CSAT vs NPS: Which Metric to Use and When

CSAT and NPS answer different questions. Customer satisfaction score (CSAT) tells you how a customer felt about one specific interaction, such as a support ticket, an onboarding call or a delivered project. Net Promoter Score (NPS) tells you how loyal customers are to your company overall, based on how likely they are to recommend you. Most growing teams need both, and the real work is deciding when each survey fires, who sees the answer and what happens next.

TL;DR

• CSAT is transactional: it scores a single touchpoint and is best for spotting what to fix quickly.
• NPS is relational: it scores the whole relationship and is best for tracking loyalty and churn risk over time.
• CSAT is usually the percentage of positive ratings (4 or 5 on a 5-point scale). NPS is the percentage of Promoters minus the percentage of Detractors, on a scale from -100 to +100.
• Use CSAT after events, NPS on a recurring schedule, and store both on the customer record so you can compare them by account.
• The metrics only matter if a low score triggers a task for a named person. Automate the routing, not just the survey.

Table of Contents

• What each metric measures
• How CSAT and NPS are scored
• CSAT vs NPS at a glance
• When to use CSAT, NPS or both
• Common mistakes
• Running CSAT and NPS inside a CRM
• Survey design that keeps data usable
• Perspective: what we see teams get wrong
• Run both surveys in one workspace
• A starter plan for the first 30 days
• FAQ
• Sources

What each metric measures

The cleanest way to separate the two is by time horizon and scope. CSAT measures satisfaction with a specific interaction or service call, a real-time snapshot of how happy a client is with a particular project or task. NPS measures loyalty and the likelihood that someone will recommend the company, which gives a broader view of reputation and long-term sentiment.

In practice that means:

• CSAT asks: “How satisfied were you with what just happened?” The answer reflects the quality of one moment. A great support rep can lift it, and a slow reply can sink it, even when the customer loves your product.
• NPS asks: “How likely are you to recommend us to a friend or colleague?” The answer reflects everything the customer has experienced so far: the product, the pricing, the people, the renewals and the support history.

IBM describes CSAT as a view of a specific product or service experience, but not a full, ongoing picture of customer sentiment or the long-term relationship. That limit is the main reason NPS exists alongside it.

There is also a difference in who answers. Transactional CSAT reaches whoever just had an interaction with you, which might be a end user who filed a ticket. Relationship NPS is usually sent to a chosen sample of contacts on each account, often the buyer or champion. Those can be different people with different opinions, so a high CSAT and a middling NPS on the same account is not a contradiction. It is a prompt to look closer.

How CSAT and NPS are scored

CSAT

A standard CSAT question uses a 1 to 5 scale, from very unsatisfied to very satisfied. The score is the share of positive responses. IBM explains the common calculation: add the 4 and 5 ratings on a five-point scale, divide by total responses and multiply by 100.

A worked example: 100 customers answer a survey after their onboarding call. 82 choose a 4 or 5. Your CSAT is 82%. The 3s count as neutral and the 1s and 2s as negative, but only the positive group appears in the numerator.

Teams define “positive” differently depending on the scale they choose, so write your threshold down. If one team counts only 5s and another counts 4s and 5s, their scores cannot be compared.

NPS

NPS uses a single 0 to 10 question. IBM lists the standard groupings: 9 to 10 are Promoters, 7 to 8 are Passives and 6 or below are Detractors. The score is the percentage of Promoters minus the percentage of Detractors. It ranges from -100 to +100, with -100 meaning every respondent is a Detractor and +100 meaning every respondent is a Promoter.

A worked example: of 100 respondents, 50 are Promoters, 30 are Passives and 20 are Detractors. NPS is 50 minus 20, which equals 30. Passives never appear in the formula, but they affect the percentages because they are part of the total.

NPS is not a percentage and should not be written with a percent sign. That is the most common formatting mistake in dashboards, and it makes an NPS of 30 look like a 30% approval rating.

