Customer Satisfaction Survey Rulebook

Written by Kayla Brehm | Feb 25, 2015 9:59:25 PM

It all comes down to customer experience. From the shipping speed for the shirt you just bought online to the interaction with your waiter at that awesome new restaurant downtown- we want the best experience possible. Despite the knowledge that the best case scenario (your shirt arrived two days early!) can’t always happen, it’s always frustrating when it doesn’t.

Many companies today are going to great lengths to track both the good and the bad of the customer experience. Customer Satisfaction Surveys, alongside tracking metrics like NPS and CSAT, are a great way to monitor company and employee performance. When figuring out (for the buzzword lovers) best practices, I’ve found it incredibly useful to revisit it from an elementary perspective. Using the scientific method as a review model, we’ll break down the best and worst ways to implement customer satisfaction surveys.

Scientific Method

1) Ask a Question
2) Do Background Research
3) Construct a Hypothesis
4) Test Your Hypothesis by Doing an Experiment
5) Analyze Your Data and Draw a Conclusion
6) Communicate Your Results

1. Ask a Question

Do you want to improve? Let’s be honest - who doesn’t? Customer satisfaction surveys are a great way for companies to evaluate their progress and better understand their customers on all levels. By asking “what can I do better”, you open yourself up to improvement. Examine all areas that you want to work on and ask your customers for feedback.

When to question

It’s important for companies to gather information both periodically and in real time. A great example of real time is surveying a customer after a contact with the front line. This information will give you an indication of how well issues are being resolved and what you could need to work on from an individual and corporate level. Periodic surveying can be done to examine a customer’s opinion of a company or product as a whole. Below we’ll examine the best approach for both.


2. Do Your Research

No one should know your customer better than you. Where do they shop? Do they prefer coffee over tea? Do they use Apple or Microsoft products? Use this information to design a survey that appeals to them.

What works?


Consider the length of your survey. Your customers are doing you a favor by participating- don’t make it a challenge for them by asking more questions than necessary. It should only take five minutes or less to complete.

Response Rates

To first understand response requirements, you’ll need to better understand your customer population. To get an accurate reading, you have to have a random sample. Survey Gizmo breaks it down in the survey below by population size, or your customer base. Representing a confidence level of 95% and a confidence interval (margin of error) of 5%. If your customer base is small, you should consider surveying a large number to get a more accurate reading.


It’s difficult to establish a precedent for how frequently a company should survey their customers. It will vary from company to company, depending on size and survey function. In order to avoid survey burnout, it may make sense to limit 1-2 surveys per customer per year.


Work with your team to evaluate the use of incentives. Companies like Zappos and Warby Parker have had great success in furthering brand loyalty by establishing an incentive process for those participating in customer feedback surveys.

What doesn’t work?

Restrictive multiple choice

Not all problems are universal. Your survey shouldn’t be tailored to only common issues that pop up in your company. If multiple choice is your preferred method, include an additional text box at the end of the survey. This is the opportunity for the customer to tell you what you could be doing better and provide any outlying information that may not have been covered in the survey.

Two part answers

An extension of restrictive multiple choice, often “yes” and “no” aren’t adequate enough to describe a customer’s experience. If you’re set on including these two responses, include an additional text box for “other” and allow the customer to let you know why the two responses didn’t meet their experience.

Mandatory responses

A completed customer survey is a gift from the customer- not a requirement.


3. Estimate Your Results

It’s always a good idea to have a benchmark of where you want your company to be. Net Promoter Score is a great tool because it allows you to compare your status with others in your industry to better determine your strengths and weaknesses. What do you think needs improving? What do you think your company is doing right? How have you improved from your last survey? Your findings may surprise you. is a great resource for those interested in scoping out how others are rated. The image below, for example, showcases recent NPS Scores for Telecommunications companies.


4. Survey

Ask the right questions!

Ticket resolution

Ticket resolution and quality of service provided is an area that requires constant improvement. In a recent survey, when asked what creates the best customer experience situation, 86% responded that they're happiest when their problems are resolved.  It sounds simple, but ticket resolution is complex. If it’s your company’s ultimate goal to increase the quality and speed of each call, ask questions like below.

- Did the agent resolve your issue today?
- How many times did you have to contact the company to resolve your issue?
- Would you say that our customer service representative solved your problem or answered your question quickly, slowly, or neither?
- How knowledgeable was the customer service representative?
- During your last interaction with us, you contacted a member of our Customer Loyalty Team. On a scale of 0-10, if you had your own company that was focused upon service, how likely would you be to hire this person to work for you? (a question pulled from the Zappos customer satisfaction survey).

Product Review

If your goal is to evaluate product performance and standards, the following are great questions to include on your survey.

- How likely are you to recommend our product to a friend or coworker? (NPS Benchmark)
- How well do our products meet your needs?
- How likely are you to purchase any of our products again?
- What would make you more satisfied about this product?

Company Gut-Check

If your goal is to analyze your customer’s opinion of your company’s overall performance, you can get creative with your survey. A lot of companies have started to ask questions like “If our company were a car, what type of car would we be?” While, out of the box, the question allows a company to gain a unique insight into how a customer perceives your brand. Other common survey questions could include:

- How likely are you to recommend us to a friend or coworker? (NPS Benchmark)
- If you had to name something that we could improve upon, what would that be?


5. Analyze Your Data

The survey is sent out, but only half of the battle is done. Establish value in surveying by analyzing your results. Look for trends- they could pinpoint areas that you company may need to improve.

What to look for

If you’re interested in examining responses to a particular question, start by calculating the mean (total average of the responses). From there, establish a z-score (normal score) by subtracting the reasonable benchmark from from the mean (Nielsen & Levy found appropriate benchmarks to be 80% of the points in a scal-so, 4 if there are 5 points in the scale).

Assuming the mean was 4.5, the end result of this would be .5. For the sake of the example, say there are 10 numbers (4,4,4,4,4,5,5,5,5,5) with a standard deviation of 0.52705. At this point you divide the difference (.5)  by the standard deviation and get .9486766 (the z-score). Using the z-score we can find that 34% of responses will fall outside of a standard deviation of the mean. In this way, we’re able to see that there’s a large range of “normal” that can be used to help companies understand where to look (the outliers).

The z-score is great because it accounts for variability (as Jeff Sauro explains in great detail here) but z-scores are just one of many options to interpret data. The point is that you want to see what’s average and where the outliers are.


6. Communicate Results

Follow up!

For NPS folks, follow up with your passive (0-6) and detractor (7-8) customers. In an article with HubSpot, Fred Reichheld, Bain fellow and founder of the Net Promoter System, stated that, for B2C’s, it’s becoming increasingly more popular for companies to close the loop when a detractor provides feedback.

They tend to share their opinions much more widely via social media tools -- and they are relying on recommendations of peers as the most reliable source for choosing new suppliers.”

Going the extra step proves to the customer that you have their best interests in mind. If the customer survey stated that they were unhappy because their item was damaged or their question was unanswered, reach out to see if there’s anything more you can do. That extra step could be all it takes to turn a passive into a promoter.

Immediate Changes

Your survey results could indicate a process that requires immediate change. Work with your team to establish a process for quick communication. Whether implementing a JIRA ticket for your engineering team to notify them of the common trend, or a simple chat to let them know something may be amiss, find a process that works for you.