Net Promoter Score: How to Calculate NPS and Actually Act on It

Net Promoter Score reduces the messy question of "do customers actually like us?" to a single number between -100 and 100. That is its strength and its trap. Used well, NPS is an early-warning system that tells you a relationship is souring months before the customer cancels. Used badly, it becomes a vanity number a team games for a quarterly bonus. This article covers how to calculate NPS correctly, how to read the result honestly, and what to do the day after the survey closes.

What NPS actually measures

NPS comes from one question: "On a scale of 0 to 10, how likely are you to recommend us to a friend or colleague?" The genius is that recommendation is a high-bar signal. People do not risk their reputation vouching for something mediocre. A customer who would stake their name on you is worth far more than one who merely says they are "satisfied."

Responses split into three groups:

• Promoters (9-10): loyal, enthusiastic, and the source of most referrals and word of mouth.
• Passives (7-8): content but unattached. They will leave the moment a competitor is slightly cheaper or shinier.
• Detractors (0-6): unhappy. They churn, and worse, they tell other people why.

The reason 7 and 8 count as passive, not positive, surprises people. It is deliberate. NPS is tuned to be demanding because lukewarm customers do not drive growth.

It also helps to remember what NPS is not. It is not a measure of how many customers you have, how much they pay, or whether they are actively using the product this week. It is a measure of sentiment — of willingness to vouch for you — and sentiment leads the hard numbers. A promoter base that is quietly shrinking is often the first visible sign of trouble that later shows up as churn and slowing referrals. That lead time is precisely why the metric earns its place next to revenue.

How to calculate NPS, step by step

The formula is simple: NPS equals the percentage of Promoters minus the percentage of Detractors. Passives are counted in the total but otherwise ignored.

Work an example. Say 500 customers respond:

• Promoters (9-10): 260 responses = 52%
• Passives (7-8): 140 responses = 28%
• Detractors (0-6): 100 responses = 20%

NPS = 52% - 20% = +32.

A few things fall out of the arithmetic that matter in practice. First, passives are dead weight in the score — you can improve your NPS just as much by converting a passive into a promoter as by rescuing a detractor. Second, the score hides the distribution. A +32 built on 52% promoters and 20% detractors is a very different company from a +32 built on 40% promoters and 8% detractors. Always look at the three group sizes, not just the headline number.

What counts as a good score

Resist the urge to compare your number to a blog post's "world-class is 70." NPS varies wildly by industry, geography, and how you ask the question. A +32 might be excellent in one market and mediocre in another. The only benchmark that reliably means something is your own score last quarter. If it is trending up, whatever you are doing is working. If it is trending down while revenue still looks fine, you are watching churn load the chamber.

The part everyone skips: closing the loop

A score you collect and file is worthless. The value of NPS is entirely in what you do next, and that work divides cleanly by segment.

• Detractors are a to-do list. Every 0-6 is a specific person telling you something is broken. Reach out, personally, within days. A large share of detractors will stay if they feel heard and see a fix. Ignoring them guarantees the churn.
• Passives are your cheapest growth. They already do not dislike you. A small nudge — a missing feature shipped, a better onboarding, a check-in call — can flip a 7 into a 9 far more easily than winning a brand-new customer.
• Promoters are underused. Most companies never ask their happiest customers for anything. That 9 or 10 is the moment to request a referral, a review, or a case study, because the goodwill is already there and measured.

A worked follow-up plan

From the 500 responses above, you have 100 detractors and 140 passives. Suppose personal outreach saves 30% of detractors from churning and converts 20% of passives into promoters over a quarter. That is 30 rescued accounts and 28 upgraded relationships. If each account is worth $6,000 a year, the detractor rescue alone protects $180,000 in revenue — from a survey that cost you almost nothing to run. The score did not do that. The follow-up did.

Relationship NPS vs transactional NPS

There are two ways to ask the question, and they answer different things. Relationship NPS is the periodic pulse — "how likely are you to recommend us?" asked every quarter to your whole base. It tells you the overall temperature of the relationship and is the number you trend over time. Transactional NPS fires right after a specific event: finishing onboarding, closing a support ticket, using a new feature. It tells you whether that particular experience worked.

You want both, and you want to keep them separate. Mixing a post-support-ticket score into your relationship average will distort the trend, because someone who just had a problem solved answers differently from someone reflecting on the whole product. A team that runs quarterly relationship NPS plus a handful of well-placed transactional surveys gets a far richer picture than one that blasts the same generic survey at everyone once a year.

How you phrase the follow-up matters too. The score is only half the survey; the free-text "why did you give that score?" question is where the actionable detail lives. A number tells you a customer is unhappy. The sentence underneath tells you it is your billing flow, or your mobile experience, or a feature a competitor ships that you do not. Always collect the reason, and read the reasons in bulk — patterns in the comments are usually more useful than the score itself.

How to track this in Inleads

NPS only works if you can ask the question inside the moments that matter and route the answers to a human fast. Inleads treats NPS surveys as a first-class capture channel, right alongside web forms, WhatsApp, Facebook Lead Ads, and LinkedIn, so a score is not a spreadsheet you export once a quarter — it is a live event attached to a customer profile.

Because every response lands on that customer's record in the customer data platform, a detractor is not an anonymous number; it is a named account with its full history of activity, deals, and past scores. That context is what makes follow-up personal instead of generic. You can wire workflow automation so that any score of 6 or below triggers an instant WhatsApp or Slack alert to the account owner, turning "we should follow up with detractors" into something that actually happens the same day.

NPS also slots naturally into the referral stage of the AAARRR pirate-funnel analytics: promoters are your referral engine, and tracking how many of them actually refer closes the loop between satisfaction and growth. None of this requires a separate survey tool bolted onto your CRM — the measurement and the action live in the same place, which is the whole point.

Keep it honest

NPS is a thermometer, not a diagnosis. It tells you something is wrong; it does not tell you what. Never tie individual bonuses directly to the score, or you will get a team that games surveys instead of a product customers love. Read the number as a trend, always look underneath it at the promoter, passive, and detractor split, and judge yourself on how quickly you act on what it reveals. A modest score you respond to beats a great score you frame on the wall.