Measuring customer satisfaction in a service company: CSAT, NPS and CES, with a short survey to copy

How do you measure customer satisfaction?

You measure customer satisfaction with a short question at a fixed moment, counted the same way every time. The three best-known measures are CSAT (how satisfied were you?), NPS (would you recommend us?) and CES (how easy was it?). Put the answers next to figures from your own operations, such as arriving on time and fixing the problem on the first visit.

Why measure customer satisfaction?

• A customer who does not complain is not necessarily satisfied. They can just as easily ask someone else for a quote.
• The wrong person answers. The resident who let the technician in is satisfied. The facility manager who decides on the contract is annoyed by reports that arrive late.
• Without a fixed question, all you have is an impression. You cannot compare an impression from one quarter to the next.

Measuring customer satisfaction: which KPI?

• CSAT fits after a single intervention: the question is about that visit.
• NPS fits the annual question to contract customers, about the whole relationship. Those who give 9 or 10 are promoters, those who give 0 to 6 are detractors; 7 and 8 only count in the total. The score lies between −100 and +100.
• CES fits after a breakdown or a request to the office: how much effort did it take to get help?

A worked example, not a standard. 40 contract customers answer the NPS question: 22 give 9 or 10, 12 give 7 or 8, and 6 give 0 to 6. Promoters: 22 / 40 = 55%. Detractors: 6 / 40 = 15%. NPS = 55 − 15 = 40.

Another company’s score was asked and counted differently: compare with your own previous measurements. With few answers, the score swings a lot, and the open question tells you more.

When do you ask?

Do not ask the same person too often, for example at most once a month. After a breakdown, only ask once the problem is solved. And after a complaint, you do not send a survey: you call.

A short survey to copy

After an intervention, four questions are enough. Adapt the words in square brackets.

1. How satisfied are you with the intervention on [date] at [address]? (1 = not at all satisfied, 5 = very satisfied)
2. Did the technician arrive within the agreed time window? (yes, no, don’t know)
3. Was the problem solved after the visit? (yes, partly, no)
4. What can we do better next time? (free answer)

For the annual question to contract customers:

1. How likely is it that you would recommend us to a colleague? (0 to 10)
2. What is the main reason for your score?
3. What should change next year?
4. May we call you about your answer? (yes, no)

In the accompanying email, say who reads the answers and that this person will call if something was not right. You can also get the answers to questions 2 and 3 from your own records: if the customer’s answer differs, that is already useful information.

Customer satisfaction in B2B: several contacts per customer

• Map the roles. Who reports breakdowns, who lets the technician in, who pays, who decides on the contract?
• Ask each role about what that role experiences. The caretaker about the visit, the accounts department about the invoices, the decision-maker about the whole.
• Note who gave each answer, and talk to the decision-maker once a year. A half-hour conversation tells you more than a score alone.
• Look site by site for a customer with many buildings. An average hides the building where things go wrong every time.

Four figures from your operations that warn you in advance

A survey comes afterwards. You get these figures every week from your own records, and they are about what a customer notices straight away during an intervention.

Put them next to the answers to your survey. How to count first-time fix honestly is explained in calculating the first-time fix rate, and which response time to promise in SLA and response time. The other figures for your department are in KPIs for a technical department.

Closing the loop with an unhappy customer

1. Call quickly, for example within two working days. Preferably someone who is allowed to make decisions.
2. Listen first. Ask what happened and what the customer expects.
3. If something went wrong, treat it as a complaint, with one person responsible and a deadline. See handling complaints.
4. Say what you are changing, even if it is small, and call back after the next intervention.
5. Record the cause in a fixed list, and go through that list every month with the team, technicians included.

An example: a water treatment company

A water treatment company starts sending the short survey after every breakdown, and an annual question to its contract customers, two months before the renewal.

A property manager gives a 5 out of 10 and writes that he has to chase the report every time. The office manager calls him the same week. She sees that the office sent the reports all at once at the end of the week. They agree that from now on the technician will send the report on site. In her list she records the cause as “report”, and at the next annual question she calls the property manager again.

How it works in Digitalnatie Service Suite

Service Suite does not send satisfaction surveys and does not store a satisfaction score: you do the survey with your own form or a phone call. What you do find are the operational figures, and a place to record the customer’s answer.

• Response time: every intervention has a priority, with a default response time of 72 hours (normal), 48 hours (important), 24 hours (urgent) or 8 hours (critical). These times, set by your organisation, determine the deadline on the intervention, with an SLA deadline signal at the top of the intervention.
• What is late: the service overview shows late interventions and those to tackle first; the list of interventions has views for overdue interventions and for urgent and overdue ones. In the reports, the technicians tab shows the backlog for each technician.
• Arrival: the progress on the intervention shows the time of each step, such as arriving on site. The reports have no figure for arriving on time: you compare the arrival time with the time window yourself.
• First-time fix: it is not recorded, and the reports show that no data is available. You count it yourself: filter the list of interventions on breakdowns only and a period, and export it as a CSV file. The outcome of each visit is on the intervention.
• Report sent: a view of the list shows completed interventions whose report has not been sent yet. In the reports, on the interventions tab, you filter on signature and on sent.
• The customer’s answer: record it as a note in the timeline of the customer file, visible only to your staff. Each contact person has their job title.

Good to know:

• The automations have no action for a survey. They send an appointment or payment reminder by email, or create an internal notification, a follow-up activity, an intervention to schedule or a draft invoice.
• In the customer portal, a customer reports a problem with an installation. A survey, a free-form request or a message thread is not possible there.
• The deadline follows the priority. If you have agreed a shorter response time with a contract customer, choose the matching priority when they report an issue.

You schedule the work behind these figures in the planning board. How to keep customers in the long run is explained in customer retention.

Frequently asked questions

How do you measure customer satisfaction in a service company?

Ask a short question after every intervention, for example how satisfied the customer was on a scale of 1 to 5, and ask contract customers once a year whether they would recommend you. Count the answers the same way every time, and put them next to your own figures, such as arriving on time and first-time fix.

What is the difference between CSAT, NPS and CES?

CSAT measures how satisfied a customer is with one intervention, usually on a scale of 1 to 5. NPS measures whether they would recommend you, on a scale of 0 to 10, and is about the whole relationship. CES measures how much effort it took them to get help.

How do you calculate NPS?

Ask how likely it is that the customer would recommend you, on a scale of 0 to 10. Calculate the share of answers 9 and 10 and the share of answers 0 to 6, and subtract the second from the first. The score lies between −100 and +100.

Which KPIs go with customer satisfaction?

Besides a score such as CSAT or NPS: the share of interventions with arrival within the time window, the first-time fix rate, the share of reports sent on the same day and the share of issue reports with a response within the agreed time. Compare them with your own previous months.

Can Service Suite send a satisfaction survey?

No. Service Suite does not send surveys and does not store a satisfaction score; you do that with your own form or a phone call. You record the answer as an internal note in the customer file.