CSAT vs NPS at a glance

	CSAT	NPS
Question	How satisfied were you with this interaction?	How likely are you to recommend us?
Scope	One touchpoint or event	The whole relationship
Typical scale	1 to 5	0 to 10
Result	Percentage of positive responses	Promoters minus Detractors, -100 to +100
When it is sent	Right after the interaction	On a recurring schedule
Best for	Fixing specific processes	Tracking loyalty and churn risk
Weakness	Moves with mood and one-off events	Slow to diagnose a cause on its own
Typical owner	Support, onboarding, delivery	Customer success, leadership

When to use CSAT, NPS or both

Use CSAT when you want to evaluate a particular touchpoint. A transactional survey suits support, sales, product, onboarding and agent interactions, and the feedback is most useful when it arrives while the experience is still fresh. Good moments to trigger it:

1. A support ticket is closed.
2. An onboarding or kickoff call ends.
3. A project milestone or deliverable is accepted.
4. A first invoice is paid or a plan is changed.

Use NPS when you want to understand the health of the relationship, loyalty and possible churn risk across an account. It belongs on a calendar, not on an event. Many teams run it quarterly, and a few add a pulse shortly after the first 90 days of a new customer. Keep the cadence steady so the trend line means something.

Use both when you need to explain a movement. A dip in NPS tells you something changed. A CSAT breakdown by touchpoint, team or product tells you where. The two are complementary, not interchangeable, so a team that picks only one will either see problems without context or see context without problems.

A simple rule for small teams: if you can only run one survey this quarter, run CSAT on the touchpoint you suspect is weakest. If you already have a stable base of customers and want a board-level number, add NPS.

Common mistakes

• Treating them as the same number. An 85% CSAT after support does not mean your NPS will be high.
• Surveying too often. If the same contact gets a CSAT after every ticket and an NPS every month, response rates fall and the answers get careless. Set a cap per contact.
• Skipping the “why”. A score without a comment tells you what happened but not why. Add one open text question after the rating.
• Averaging across segments. A blended score hides the enterprise accounts that are unhappy behind the SMB accounts that are delighted. Always segment by plan, owner or lifecycle stage.
• Collecting without acting. A low score with no owner is a lost customer waiting to happen.

Running CSAT and NPS inside a CRM

Most CSAT and NPS programs fail at the handoff gap: the survey lives in one tool, the customer record lives in another and the person who could fix the problem never sees the answer. Running surveys from your CRM closes that gap, because the response lands on the same record as the deals, tickets and projects behind it.

Store scores on the record, and keep the history

Create fields or a custom object for survey responses and link each one to a contact and an account. Do not overwrite the previous value with the newest one. Keeping every response as its own record lets you chart a single account over time and compare by product, owner or process. If you are designing this structure, how a CRM data model works is worth reading first, and CRM data mapping covers keeping picklists and fields clean as the data grows.

A minimal response record usually holds:

• The metric (CSAT or NPS) and the raw score
• The date and the trigger (ticket closed, quarterly pulse, milestone accepted)
• The contact, the account and the owner at the time
• The open text comment
• A follow-up status (open, contacted, resolved)

Trigger CSAT from events

Tie the survey to the event that should cause it. A closed ticket, a completed onboarding task or an accepted deliverable fires the survey by email or text shortly afterward, while the experience is still fresh. This is standard CRM automation: a trigger, a condition (for example, skip contacts surveyed in the last 30 days) and an action.

Schedule NPS by segment

For NPS, the trigger is a date, not an event. Build a recurring rule that selects the contacts due for a pulse, such as the primary contact on each active account who has not been surveyed in 90 days, and sends the question in batches rather than all at once. Batching protects your support team from a flood of replies and makes it easier to see the effect of a change.

Route low scores to a person

This is the step that earns the program its keep. When a response comes in at 6 or below on NPS, or at 1 or 2 on CSAT, create a task for the account owner, notify their manager and mark the account for review. When a Promoter answers 9 or 10, route that to marketing or customer success as a candidate for a case study, a referral or an expansion conversation. Teams that care about retention versus expansion get a lot from this second branch.

The pattern is common across survey platforms: an interaction triggers a workflow, the survey goes out, and a negative response creates a follow-up item for a human to close. What differs is whether all of it happens in one system or across three.

Report on it

Once the scores sit on CRM records, you can report on them next to renewal dates, plan, revenue and lifecycle stage. Ask questions like: do accounts with NPS Detractors renew at a lower rate, and does CSAT after onboarding predict first-year expansion? Put the answers into your regular SaaS reporting and review them with the same cadence as pipeline. The customer lifecycle gives a useful frame for deciding which survey belongs at which stage, and customer success operations is where ownership of the process typically sits.

