How to Write Clear and Effective Survey Questions for Accurate Results.
Quality output depends on quality input, and surveys are no exception. Do you want your output to be an actionable customer survey? Then, your input requires incisive, compelling questions. Want to know how to write good survey questions? Read on.
Prashanth Southekal, head of the Data for Business Performance Institute, says, “Companies...have tons and tons of data, but [success] isn’t about data collection, it’s about data management and insight.”
Well said, Prashanth! Let’s set your sights on a customer (or employee) survey that gleans actionable insights, not one that simply amasses more data. After all, you’re probably swimming in data already.
While survey software like Qualtrics helps you to implement your surveys, it does a poor job of writing survey questions for you. And survey questions lifted from another company’s survey won’t help you get to the heart of your customer’s experience. That leaves it up to you to learn how to write good survey questions or hire an expert to do it for you.
Eight tips for how to write good survey questions.
Tip 1: Ask your customers if they want to answer anonymously.
One of your opening survey questions should ask respondents if they want to answer anonymously. When given the option, about half of your respondents will usually choose to take your survey this way.
Of course, it's human nature to want to know who's giving you which ratings, but customers want to avoid being bothered by follow-up calls or possible retaliation. And everyone’s feedback is valuable, anonymous, or not.
Tip 2: Keep your questions focused.
Unify your questions around specific topics. Don't let your questions veer off in a dozen directions. It's tempting to ask about everything you can think of, but sticking to a small set of topics is far better. This improves usability for your customers and actionability for you!
Tip 3: Scrub leading constructs from your questions.
Bias can show up in numerous ways and skew your data. Never write questions that assume your respondents feel a particular way. For example, don’t ask, “How satisfied were you?” because this question assumes customers are at least somewhat satisfied.
Tip 4: Write questions that reflect how your customers talk.
Can you think of a friend who tells you they “neither agree nor disagree” or “somewhat agree” with you? Of course not; nobody talks like that. Instead, when possible, write your surveys as though you’re talking in person.
For instance, ratings scales that use “Bad,” “Poor,” “OK,” “Good,” and “Great” tend to collect accurate rating data because those are familiar, plain speak words.
Tip 5: Ask questions at the right time.
The principles of sound science apply to good surveys, so create an environment where customers can answer objectively without pressure. For example, issue your surveys after the interaction has concluded, not while the customer is still on the phone with your representative or in the car with your driver.
Customers feel social pressure to give high ratings if the person they’re rating is still present. Wait to send surveys until an interaction has concluded because this helps your customers provide truthful answers.
Tip 6: Avoid double-barreled questions.
Double-barreled questions combine two questions into one, so your customer is unsure which one to answer.
For example, "Was your server prompt and polite?" asks about two separate behaviors in one question and will result in flawed data that obscures the information you are collecting and the next steps you should be taking.
Tip 7: Don’t force customers to answer irrelevant questions.
If a question doesn’t apply to your customers, don’t demand they answer it. Asking a customer what their favorite coffee shop is when they may not be a coffee drinker requires them to say something untrue.
Instead, give your survey respondents an opt-out answer like Not Applicable” or “None of the Above,” or make sure it’s not a required question. Even better, use logic gating to ensure customers only get questions that are germane to them.
Tip 8: Ask ‘answerable’ questions.
There are many questions that customers simply can’t answer. For example, we’ve seen surveys ask respondents, “Rate the white space to text balancing on our website.” Questions like this are best answered by your design team, not your customers. They don’t belong in surveys.
Great survey examples.
It’s not a good practice to copy questions from another company (or even from this article.) Nevertheless, examples can give you ideas for how to write good survey questions.
Ranking questions.
These types of questions usually include two columns. The left column lists key factors or departments. Your customers select what matters most to them by dragging those selections to the right column in priority order.
Which 2 areas impact your business the most? Drag them to the right, in order of their importance.
Rating questions.
“Overall, how would you rate Company XYZ?” Or, “Overall, how would you rate Product XYZ? – or Service XYZ?” Simple rating questions are often part of a long-term strategy for tracking customer satisfaction over time.
Comparison questions.
“Who is Company XYZ’s closest competitor?” Then, follow this question with, “What are areas in which <COMPETITOR> performs better than Company XYZ?”
Since your customers are constantly comparing your company to your competitors, you should be, too.
Find weak points.
“What challenges (if any) did you encounter while using Company XYZ’s products/services?” If your customers are unhappy about something, you want to hear about it right away.
Identify room for improvement.
“What could XYZ do to make your experiences easier or better in some way?” Always be on the lookout for ways to improve.
Pro tip: How to write good survey questions.
Include a mix of question types like open-ended questions, rating scales, and multiple-choice questions to provide a comprehensive view of the customer experience. This also makes your survey more engaging. But don’t include so many different types of questions that your customers will feel confused!
How to measure customer satisfaction.
Every customer interaction is comprised of multiple elements that impact the customer’s overall opinion. But these elements aren’t equally important to the customer, so don’t treat them as equally important in your surveys.
For example, coffee shop customers typically care that they get their order quickly, not whether their name is spelled right.
Because it’s not all equally important, weighting your data is critical to measuring customer satisfaction accurately. There are two main methods for establishing a weighting factor:
Run a correlation study to determine what impacts your scores the most.
Use a question format, like drag-and-drop ranking, as shown in the prior section, to easily identify your customers’ priorities.
Now that you have a weighting factor, take your survey data and apply a multiplier to reflect what matters most to each customer and their situation.
Again, because every aspect of the customer experience is not equally important, this is the only way to get an accurate customer satisfaction score.
Financial calculations measure satisfaction, too.
Many financial metrics are proxies for satisfaction, but the three best ones are these:
Repeat Purchase Rate, or the frequency with which customers buy, buy more, and buy again, is an excellent window into your customers’ happiness.
Customer Lifetime Value measures how valuable a customer is to your company. A high CLV signals you have satisfied customers who engage consistently with your company over time.
Churn Rate is the rate at which customers stop buying from a company. A high churn rate could be due to price or marketplace forces but often signals low customer satisfaction.
Target Customer Satisfaction (CSAT) Scores.
It's pointless to think about how to write good survey questions if you don't have target scores in mind. In our years of experience, we know that you want nearly 100% of your customers to rate you in the top tiers, but for most companies, that's unrealistic.
Instead, consider that a superior CSAT score generally means that 65 percent of your customers rate you in the top tiers – and only a few rate you in the lowest tiers.
For example, with Net Promoter Score, the top satisfaction tier comprises 9s and 10s. You are doing well if 65% of your customers are in that top tier and only a small number rate you six or below.
If you use a 5-point scale like 1, 2, 3, 4, 5 OR “Bad,” “Poor,” “OK,” Good,” Great,” the top satisfaction tier is comprised of 5s OR “Greats.” If 65% of your customers rate you in that top rung with just a few detractors (1s and 2s OR “Bads” and “Poors,” you are offering a superior experience.
A company with exceptional customer satisfaction will have 80 percent or more of its customers in the top tier and, again, very few in the bottom tiers.
Get out what you put in.
There’s no substitute for bias-free, scientific survey questions. And remember, surveys don’t just measure satisfaction, they are part of the customer experience too! If you’re looking for advice on how to write good survey questions, these tips are just the beginning. There’s a lot to know about how to write great surveys!
Other related posts