Customer satisfaction score (CSAT) is a survey-based metric that captures how satisfied customers are with a specific experience, such as a support call, a chat, a delivery or a product feature. Customers answer a direct question like “How satisfied were you with your experience today?” on a short scale, and the score is usually the percentage who chose a satisfied answer. CSAT is the most widely used pulse check in contact centers and service teams because it is simple to ask, simple to understand and quick to react to.
At a glance
- CSAT asks customers to rate their satisfaction with a recent interaction or product, typically on a 1–5 or 1–7 scale.
- The score is most often the percentage of satisfied responses, though some companies report the average rating.
- It works best when sent right after a defined moment and kept to one question plus an optional comment.
- Low response rates, survey timing and selective sending can skew results.
- CSAT is usually read alongside NPS, CES and operational metrics such as first contact resolution.
What problem it solves
Managers need to know whether customers are leaving interactions happy, and they need to know quickly, by team, channel and process, not once a year. Without a consistent measure, judgments rest on complaints and anecdotes, which overrepresent the worst cases and miss slow declines.
CSAT gives a fast, comparable signal tied to specific moments. A drop after a policy change, a new IVR menu or a platform migration shows up within days. Paired with customers’ comments, it points to what went wrong. The common mistake is treating CSAT as a number to celebrate rather than a prompt for action: a high average can hide response bias, poor timing and the customers who never answer.
How it works
The question. One satisfaction question, such as “How satisfied were you with [the interaction]?”, with a consistent scale across channels. An optional “Why?” free-text box usually gives the most useful information.
The calculation. On a 5-point scale, responses of 4 and 5 are commonly counted as satisfied: CSAT equals satisfied responses divided by total responses, times 100. On 7-point scales, many teams count 6 and 7. Whatever the rule, keep it fixed and state it in reports.
Timing. Surveys are triggered automatically right after the moment being measured, such as the end of a call or chat, a completed order or a resolved ticket. Delays reduce response rates and blur what the customer is rating. Frequency caps prevent over-surveying.
Sampling and bias. Very happy and very unhappy customers respond more often than the rest. Compare like with like (phone with phone, chat with chat), wait for enough responses before drawing conclusions, and watch for selective surveying.
Closing the loop. Very negative responses trigger follow-up. Comments are tagged by theme, by people or with sentiment analysis, and recurring themes go to the teams who own the cause. In the contact center, CSAT is often combined with quality management (QM) reviews for coaching.
What drives it. Fast resolution, clear answers, competent and courteous help and consistent policies. Routing through the automatic call distributor (ACD), staffing from workforce management (WFM), up-to-date knowledge and good call quality all feed into it. First contact resolution (FCR) is closely linked to satisfaction in most service settings, so track both.
When it matters for buyers
- Choosing a contact center platform. Check that post-interaction surveys are built in across channels and that results link to the interaction, agent and recording.
- Contracting with an outsourcer. If CSAT is a performance target, specify the question, scale, formula, sampling and who sends the survey.
- Measuring a change. Capture a baseline before a platform migration or process change so you can see its effect.
- Voice and video quality. Choppy calls depress satisfaction; network choices such as SD-WAN and dedicated internet access at contact center sites can help.
Our contact center as a service overview covers platforms with built-in surveys, recording and analytics.
Questions to ask vendors
- Can CSAT surveys be triggered automatically after calls, chats, emails and messages, with one consistent scale?
- Can we customize the question, scale and definition of “satisfied”?
- Are results linked to the interaction record, agent, queue and recording or transcript?
- What alerts and workflows exist for following up on negative responses?
- Do you analyze free-text comments, and can we export raw survey data?
- For outsourcers: who controls survey sending, and how do you prevent selective surveying?
How it differs from NPS and CES
Net Promoter Score (NPS) asks how likely a customer is to recommend the company, which is meant to reflect loyalty to the business as a whole; it is often sent periodically rather than after each interaction. Customer effort score (CES) asks how easy a task was, which is better at pinpointing process friction. CSAT sits between them: it asks how satisfied the customer was with a specific moment, capturing their overall reaction, including outcome and tone. Many organizations use all three, each matched to the decision it informs. None of them directly measures whether an issue was resolved, which is why CSAT is often paired with first contact resolution.
Problems this helps with
- No reliable way to know whether support interactions went well
- Satisfaction surveys get few responses and nobody acts on them
- Outsourcer reports high satisfaction but complaints keep rising
- Need to show whether a new platform or process improved service
