Customer Service Guide

Published: 2026-10-30 · Reviewed: 2026-11-04

How to Run a Customer Satisfaction Survey: A Practical Guide

Feedback forms capture one transaction; a satisfaction survey measures the whole relationship. It tells you whether customers are happy with your business overall, where you stand against their expectations, and whether they would come back or recommend you. Done properly, it gives you a small set of numbers you can track quarter after quarter, and a list of specific fixes.

When to Use This

Use the survey once or twice a year with your active customer base, or quarterly if you are in a fast-moving service business. It is also the right tool after a major change, a new system, a price change, a new branch, when you need to know how the change landed. Send it to customers who have bought from you in the last six to twelve months, not to your entire mailing list going back years.

Do not use the survey as a complaint channel, that is what the complaint log is for, and do not send it immediately after a service failure, when answers will reflect one bad day rather than the overall relationship. Keep it separate from marketing emails too; a survey that opens with a sales pitch will not be taken seriously.

How to Fill It Out

Section 1: Survey header and purpose statement. Before sending, write a two-line introduction: who you are, why you are asking, and how long it takes. “We’re Brightline Office Supplies. We’d like 3 minutes of your time to understand how we’re doing, so we can serve you better.” A clear purpose and a time estimate raise completion rates dramatically, people open a “3-minute survey” far more readily than a mystery link.

Section 2: Overall satisfaction and loyalty questions. The core scored questions, all on the same 1-to-5 scale: overall satisfaction with your business, likelihood to buy again, likelihood to recommend to a friend, and value for money. These four give you your headline numbers. Keep the scale identical across all questions and label the endpoints clearly (“1 = very dissatisfied, 5 = very satisfied”) so a 3 means the same thing everywhere.

Section 3: Service-area ratings. Break the experience into the areas you can actually manage: product quality, delivery speed, customer service responsiveness, ease of ordering, and after-sales support. For example, a respondent might score product quality 5 but delivery 2, and now you know exactly where to invest. Do not include areas you cannot change; asking about something you will not fix just breeds cynicism.

Section 4: Open-ended questions. Two questions, no more: “What do you value most about doing business with us?” and “If you could change one thing, what would it be?” Limit the space so answers stay focused, but read every single one. The scored questions tell you the size of the problem; these answers tell you what the problem is. One respondent writing “your phone line closes at 4pm and I’m at work until 5” is worth more than fifty 3-out-of-5s.

Section 5: Customer segment details. A few classification questions so you can slice the results: how long they have been a customer, roughly how often they buy, and which branch or channel they use most. This is how you discover that new customers love you but customers of three years are drifting, a pattern invisible in the overall average. Keep it to three questions; more feels like data mining.

Section 6: Follow-up permission and incentive note. Ask whether you may contact them about their answers. If you offer an incentive, a discount code or a prize draw, state the terms plainly: what it is, when it is drawn, and that completing the survey is all that is required. Never make the incentive conditional on a positive answer; that corrupts your data and your credibility.

Common Mistakes

1. Asking questions you have no intention of acting on. Every unanswered survey finding is a small betrayal. If you ask about delivery speed and then ignore three quarters of bad scores, customers learn that your surveys are theatre. Before sending, get the manager’s commitment: the top three findings will get owners and deadlines.

2. Surveying only your happiest customers. Hand-picking recipients, the regulars, the ones who complimented you last week, produces a lovely score and zero information. Send to a fair cross-section, including customers who have gone quiet. The quiet ones are the ones the survey is for.

3. Changing the questions every time. If you rewrite the survey each year, you cannot compare this year to last, and the trend, the most valuable thing a survey gives you, disappears. Keep the core questions identical year to year; add or change only the service-area section if the business has genuinely changed.

4. Sitting on the results for months. A survey analysed in June and discussed in November is history, not insight. Aim to have results compiled within two weeks of closing, reviewed by management within a month, and the first actions communicated to customers within six weeks. Momentum matters, a fast, visible response teaches customers that answering was worth their time.

Who Needs to Sign Off

The customer service manager (or business owner) signs off the survey questions before it goes out, the questions commit the business to listening, so they need senior eyes. The compiled results and the proposed action plan go to the same person for sign-off before actions are assigned. If the survey covers regulated areas or collects personal data beyond contact details, check your local consumer protection and data privacy rules first, and have the owner confirm the survey complies.

How Often to Review and Update

Run the survey on a fixed rhythm, annually at minimum, twice a year if you can act on the results that fast. Analyse and act within weeks, not months, of each round. Review the questionnaire itself once a year: keep the core questions stable for trend comparison, but refresh the service-area questions to match what the business actually offers now. Between surveys, track the headline numbers from your regular feedback forms so there are no surprises when the big survey lands. A survey is a health check, useful on schedule, useless as a one-off panic measure.


Download the free Customer Satisfaction Survey.