Survey design that keeps data usable

• Keep it short. One rating and one open question is enough for most CSAT surveys. Add a few driver questions only when you know what you will do with the answers.
• Carry context with the response. Record the product, topic or assigned rep with each score so you can slice later, without asking the customer to repeat it.
• Pick one scale and keep it. Changing from a 5-point to a 7-point scale breaks your trend line.
• Send immediately for CSAT. The longer you wait, the more the answer reflects the customer’s general mood rather than the interaction.
• Cap frequency. Decide how many surveys one contact can receive in a month, and enforce it in the automation rather than relying on people to remember.
• Close the loop. Reply to Detractors personally, and tell customers when a change came from their feedback.

Perspective: what we see teams get wrong

We talk to a lot of revenue and customer teams about this, and the pattern is consistent: the argument about which metric is better is usually a stand-in for a different problem. The team has a survey tool in one place, customer records in another and no automation connecting them, so every score arrives as a spreadsheet export weeks after it could have helped anyone.

Our view is that the metric matters less than the plumbing. A CSAT survey that reaches the customer within minutes of a closed ticket and creates a task for the right owner when the score is low is worth more than a sophisticated NPS program that gets reviewed once a quarter. That is why we built surveys, messaging, projects and automations as native modules on the same platform as the CRM, so a response can update a record, trigger a Flow and notify a person without a connector in between.

If you are starting from scratch, we would pick one CSAT trigger and one quarterly NPS pulse, store every response as its own record, and set a rule that every Detractor gets a human reply within a set number of business days. Add complexity only after that loop works.

, Josh

Run both surveys in one workspace

Strkr includes Surveys as a native module, so CSAT and NPS responses can sit on the same contact and account records as your pipeline and projects. You can trigger them with no-code Flows, send them by email or by text through Messaging, and report on them next to deals and renewals.

Pricing is published. Surveys is $59 per month per tenant, and the CRM starts at $29 per month per seat on Starter, with Pro at $59 and Scale at $119 per seat per month. See the full pricing, the Surveys module and the SMS messaging details to estimate your own setup.

If you are still comparing platforms, our practical CRM buying framework lists the criteria worth checking, including how well a CRM handles feedback and automation.

A starter plan for the first 30 days

1. Week 1: define. Choose your CSAT scale and positive threshold, and write down the NPS cadence. Decide who owns follow-up for each team.
2. Week 2: model. Add a response record with the fields listed above, linked to contacts and accounts.
3. Week 3: automate. Build one event-triggered CSAT flow, one quarterly NPS flow and the low-score routing rule. Test them on internal contacts first.
4. Week 4: review. Check response rates, read every open comment and fix anything that makes the survey annoying. Then publish the first baseline score and a target.

Do not set a target before you have two cycles of data. A baseline that rises or falls for reasons you cannot explain is not yet a metric.

FAQ

Is CSAT the same as NPS?

No. CSAT scores satisfaction with a specific interaction, usually on a 1 to 5 scale, and is reported as a percentage of positive answers. NPS asks how likely someone is to recommend you on a 0 to 10 scale and is reported as a number from -100 to +100. They measure different things and should be read together, not swapped.

Is CSAT better than NPS?

Neither is better in general. CSAT is better for diagnosing and fixing individual touchpoints because it arrives right after the event. NPS is better for tracking loyalty and churn risk across the relationship. The right choice depends on whether you need a quick fix or a long-term trend.

Can you convert NPS to CSAT?

Not reliably. The scales, questions and groupings differ, so any conversion is a rough guess. If you need both numbers, ask both questions. Keeping them separate in the CRM also lets you see when they disagree, which is often the most useful signal.

How often should you send each survey?

Send CSAT after the event that you want to measure, with a cap on how many a single contact receives in a month. Send NPS on a fixed recurring schedule, and many teams choose quarterly. Whatever you choose, keep it consistent so the trend can be compared period to period.

Should NPS and CSAT live in the CRM?

Yes, if you want to act on them. When scores sit on the contact and account records, you can route low scores to the owner, compare them with renewals and revenue, and keep the history. Surveys stored in a separate tool tend to end up in spreadsheets that nobody reviews.

Sources

• CSAT vs. NPS, NSCA News
• What is CSAT and how to calculate it, IBM
• What is Net Promoter Score (NPS), IBM
• Relationship vs transactional surveys for measuring the customer experience

Recommended

• Measuring Sales Productivity: The Metrics That Actually Matter
• What Is Conversion Rate? Funnel Benchmarks and Calculation
• What is Customer Success Operations (CS Ops)?