Net Promoter Score (NPS) is a loyalty metric built from one survey question: "How likely are you to recommend us to a friend or colleague?", answered on a 0-10 scale.
You subtract the percentage of detractors (0-6) from the percentage of promoters (9-10), which gives a score between -100 and +100.
Fred Reichheld of Bain & Company introduced it in "The One Number You Need to Grow" (Harvard Business Review, December 2003).
Net Promoter and NPS are registered trademarks of Bain & Company, NICE Systems and Reichheld.
How an NPS survey is scored
| Score | Group | Counts toward NPS? |
|---|---|---|
| 9-10 | Promoters | Yes, added |
| 7-8 | Passives | Only in the total respondent count |
| 0-6 | Detractors | Yes, subtracted |
Formula: NPS = (promoters / respondents x 100) - (detractors / respondents x 100).
Worked example: 200 people answer. 110 score 9-10, 50 score 7-8 and 40 score 0-6. Promoters are 55%, detractors are 20%, so NPS = 55 - 20 = 35. The passives lower both percentages but never show up in the score directly.
Customer effort score (CES)
Customer effort score asks how easy it was to get something done, such as resolving a support issue or completing a task.
Matthew Dixon, Karen Freeman and Nick Toman of CEB (now part of Gartner) made the case for it in "Stop Trying to Delight Your Customers" (HBR, July-August 2010), arguing that reducing effort builds loyalty more than delighting customers does.
CEB introduced CES in that article and argued it predicted loyalty better than NPS or CSAT.
Customers rate how easy or hard a single interaction was.
The two metrics answer different questions. NPS tracks overall sentiment toward a brand.
CES measures one interaction, which makes it closer to the friction that CRO teams try to remove.
NPS in CRO and A/B testing
NPS is a poor primary metric for an A/B test.
Few visitors answer the survey, scores move slowly, and the result says nothing about whether a change raised conversion rate or revenue.
The Keiningham et al. study (Journal of Marketing, 2007) used data from 21 Norwegian firms and could not replicate the claim that NPS predicts growth clearly better than other satisfaction measures.
When an NPS survey does appear in a test, it usually works as a secondary or before-after signal.
Avid moved its left-side hover navigation to a top navigation bar with mega-menus and saw +49% NPS supporters in a VWO case study.
That was a before-after comparison from an on-site NPS survey, not a randomized test of conversion, so it shows sentiment moved but not that sales did.
Survey answers are more useful as a source of test ideas.
The free-text follow-up ("Why did you give that score?") and post-purchase questions show which objections are holding buyers back.
Kasper Dolk turned survey objections such as "is it durable?" into benefit bullets in a product page hero, which reported +$763,679.14 profit per year.
Oliver Kenyon used survey answers to write three reassurance lines above a collection grid for a bird food brand, worth $44,890 a month.
@oliverkenyon
@kasper_dolkRelated terms
- Survey questions for CRO research: how to write the follow-ups that turn scores into test ideas.
- Website survey tools: tools that run on-site NPS and CES surveys.
- The full CRO glossary.