant to the demographic you’re planning to survey.
- Avoid leading questions, suggestive questions, sensitive topics, highly technical language, and questions that cause survey fatigue.
Survey Formatting
The formatting areas of color, branding & images, and non-question text are where you’ll focus next. These areas give you the look and feel of your survey, and they have a substantial impact on the overall respondent experience.
If the respondent has a positive experience, after all, it’s more likely that you’ll get the data you are looking for.
Color Scheme
When choosing colors for your survey, you’ll want to consider the emotional impact each color has, and what type of impact you’d like to have on your audience. (Please see The Visual Language of Great Survey Design for more information.)
In selecting your color scheme, you’ll also want to keep accessibility in mind. It will benefit respondents with visual impairments if you stick to using high contrast colors that are easier to see.
Contrasting colors also have the ability to increase response rates in general for those without visual impairments as well.
Branding & Images
If you plan to use the color scheme of your company’s logo or website, it will save you a step from having to decipher color choice on your own. To grab your company colors, you can import CSS files from a particular URL, use a color picker tool, or type in the hex codes for your corporate colors.
If you’d like to add your logo to the top of your survey, be sure to keep the size in check. You don’t want your respondents to have to scroll down past your logo just to get to your survey questions.
Lastly, to maintain full accessibility for your audience, make sure to add captions or alternative text to your images. This will ensure that respondents can at least hear a description of the image if they are not able to see it.
Non-Question Text
The non-question text portion of your survey, forward & back buttons, welcome/thank you pages, and error messages are all important pieces of your survey that require attention.
They must all convey a consistent tone.
Avoid The Four Horsemen Of The Survey-pocalypse
No matter how careful you are with planning your survey, there are still some problems that can creep into your design and cause issues for your data.
Be aware of these four common pitfalls that can create havoc for your survey:
- Lack of Focus
- Bias
- Respondent Fatigue
- Miscommunication
Lack of Focus
Your survey can lose focus easily if you try to cover too many diverse topics (this is why we suggested limiting your learning objectives to five in our post on survey goals). Avoid questions that are not directly in line with your learning objectives.
You should also steer clear of any questions that don’t drive actionable results with their data. Finally, adding questions that you’re curious about, but which don’t meet the survey goal, can also get your survey off track.
Question Bias
Bias is a big problem with surveys because it can skew your data without you realizing it.
Oftentimes we aren’t aware that we have a bias and are expressing it in our survey. This is where having others proofread your questions comes in very handy to help point out the issues you may be blind to.
There are four areas of question bias to avoid:
- Emotional bias
- Option bias
- Identity bias
- Conversational bias
Bias can sneak into your questions very quickly, and it can wreak havoc on the accuracy of your results, but with vigilance you can steer clear of all four of these dangerous forms of bias.
Emotional Bias
Your own opinions about a topic shouldn’t show up in the question content or phrasing. This most often happens when we ask loaded questions or neutral-seeming questions about a loaded topic.
Example: Most Americans celebrate the 4th of July. Do you?
How to Fix It: Do you celebrate the 4th of July?
Option Bias
By limiting the types of answers available, you may be creating option bias. Make sure that you avoid:
- Required, non-applicable questions
- Leading or restrictive options
- Different types of scales
- Massive option lists
Identity Bias
This most often happens when companies are conducting a branded survey about themselves. Asking “How much do you love Alchemer?” or “Do you like Alchemer?” on a page with the Alchemer logo would definitely create identity bias.
Conversational Bias
While you want to use language that your respondents will understand, you don’t want to veer too far toward turning your survey into a conversation.
Then you risk respondents giving the answer that they think you want to hear, rather than their actual opinion.
Survey Fatigue
Survey fatigue is unfortunately a common cultural phenomenon. Most respondents are tired of the barrage of emails, phone calls, and text messages from retailers requesting them to take take their post purchase survey.
The prevalence of online survey tools contribute to this problem, because it is easy for anybody can create a survey. Take the time to craft questions that are considerate of your respondents’ time and effort.
Avoid Miscommunication
To communicate effectively within your survey, you should be sure that you know your audience and use the language that they use and understand. Avoid technical terms within your questions, unless they are appropriate and will be understood by respondents.
Define terms in the survey if necessary. Keep your company’s tone consistent if the survey will be coming from your brand.
Finally, make sure to have a peer review your questions for clarity.
Design Your Way To Success
Approaching your survey design process in a strategic manner will set you up for successful survey results you can act from. From beginning to end of your creation process, keep your survey data goals in mind.
Ask yourself, “What decisions are you hoping to make from this data?” and, “Who is the audience you are speaking to?”
These questions will help to guide in your creation process.
The common pitfalls to avoid will also help you steer clear of survey potholes along the way.
Regardless of your particular approach, be sure to get sufficient review and feedback of your design from your peers and stakeholders to make sure you are avoiding bias and are coming across the best way possible to respondents.