When you design an online survey, there’s a temptation to ask respondents a lot of questions. After all, you surmise, this is your own opportunity to pick their brains! But people rarely are motivated to answer long surveys. How can you get all the data you need without chasing people away?
Few people write surveys for a living. Most businesspeople have their attention on doing their own jobs, such as organizing a successful sales conference or running a healthcare organization. As a result, you are unlikely to be adept at survey design. But we’ve done this many of thousands of times, and we’ve seen what sort of surveys attract responses… and, sadly, the sort that don’t.
You want to design surveys that gather useful data and help you make good decisions. But as we have seen – and heard, on our tech support lines – people often are frustrated by the process of writing good questions. Obviously, the quality of the answers you receive is affected by the questions you ask.
But another important factor is the survey length. You might want to ask respondents about everything possible, in great depth – you want more data to work with! However, most respondents lack motivation to participate. Unless you have a motivated community or offer some kind of valuable incentive (valuable to them), you’re lucky to capture respondents’ attention for more than a few minutes.
On the other hand, you can’t ask too few questions, or you won’t gather enough data to make good decisions. To use an only-slightly-absurd example, there’s no point in asking “How satisfied are you with the product you purchased” if you don’t record which product the respondent purchased.
Ask only the questions you need.
The key takeaway is to focus on the information that helps you make a decision – and leave out everything else.
Sometimes you have only one real question to answer. The vast majority of customer-satisfaction insight comes from answers to a single question: “How likely are you to recommend [this item or service] to a friend or colleague?” (It’s referred to as the “Net Promoter Score,” and yes of course it’s included in our question survey bank!)
I once spoke with a major magazine publisher who had stuffed her survey full of all sorts of interesting demographic questions. If the 1,500 people she targeted had answered the entire survey, she would have had a fascinating picture of her audience. But the long survey would take 20 minutes for a respondent to answer, and there was no particular reason that a reader would do so. If she got 100 responses to that survey as-written, I would have been surprised.
I began to query her about each of the questions she included. “It’s fine that you asked about gender,” I said. “But let’s say you discover that 70% of the readers are female. How would knowing that fact help you make a business decision? What would you do differently, with that number in mind?”
She admitted that she didn’t know. It was just interesting.
There’s nothing wrong with “interesting” or “nice to know.” But there’s a lot wrong with 80% of the survey respondents abandoning it partway through.
For every question you include in the survey, ask yourself, “How will knowing the answer help me make a decision?” If you don’t have a good response, leave it out.
Consider fielding more than one survey.
Do you need to get the same information from everyone? Depending on the number and nature of the questions you ask, you might consider whether it makes more sense to run, say, three separate surveys.
Sure, you can ask the same handful of base questions on all three surveys. But distribute the secondary questions across the three survey versions. For example, on SurveyA ask for product satisfaction and gender; on SurveyB ask for product satisfaction and income range; on SurveyC ask for product satisfaction and employment status.
This isn’t helpful if you need to cross-tabulate the data to compare gender and income range. But if you simply want to get a snapshot of the relationship between customer satisfaction with the respondent’s gender, it works great. (Plus, with a Sogolytics Enterprise account, you can merge multiple surveys’ responses, so you still have quite a bit of analytical power.)
Pre-populate the survey with data you already have.
Another way to minimize the number of questions you ask respondents is to use the data you already have collected about them. As I explained in How Data Pre-Population Can Improve Your Surveys, when you already have information about an individual, you don’t have to ask the respondent to re-enter it.
Skip questions that aren’t relevant.
Online survey applications let you use branching and skip logic (also known as “conditional branching” or “branch logic”) to hide questions that don’t apply to a given respondent. For example, you can ask respondents if they own a swimming pool; if the answer is “No,” you can hide follow-up questions about pool care or pool supplies.
That makes the survey visually shorter, which encourages respondents to stick around. It has other benefits as well, such as ensuring that you collect useful information (why collect pool supply answers from people who aren’t qualified?) and making sure you don’t waste anyone’s time.
If you’ll permit a small “sales” aside: This kind of logic branching usually is available only in the paid versions of online survey tools. In Sogolytics, our free version permits you to include a single branch. That can help you explore the power of this important feature.
If you need assistance, don’t be afraid to ask for help.
Sometimes you don’t need to design questionnaires from scratch. For instance, you can start with a premade design suitable to your industry (such as Healthcare or Government), or choose from a survey bank of common questions. Just keep in mind that more questions is not better!
At Sogolytics, we pride ourselves on the quality of the tech support we provide. That extends to one-on-one help in survey design. If that’s a service we can help with, please get in touch with us about survey design services.