Many SMEs know from conversations that their customers are satisfied, and are still surprised when a long-standing customer leaves. If you want to measure customer satisfaction, you do not need an expensive study: three well-chosen questions, easy access for customers and a clean evaluation are enough to begin with. This guide shows how to build a short survey with a satisfaction scale, a recommendation question (NPS) and free text, what to consider under data protection and how to read the result correctly.
Why and what to measure
Satisfaction is an early indicator. Complaints show what went wrong, and they come from those who speak up. The silent majority says nothing but may drift away. A survey collects this feedback systematically.
Decide in advance what you want to measure: satisfaction with a specific order (after completion), with the cooperation overall (once a year), or willingness to recommend you. Do not mix the feedback on an order and the annual assessment of the cooperation in one questionnaire; the recommendation question suits both. Feedback on the order steers the processes, the annual survey steers the relationship.
What ISO 9001 says
If your business works to ISO 9001 or wants to be certified, clause 9.1.2 (“Customer satisfaction”) is the point of reference. In substance, it requires you to monitor customer satisfaction. As we understand it, the clause does not prescribe a particular method such as the recommendation question; you decide how to obtain the information.
On the edition: The 6th edition, ISO 9001:2026, has applied since 16 September 2026 and replaces the 2015 edition. The clause number and the description of content above come from the 2015 edition and have not been checked against the text of the 2026 edition. According to a Swiss certification body, the changes overall are moderate, without a fundamental new direction. Check number and wording in your edition; certified companies switch within the transition period set by their certification body. This guide does not reproduce the text of the standard.
Three question types and how the NPS is calculated
Satisfaction scale. A question such as “How satisfied are you overall with our service?” with a scale from 1 (very dissatisfied) to 5 (very satisfied). You evaluate the average and the distribution of the answers. Together they say more than the average alone: an average of 4.0 can arise from nothing but fours, or from fives and ones.
Recommendation (NPS). The Net Promoter Score is a widespread method, not a statutory or normative requirement. You ask: “How likely are you to recommend us?” The answer is given on a scale from 0 to 10. According to the description of the method by Bain & Company, the answers are divided into three groups:
- Promoters: 9 or 10
- Passives: 7 or 8
- Detractors: 0 to 6
The NPS is the share of promoters minus the share of detractors, each as a percentage of all answers. Passives make no contribution of their own to the calculation but count towards the base. The result lies between −100 (only detractors) and +100 (only promoters) and is usually stated as a number without a per cent sign. An NPS of +35 does not mean “35 per cent satisfied” but: the share of promoters is 35 percentage points above the share of detractors.
Free text. An open question such as “What can we do better?” delivers what numbers do not say. It explains the numbers: why do customers give a 6? Read the free texts together with the figures and look out for repetitions.
Designing the survey
Few questions, high participation. Three to five questions, answered in 1 to 2 minutes, achieve more than a questionnaire with twenty items. The trio of scale, recommendation and free text is a good start. Make satisfaction and recommendation mandatory questions and the free text optional.
Word the questions neutrally. “How satisfied were you with our excellent service?” steers the answer. Ask about one topic per question and use the same wording each time you repeat the survey, otherwise the results are not comparable.
Timing is decisive. Ask shortly after an order is completed, while the memory is fresh. Distribute access where customers look anyway: in the completion email, on the delivery note or the invoice, as a QR code in the shop.
Note how many people you invited. Only then do you know the response rate: responses divided by invitations. Very small response rates distort the picture, because mainly the very satisfied and the very annoyed respond.
Data protection in customer surveys
The Federal Act on Data Protection (DSG) applies to personal data of natural persons (Art. 2, para. 1 DSG). Personal data is all information relating to an identified or identifiable natural person (Art. 5 lit. a DSG). Information about a company as such is not personal data; for sole proprietorships, however, the company name can make a person identifiable. The answers of a contact person who can be identified by name or via a personal link are personal data.
A few rules follow from this: obtain personal data only for a specific, recognisable purpose and process it only in a way that is compatible with it; processing must be proportionate; as soon as the data is no longer required, it is destroyed or anonymised (Art. 6, paras. 2 to 4). When obtaining the data, you inform at least about your identity and contact details, about the purpose and, where applicable, about the recipients of the data (Art. 19, para. 2). Whether an anonymous survey falls under this depends on whether persons can be identified. Be careful with free text: customers occasionally name your employees in it. Read the answers before passing them on and do not circulate them uncontrolled.
In practice this means: ask only what you need for the stated purpose. Explain in the introduction who is conducting the survey and what the answers are used for. Make details about the person optional.
Reading and using the results
After each survey, evaluate three things: the key figure (average, NPS), the distribution and the free texts. Pay attention to the base. With 40 answers each single answer accounts for 2.5 per cent; if one promoter becomes a detractor, the NPS changes by 5 points (2 ÷ 40). Small fluctuations are therefore not a trend.
Compare only like with like: the same questions, comparable periods, comparable customer groups. And ask: what do we change in concrete terms? Recurring criticism in the free texts becomes actions; from positive feedback you see what you should keep. Make the value a metric of a quality objective with a starting value, target value and date, so that management sees the development and discusses it in the annual review (management review).
