Customer feedback survey for a small business: 3 questions

Shorter than you think. The U.S. Chamber's small business site says to keep it "as brief as possible" and asks whether you would be excited to complete a survey with 20 to 30 questions (U.S. Chamber).

GatherUp, a review-management vendor, asked 500 consumers through Google Surveys how many questions they would answer. It reports that 75.4% said five or fewer (GatherUp). The company limits its own product to four questions, so it has a reason to like short surveys. Still, the direction matches the Chamber's advice.

Our post on how long a survey should be goes deeper on drop-off.

Start with what you will do with the answers. The Chamber's advice is to get clear on goals first: why you are surveying, and how you will apply what comes back (U.S. Chamber).

Here is a three-question version for someone who just bought something or used your service:

1. How satisfied were you with your visit today? A five-point scale, from very satisfied to very dissatisfied. PaySimple, a payments vendor, lists this as its first question (PaySimple).
2. What could we do better next time? An open box. This is the question that tells you something you didn't already think of.
3. Please share anything else you'd like us to know. Optional, and it catches what the first two missed.

These come from PaySimple's list of 13, with "visit" swapped in for "product/service" in the first. Its own advice is to be selective, because "jamming a lot of questions into a single survey causes survey fatigue" (PaySimple). It also says to make most questions optional, since mandatory ones can make people abandon the survey.

If you want a recommend question, add "How likely are you to recommend us to others?" on a 1 to 10 scale. That is one more question, so only add it if you will use the number.

For more on wording the open box, see how to write open-ended survey questions.

Soon after the visit. The Chamber's guide quotes HubSpot recommending customer satisfaction surveys go out as soon as possible after an interaction, while the experience is fresh (U.S. Chamber). Send it by email or text, or put a link on the receipt.

The same page says to send on a weekday during a low-peak time such as early morning, lunchtime or after work. It also warns that offering an incentive "can backfire", so a thank-you note is safer than a coupon.

Expect a minority to answer. Our post on survey response rates has sourced benchmarks.

Read all of them once. Then write a short list of groups from what people actually wrote, such as "slow service", "price", "staff were great" or "couldn't find what I wanted". Label each answer with one group and count them. Reread the answers under the biggest count, because the label is a shortcut and the sentences are the evidence.

A worked example

The numbers below are invented to show the method. They are not from a real survey.

A small cafe gets 40 answers to "What could we do better next time?". Sorted by hand:

The shares are worked out as answers divided by 40, rounded. The top group points at lunch, not the whole day. Reading those 14 answers might show that most mention the same 12:30 queue. With 40 answers, treat the counts as a direction. A gap of one or two answers is noise, and people who reply are not everyone who came in.

If you want to compare score and text, our post on reading NPS comments next to the score shows how.

There are a few honest ways to do this.

• By hand. Ask for replies by email, keep them in a spreadsheet and label each one yourself. It costs nothing and you read every word. It gets slow at a few hundred answers, and labelling by eye drifts.
• A general AI chat such as ChatGPT. Paste the answers in and ask for groups. It's quick, but a summary can merge two problems into one, so reread the answers behind each group.
• A survey tool. Google Forms, Typeform and Tally collect the answers. The Google Forms comparison shows where Google Forms and Enform differ on reading the written ones.

Enform writes the survey from a description of what you want to ask, you share one link, and it reads every reply and counts what people said. Every count opens to the replies behind it, so you can check a group against the sentences. It's a better fit once answers pile up. A spreadsheet is enough for your first twenty. See pricing, the roundup of customer feedback tools, or the product feedback template.