Respondents are grouped into three buckets based on their 0-10 score: promoters (9-10), passives (7-8), and detractors (0-6). NPS is calculated as the percentage of promoters minus the percentage of detractors, producing a score that can range from -100 to 100.

Unlike CSAT, which is tied to a specific interaction, NPS is usually surveyed periodically (quarterly or after a meaningful milestone) and reflects the overall relationship, including product quality and pricing, not just how support handled the last ticket.

Frequently Asked Questions

Why are passives (7-8 scores) excluded from the calculation?

The methodology treats them as neither actively promoting nor detracting from the brand, so they're counted in the total respondent base but don't add to or subtract from the score itself.

Can an AI agent deployment move the needle on NPS?

Indirectly. Faster, more consistent support resolution contributes to the overall relationship customers have with a company, but NPS also reflects product, pricing, and factors well outside support, so it's a slower, noisier signal than CSAT for judging an agent specifically.

What counts as a good NPS score?

It varies enormously by industry, so compare against your own trend over time and industry benchmarks rather than a single universal number. Above 0 is generally considered acceptable, above 50 is strong, but the trend line matters more than any single reading.

Should a support team optimize for NPS or CSAT day to day?

CSAT is the more actionable, immediate signal for a support or AI agent team, since it's tied to specific interactions they control. NPS is a useful long-term health check but too slow and too broad to guide day-to-day operational decisions.