Customer satisfaction: NPS, CSAT and CES explained.

In this guide I show how to measure satisfaction with the three most effective indicators (NPS, CSAT and CES) and how to turn results into improvement actions through a replicable process. The SMEs (small and medium-sized enterprises) that win on customer experience do not necessarily measure more, but they measure better and act faster.

The three fundamental indicators.

NPS (Net Promoter Score): loyalty and recommendation

NPS measures the probability that a customer recommends your company to a colleague or contact. It is obtained with a single question: on a 0-10 scale, how likely would you be to recommend our company? Those scoring 9-10 are promoters, 7-8 are passives, and 0-6 are detractors. NPS is calculated as the percentage of promoters minus the percentage of detractors.

NPS is the most widely used global indicator due to its simplicity and predictive power: it correlates directly with business growth.

CSAT (Customer Satisfaction Score): point-in-time satisfaction

CSAT measures satisfaction with a specific interaction or experience. It is obtained by asking how satisfied the customer is with a specific service, generally on a 1-5 or 1-10 scale. It is calculated as the percentage of positive responses (4-5 on a 5-scale, or 8-10 on a 10-scale) over total.

CSAT is ideal for measuring specific transactions: satisfaction with a delivery, a customer service interaction, a completed project or an after-sales service.

CES (Customer Effort Score): ease of experience

CES measures the effort a customer had to make to resolve an issue or complete a transaction. It is obtained by asking how much the customer agrees with the statement: the company made it easy for me to solve my problem. Lower perceived effort, greater loyalty.

CES is especially useful for evaluating technical support, complaints, returns and any interaction where ease is key to experience.

Comparison table: when to use each indicator.

Indicator	Measures	Sending moment	Frequency	Best for
NPS	Loyalty and overall recommendation	Periodically (quarterly/half-yearly)	Relational	Global business health
CSAT	Satisfaction with an interaction	Immediately after the interaction	Transactional	Service quality at specific touchpoints
CES	Customer effort	After support / complaint handling	Transactional	Support, after-sales, self-service processes

Best practice is to combine all three: NPS to measure the relationship, CSAT to measure key interactions, and CES for processes where ease matters.

Designing effective surveys.

An effective satisfaction survey meets five conditions:

1. It is short: 5-8 questions maximum, ideally 3-5.
2. It is sent at the right moment: immediately after the interaction for CSAT and CES, periodically for NPS.
3. It includes at least one open question inviting the customer to explain their rating. Open responses are where qualitative value lives.
4. It has a design that makes responding on any device easy, especially mobile.
5. It is not manipulative: avoid leading questions ("don't you agree that…?").

Real case: a 35-employee industrial company.

An industrial SME in the metal transformation sector in Aranda de Duero (35 employees, 140 active customers) deployed a three-tier measurement programme during one fiscal year:

• Half-yearly NPS sent by email to purchasing managers of active customers (38% response rate).
• Transactional CSAT sent by SMS 48 hours after each delivery.
• CES sent only to customers who had opened an incident.

In the first cycle they detected an overall NPS of 31, acceptable but improvable. Cross-analysis revealed that detractors concentrated on urgent orders under 72 hours. They implemented a specific urgent-orders process with three improvements (pre-assigned production slot, proactive status communication and automatic compensation on delay).

From data to action: the improvement loop.

Measuring without acting is worse than not measuring, because it frustrates customers who spend time responding and see no change. The full process includes five steps:

1. Analyse results identifying patterns (by customer segment, product, service, account manager).
2. Prioritise improvement areas with greatest impact on satisfaction.
3. Design corrective actions with owner and deadline.
4. Implement the actions.
5. Re-measure to verify improvement.

The most useful tool for prioritisation is the importance-satisfaction matrix: it crosses what matters most to the customer with their current satisfaction level, focusing resources on the quadrants of high importance and low satisfaction.

Integration with the quality management system.

Customer satisfaction measurement is a mandatory input for management review (ISO 9001, clause 9.3). Results must be analysed in management review, generate specific quality objectives, feed continuous improvement plans and feed back into risk and opportunity analysis.

Mini-glossary.

• NPS: Net Promoter Score, loyalty indicator based on a 0-10 recommendation question.
• CSAT: Customer Satisfaction Score, point-in-time satisfaction indicator.
• CES: Customer Effort Score, indicator of perceived customer effort.
• Detractor: customer scoring 0-6 on NPS, likely to speak negatively of your brand.
• Promoter: customer scoring 9-10 on NPS, likely to actively recommend you.
• Voice of Customer (VoC): structured programme to capture and analyse the customer's voice.
• Touchpoint: customer-company contact point in the customer journey.

Frequently asked questions.

Which is the best indicator if I can only measure one?

If you can only measure one, measure NPS. It is the simplest to implement (a single question), most comparable to sector benchmarks, and best predictor of future customer behaviour. Add CSAT and CES in a second phase, once you have a measurement habit and a process to respond to data.

How often should NPS be measured?

For B2B the usual frequency is half-yearly or quarterly. For low-frequency B2C (banking, insurance, utilities), once a year is sufficient. For high-frequency B2C (e-commerce, retail) rolling continuous measurement works, capping at one NPS survey every 90 days per contact.

What response rate is reasonable?

For B2B with personalised email and established relationship, 25-40% is good. For mass B2C by email, 5-15% is normal. By SMS it rises to 20-30% in B2C. If it falls below 5%, review three things: clean contact list, attractive email subject, and right send moment.

Is outsourcing measurement worth it?

For companies with fewer than 500 active customers, it is not necessary. Tools like Survicate, Delighted, Typeform or Google Forms enable a decent low-cost programme. For higher volumes or multichannel programmes with CRM integration, outsourcing the technical platform is worth it, but analysis and action must stay in-house.

How do I avoid response bias?

The main bias is self-selection: the very satisfied and the very angry respond more, leaving the silent majority out. To mitigate, segment by customer type (do not just analyse aggregate), measure at different lifecycle moments and, in key B2B, complement with personal interviews with selected customers.

What do I do with a specific detractor once identified?

Call within 48 hours (close the loop). Not to sell anything, but to understand what happened and what they need. A recovered detractor often becomes a promoter, because they see their complaint triggered real response. Document the learning and link it to the non-conformity process.

Checklist: 10 steps to implement a satisfaction measurement programme.

1. Define the three questions you want to answer (e.g. are we good at deliveries? at support? would they recommend us?).
2. Choose the indicators (NPS for relationship, CSAT for deliveries, CES for support).
3. Select a simple tool (Typeform, Survicate, Google Forms to start).
4. Design surveys with a maximum of 5 questions and one open question each.
5. Define sending moments (transactional vs relational).
6. Launch a pilot on a small segment.
7. Adjust response rate and wording based on pilot results.
8. Scale to the full portfolio with a defined cadence.
9. Establish the closing loop: who calls each detractor in under 48 hours.
10. Bring results to management review and the scorecard.

Do you want to implement a customer satisfaction measurement system integrated with your quality system? Let's talk and I'll design with you the measurement programme appropriate to your company, its size and its sector.

---

By Ángel Ortega Castro · independent consultant in strategy, quality and digitalisation for SMEs. Based in Aranda de Duero (Burgos), Castilla y León.