Practical example: Muster Treppenbau AG
Muster Treppenbau AG is a fictitious company. It surveys customers after installation is completed with a scale question, the recommendation question and a free-text question. 160 customers were invited, 40 responded; the response rate is 40 ÷ 160 = 25 per cent.
The recommendation question gives:
| Group | Answers | Share |
|---|---|---|
| Promoters (9 or 10) | 22 | 55 % |
| Passives (7 or 8) | 10 | 25 % |
| Detractors (0 to 6) | 8 | 20 % |
| Total | 40 | 100 % |
NPS = 55 % − 20 % = +35. Calculated by count: (22 − 8) ÷ 40 = 0.35.
The scale question gives: 1 point once, 2 points twice, 3 points five times, 4 points 16 times and 5 points 16 times. The sum is 1 + 4 + 15 + 64 + 80 = 164, divided by 40 answers gives an average of 4.1.
In the free texts, five respondents mention postponed installation dates. The company sets itself a quality objective: the NPS is to rise from 35 to 40 by the end of 2027. A second survey with the same questions follows after six months. The postponed dates are then evaluated first; if the number of detractors falls, it shows in the metric.
Checklist
- Aim of the survey defined (order or cooperation overall)
- No more than three to five questions, worded neutrally
- Scales clearly labelled (1 to 5 or 0 to 10)
- Free text optional
- Introduction names the controller and the purpose
- Access via link or QR code placed where customers see it
- Number of invitations noted (for the response rate)
- Answers evaluated: key figure, distribution, free texts
- Metric recorded as a quality objective with starting value, target value and date
- Actions derived from recurring criticism
- Repetition with the same questions planned
How KMU | base helps
In QMS | base you find the surveys in the menu under Operations and records › More › Customer surveys. The page is visible to people with the right to manage surveys.
Surveys with scale, NPS and text. You create customer surveys with an introduction, suggested questions, mandatory questions and an end date. Three question types are available: scale 1 to 5, recommendation 0 to 10 (NPS) and free text, with suggestions for the questions. You decide the order yourself. A survey starts as a draft and accepts answers only once you activate it; with “Close” you close it again.
Public link with QR code. Customers answer without signing in via a link or a QR code; spam protection and a limit per IP address are built in. The answers are stored without names, and giving the company is optional; this is also stated on the page for respondents.
Evaluation as a target metric. You see the evaluation per question: number, average, NPS, distribution and free texts, plus the list of companies named voluntarily. With “Adopt as metric” you set the result as the actual value in the quality objective that you linked to the survey. The NPS is adopted if the survey contains an NPS question, otherwise the average of the scale questions in per cent (1 corresponds to 0 per cent, 5 corresponds to 100 per cent). In the example that would be 35 with an NPS question, and without one (4.1 − 1) ÷ 4 × 100 = 77.5 per cent; the software shows the value as a whole number.
Measurable quality objectives and actual values. An objective has a metric, unit, direction, starting value and target value, tolerance, due date, process and responsible person; traffic light and progress result from target and tolerance.
What the software does not do: The link is a shared access, not an individual mailing. QMS | base sends no invitations and does not assign answers to individual customers; you handle the mailing and the number of people invited yourself, so the software does not calculate a response rate either. Once the first answer is in, the questions can only be reworded, no longer added, removed or reordered. The evaluation covers all answers of a survey; for the second round you therefore create a new survey so that the values are not mixed. Comparison by customer group or period is not provided. The free texts appear collected per question; they cannot be filtered by promoters, passives or detractors.
The guide to QMS | base explains how to use it. Take a look at the demo or open a trial account. If you have questions, the contact page will help.
Sources
- SNV: SN EN ISO 9001:2026, Quality management systems – Requirements, valid since 16 September 2026, replaces the 2015 edition; accessed on 27.10.2026. The clause reference 9.1.2 comes from the 2015 edition and has not been checked against the 2026 text.
- ISO: ISO 9001:2026 (Edition 6), published on 16 September 2026; accessed on 27.10.2026.
- SQS: ISO 9001:2026 revision of the standard – all information, accessed on 27.10.2026.
- Bain & Company: Measuring Your Net Promoter Score, description of the method (scale 0 to 10; promoters 9 to 10, passives 7 to 8, detractors 0 to 6; NPS as the difference of the shares), accessed on 27.10.2026. Net Promoter®, NPS® and Net Promoter Score are trademarks of Bain & Company, Inc., NICE Systems, Inc. and Fred Reichheld. The NPS is a practical method, not a statutory or normative requirement.
- Federal Act on Data Protection (DSG, SR 235.1), Art. 2, para. 1, Art. 5 lit. a, Art. 6, paras. 2 to 4 and Art. 19, para. 2, as at 7 July 2025: fedlex.admin.ch (German original), accessed on 27.10.2026.
The questionnaire structure, example figures and recommendations are editorial practical aids. They do not reproduce ISO standard text and do not demonstrate conformity for certification.
This guide does not replace legal or tax advice. Information as at 27 October 2026.