NPS vs CSAT in hospitality: what each metric is actually for

The NPS vs CSAT choice depends on what you are trying to diagnose. NPS measures a guest’s willingness to recommend the business, making it useful as a broad loyalty signal across sites; CSAT measures satisfaction with a specific experience, making it better for operational touchpoints such as check-in, cleanliness or speed of service.

Neither score explains what happened. Guest experience metrics become useful only when a number is paired with an open comment and segmented by venue, visit type and time of day.

What does NPS measure in hospitality?

NPS is a relationship metric, not a diagnostic tool.

The standard NPS question is:

“Based on your most recent visit, how likely are you to recommend [venue or brand] to a friend or colleague?”

Guests answer on a 0 to 10 scale. Scores of 9 and 10 are promoters, 7 and 8 are passives, and 0 to 6 are detractors. NPS is calculated by subtracting the percentage of detractors from the percentage of promoters, producing a score between -100 and 100.

For a 30-site pub group, that gives the ops director a consistent measure for comparing broad guest sentiment across the estate. If one pub’s NPS has fallen for three consecutive periods while comparable sites remain stable, there is a reason to investigate.

But NPS does not tell you whether the cause was slow drinks, a cold dining room or a disappointing Sunday roast. A guest can also recommend a lively city-centre restaurant despite finding the toilets poor, or refuse to recommend an excellent hotel because its location did not suit their trip.

Treat any hospitality NPS benchmark carefully. A benchmark is meaningful only when the question wording, invitation timing, guest mix and collection method are comparable. A score collected from loyalty members after a complimentary stay should not be compared with one collected from every direct-booking guest.

NPS is best used to spot movement and differences, not to declare that one site is objectively “good”. Once the score changes, the operational work begins: reading the comments, finding the pattern and briefing the site team on what to fix.

When should you use CSAT instead of NPS?

CSAT is the better metric when a manager can name the exact part of service being assessed.

A CSAT question usually asks guests how satisfied they were with a specific experience. For example:

“How satisfied were you with the speed of service during your visit?”

A five-point response scale might run from “Very dissatisfied” to “Very satisfied”. The customer satisfaction score is commonly reported as the percentage selecting the top two positive responses. Some teams use an average instead, which is acceptable if the calculation remains consistent and clearly labelled.

Consider a hotel testing a new check-in process. NPS is too broad because the guest’s recommendation could be influenced by the room, breakfast, location and price. A CSAT question about check-in gives the front office manager a cleaner signal.

The same applies to a restaurant investigating Saturday dinner service. Ask about speed of service, food temperature or bill accuracy separately. Do not ask “How satisfied were you overall?” and then pretend the answer diagnoses kitchen performance.

Specificity has a cost. Asking guests to rate twelve touchpoints produces a long, repetitive survey and encourages straight-line answering. CSAT works when it is selective, not when every department adds another compulsory rating question.

Where does a CES score fit?

Customer Effort Score measures friction, particularly when guests are trying to complete a task or resolve a problem.

A typical question is:

“How easy was it to change your booking?”

The response scale could run from “Very difficult” to “Very easy”. Keep the scale direction consistent. Reversing it between questions is an excellent way to create bad data from attentive guests.

A CES score belongs around transactional moments: booking a table, amending a hotel reservation, using mobile ordering, requesting a refund or resolving a complaint. It is less useful for judging whether someone enjoyed a meal or felt welcomed by the team.

Imagine a guest whose hotel room was excellent but who spent 25 minutes trying to move the booking by one night. Their CSAT for the stay could be high and their effort score poor. Both answers can be true.

CES is especially valuable after service recovery. “Was your issue resolved?” measures outcome, while “How easy was it to get your issue resolved?” exposes the six emails and three phone calls hidden behind that outcome.

Why do teams argue about the score instead of the root cause?

Metric arguments usually begin when a broad score is asked to do a specific operational job.

A regional manager sees NPS fall at a restaurant and asks the general manager to fix it. The general manager points out that the terrace was closed for refurbishment and disputes the result. Marketing argues that public reviews remain strong. Everyone debates the number because nobody collected enough detail to test the competing explanations.

Three mistakes make this worse.

The first is changing scales or wording. Moving a CSAT question from five response options to ten breaks the trend line. Rewording NPS from “recommend” to “visit again” changes the construct entirely.

The second is mixing overall and specific measures. “How satisfied were you with your visit?” cannot diagnose reception, housekeeping or breakfast. Overall CSAT is another broad sentiment measure, not a shortcut to department-level evidence.

