Opinions expressed by Entrepreneur contributors are their own.
As a customer service consultant and customer service transformation expert, I’ve felt like cringing a few times (or more than a few!) when my client companies send out customer satisfaction (CSAT) surveys that are less accurate and customer-friendly than I’d want them to be.
This is risky for any business, so I work with them immediately to make the needed improvements.
Sending out poorly designed surveys can try the nerves of even your most loyal customers, mislead you with spurious results and waste the time of everyone involved — those who send it out and those who receive it. When designed and deployed correctly, however, surveys can reveal essential insights into how customers view their experience with your company and allow customers to vent!I encourage you to spend a few minutes with me learning the do’s and don’ts of designing and deploying customer surveys.
1. Every survey question should be clearly worded and easy to answer
It shouldn’t require your customer to do math or think too much about the inner workings of your company. Avoid anything like, “Compare this interaction with interactions you’ve had at similar departments at other fintech companies in our broadly competitive cohort.” Also, don’t ask questions you don’t care about and already know you’re not going to act on. (This seems obvious, but it happens all the time.)
2. Don’t ask your customers to grade you on a scale of 1–10
When you request their opinions on a scale of 1-10 (or 0-10), you’re confounding your customers at best. Why? You’re essentially asking your customers to determine the difference between a “six” and an “eight” or a similar nuanced gradation when choosing how to rank you. Provide your customer with no more than approximately five choices. (Why do I say approximately five? Well, I’d go with five, but I confess that there is an argument to be made for making it four or six: If you choose an even number like this, you take away an easy midpoint response, thus perhaps getting more reasoned answers.)
3. The order in which you ask your questions matters
The order matters because a prior question brings up images in a customer’s mind that will influence their answer to the next one. So, be sure to ask for your customer’s overall impression first. You don’t want to influence how a customer answers this central question by asking more nitpicky questions before you get to the most important, broad one.
Asking several individual questions before asking for an overall rating will tend to color that overall rating, perhaps quite significantly. For example, if the question the customer encounters just before the big one asks about the cleanliness of your restrooms, which was just so-so, this will likely reduce your overall rating since you’ve left their mind in the toilets.
Conversely, if they’ve just been asked about the availability of parking and parking was abundant, this is likely to artificially increase your overall rating since they are thinking about something positive (how easy it was to park).
4. Include at least one open-ended question
Doing this is valuable both to harvest customer insights and to let customers know you value and are curious about their thoughts and insights. For example, “Please share any thoughts you may have; we promise to read all of these!”
The CEO of a major corporation told me that he transformed his entire level of customer service success by reading every one of these so-called “verbatims.” In these, he found a “staggering” level of nuance about his current operations and even some promising suggestions for innovations for the future.
5. Word choice matters
I’m a fan of emotive rating options on surveys, such as “fantastic!” (for your top score), “meh” (for somewhere in the middle) and even “Are you sure you can handle the truth?!” (for your lowest). Only consider this approach if it conforms with your brand style! It wouldn’t be appropriate for a traditional jeweler or a business in a life-and-death industry like healthcare or mortuary services.
6. Pay attention to the number of top ratings (5 on a scale of 5) that you receive
While it’s nice to know how, on average, customers perceive you, It’s arguably more important to know the number of customers who give you a top (5 on a scale of 5) rating. This may be more important than your average score because the number of people who rate you as tops are the best representation of the number of truly loyal customers you have — or, at least, the number of customers well on their way to true loyalty. Of course, most important here is the trend: are you getting more loyalists than in the past, or is customer enthusiasm flagging?
7. Don’t ask nosy questions
Nosy questions include questions on income, sex or how old they are. First, you can never assume respondents will trust your privacy practices. Second, unless you’re a casino operator, cannabis dispenser or operate another type of business limited by law to serving adults, you don’t have a reason to ask for a complete birthdate. If you are trying to set yourself up to send out birthday cards or offers later, please at least stop asking for the year of birth. A complete birthdate is probably none of your business and makes identity theft all too easy in the event of a breach.
8. Skim through your surveys right away, looking for any complaints or ultra-low scores
Then respond personally and immediately to these upset customers. Don’t make them wait without a response, stewing in their own frustration, until such time as you’ve batched all your surveys for review.
9. Put thought and attention into any preamble that accompanies your survey
Your introductory note or cover letter, just like the survey itself, should be friendly, gracious and brand-appropriate. This way, whether or not the recipient chooses to respond, they’ll be left with a positive impression.
10. Please don’t hound your customers if they don’t respond to your survey request(s)
I would make one follow-up reminder the limit — or even zero. Once you’ve surveyed a particular customer, suppress future surveys of that same customer for at least 30 days.