Identifying topics to be explored is the tip of the iceberg when writing a survey. The most difficult step comes in the phrasing of questions.
Online survey producer Survey Monkey says good survey questions are simple and unbiased.
No customer should be required to fill out a survey, which means every second they spend doing so is time they’re not spending on their business. It also means they are unlikely to complete with a survey that asks questions they can’t easily understand or answer.
Combining multiple queries into one question is one common mistake that can doom a customer survey, because it makes it “impossible for you to interpret [customers’] answers,” Survey Monkey says.
A company that wants to gauge its sales team’s response to customers via phone, email and text should ask different questions specific to each communication tool. This ensures a responder doesn’t merge its thoughts on the three tools to one response, and gives the surveyor insights on each method if overall customer opinion is different.
Simplicity also means avoiding excessive caveats or stipulations to questions every responder is expected to answer. Asking a customer a yes or no question and following with a direct secondary question can be valuable; but should not be overused. A good survey also should be written to allow customers to clearly pass over a secondary question if it does not apply to them.
And when a survey is updated to ask better questions, positive results are seen immediately, says Liz Macpherson, customer service manager at Haldex.
“We recently tweaked some questions on our annual customer survey and we almost tripled our response rates,” she says. “We continued to ask about the same issues, we just changed the questions to make them more concise.”
Bias can be another killer of a customer survey.
Injecting a company’s opinion of itself into a question can pressure a responder into agreeing with a statement they may not wholly support, Survey Monkey says.
For example, a supplier that updates its website should not precede survey questions about the update with phrases such as “How has our improved website” or “Since improving our website.” The “improve” statement alters the questions in such a way where a customer isn’t giving an opinion so much as confirming the company’s positive assessment of the website.
A website can improve and still be sub-par, but a biased question minimizes the likelihood of a customer responding as such.
Good survey questions also should be open-ended to attract responses.
The yes or no question previously referenced works because it feeds responders to a more focused and specific secondary question. It directs customers the question most relevant to them.
When the goal is not a simple response, a one-word answer is nearly worthless.
An example of this can be seen as follows. The question, “Are you satisfied with our shipping performance?” is formed in such a way that a survey responder may believe only a yes or no is required, or choose to respond with either a yes or no because no more information is explicitly requested.
Conversely, the question “How does our shipping performance impact your business?” invites a customer to respond with a complete sentence, or more, focusing on their level of satisfaction with the shipper’s performance and how it benefits them.
One extra word and a slightly altered statement produces considerably more information.