The third is chasing targets. Once a venue is told to reach a particular NPS, staff can start coaching guests, selectively promoting the survey or treating every detractor as unfair. The score improves while the measurement gets worse. Survey bias deserves as much attention as the dashboard itself.

The NPS vs CSAT argument is therefore the wrong starting point. Begin with the decision a manager needs to make, then choose the metric capable of informing it.

How should NPS and CSAT work together?

The strongest hospitality surveys use one broad signal, one open explanation and only the specific follow-ups the response justifies.

Start with NPS when you need an estate-wide loyalty signal. Follow it immediately with:

“What was the main reason for your score?”

That comment carries the operational evidence. A guest who gives a hotel 6 and writes “Lovely room, but check-in took ages and nobody explained breakfast times” has identified two concrete issues. The next question should explore check-in, not ask for another generic overall rating.

This is where static survey design becomes wasteful. A fixed form asks every guest about food, service, cleanliness, atmosphere and value, even when their opening comment has already covered three of those subjects.

Service Monitor Adapt is an AI-adaptive customer feedback platform built for hospitality businesses. It cross-checks the opening score against the guest’s written sentiment, then adjusts the follow-up questions according to what the guest has and has not mentioned.

If someone scores a pub 2 but writes “Had a great time”, the survey can ask them to confirm the apparent mismatch. If they say “The food was bad”, it can ask what specifically was wrong rather than accepting a complaint the kitchen cannot act on.

The principle is simple: scores classify the experience, while comments explain it. Adaptive questioning then fills the important gaps without forcing every guest through the same questionnaire.

Which metric should your hospitality business lead with?

Lead with the metric closest to the management decision, then keep the wording and scale stable.

Use this decision framework:

1. Lead with NPS for brand and site comparison. Use it when senior leaders need a consistent signal across a pub, restaurant or hotel estate. Pair it with an open comment and analyse movement over time rather than obsessing over a single month’s rank.

2. Lead with CSAT for a defined service touchpoint. Use a focused CSAT question when a manager owns the process, such as breakfast quality, room cleanliness, order accuracy or check-in. Avoid combining several concepts in one question, such as “How satisfied were you with the food and service?”

3. Lead with CES for friction and resolution. Use it after booking, digital ordering, amendments, complaints and refunds. CES should answer whether the process was easy, not whether the overall visit was enjoyable.

Whichever side of the NPS vs CSAT debate you started on, segment the results by venue, visit type and time of day. A restaurant’s overall CSAT can look stable while Friday dinner satisfaction falls sharply and weekday lunch improves. A hotel’s NPS can conceal a gap between business travellers staying midweek and families arriving on Saturday afternoon.

Add other segments only when they inform a decision: first-time versus returning guests, direct versus third-party bookings, dine-in versus takeaway, or event guests versus regular visitors. Do not slice a small response set into impressive-looking but unreliable fragments.

If you want to combine an estate-wide score with shorter, response-specific follow-ups, Adapt can help you build a feedback flow that gives senior leaders consistency and site managers usable detail.

Frequently Asked Questions

Is NPS or CSAT better for hospitality?

NPS is better for tracking broad recommendation intent across venues and over time. CSAT is better for measuring satisfaction with a specific touchpoint such as check-in, cleanliness, food quality or speed of service.

What is a good hospitality NPS benchmark?

There is no useful universal hospitality NPS benchmark because collection methods, guest profiles and survey timing differ. Compare like-for-like sites using consistent wording, then focus on trends and the reasons guests give for their scores.

What is the best CSAT question for a restaurant?

The best CSAT question names one operational area, such as “How satisfied were you with the speed of service during your visit?” Avoid broad or double-barrelled questions that combine food, service and atmosphere.

Can NPS, CSAT and CES appear in the same survey?

Yes, but they should not be added automatically. Use NPS for the overall relationship, CSAT for relevant touchpoints and CES after a process involving effort, then skip questions that the guest’s comment has already answered.

How often should hospitality operators change survey questions?

Keep core metric wording and scales stable so results remain comparable. Change diagnostic follow-ups when operations change, but record the date and avoid presenting pre-change and post-change results as one uninterrupted trend.

Why should a score always be paired with an open comment?

The score shows the direction and strength of the guest’s reaction, but not its cause. An open comment turns “6 out of 10” into evidence about a breakfast queue, cold chips, a helpful receptionist or a confusing booking process.