Net Promoter Score (NPS)

Purpose

Measure users' likelihood to recommend the product to others, producing a single score (-100 to +100) that serves as a proxy for overall satisfaction and perceived value. NPS answers a different question from SUS (the referenced method): SUS asks “Is this usable?”; NPS asks “Is this valuable enough that you would stake your reputation on recommending it?”

In A3.2, NPS provides a post-testing loyalty signal. In A3.3, longitudinal NPS tracking (Weeks~2, 4, 6, 8) reveals whether perceived value increases with sustained use (healthy) or declines after novelty fades (concerning). NPS is a leading indicator of retention: Promoters retain; Detractors churn.

When to Use

Use NPS when:

• After A3.2 testing sessions (post-SUS, as a final question)
• At regular intervals during A3.3 pilot (Weeks~2, 4, 6, 8) for longitudinal tracking
• Comparing prototype variants (which version generates more Promoters?)
• Benchmarking against industry or competitor NPS
• Providing A3.4 evaluation with a standardised loyalty metric

Do NOT use when:

• As the sole satisfaction measure—NPS measures loyalty intent, not usability; combine with SUS
• With fewer than 20 respondents—NPS is highly sensitive to small sample sizes (one Detractor in a sample of 10 swings NPS by 10 points)
• Without the follow-up question—a score without context is a number without meaning
• To diagnose specific problems—NPS identifies overall sentiment, not which features cause it (use think-aloud, the referenced method, and user interviews, the referenced method, for diagnosis)

Sample Size and Duration

A3.2: 20–50 respondents (all test participants); single measurement post-session

A3.3: 50–250 respondents per survey wave; 4 measurement points over 8 weeks

Administration time: 1 minute per respondent

Analysis time: 1–2 hours per measurement wave (scoring + follow-up theme analysis)

Prerequisites

• User interaction: Respondents must have meaningful experience with the prototype (completed tasks in A3.2, or ≥1 week of pilot use in A3.3)
• Standardised question: The 0–10 scale and exact wording must be preserved for benchmarkability
• Follow-up question: “What is the primary reason for your score?” (open-ended)
• Sample: ≥20 for directional signal; ≥50 for reliable estimates with confidence intervals

Complete Procedure

Step~1: Administer NPS (1 minute per respondent)

A3.2: Administer after SUS, at the end of the testing session. Two questions only:

1. “On a scale of 0 to 10, how likely are you to recommend [product] to a friend or colleague?”
2. “What is the primary reason for your score?”

A3.3: Send NPS survey via email or in-app prompt at Weeks~2, 4, 6, and 8. Keep the survey to these two questions—response rates drop dramatically with additional items.

Step~2: Calculate NPS (15 minutes per batch)

1. Count Promoters (9–10), Passives (7–8), Detractors (0–6)
2. Calculate percentages of total respondents
3. NPS = %Promoters - %Detractors

Example (A3.3 C001 Smart Checkout, Week~4, n = 80):

Step~3: Interpret Against Benchmarks (15 minutes)

A3.2 target: ≥30 (indicates users would recommend after initial testing experience).

A3.3 target: ≥30 sustained through Week~8 (indicates value persists beyond novelty).

Step~4: Analyse Follow-Up Responses (1–2 hours)

Categorise open-ended responses:

• Promoter themes: What drives enthusiasm? (features, speed, simplicity)—these are the value proposition's strongest elements; protect in A4
• Detractor themes: What drives dissatisfaction? (missing features, bugs, confusing UX)—these are the highest-priority iteration targets
• Passive themes: What would convert Passives to Promoters? (often small improvements with high leverage)

Step~5: Track Longitudinally (A3.3 only)

Plot NPS over time (Weeks~2, 4, 6, 8):

• Rising NPS: Users discover more value over time (strong signal)
• Stable NPS: Consistent satisfaction (healthy)
• Declining NPS: Novelty wearing off; underlying problems emerging (investigate immediately)

Cross-reference with cohort retention (the referenced method): declining NPS typically precedes retention decline by 1–2 weeks, making NPS an early warning system.

Quality Criteria

1. Standardised question: Exact wording and 0–10 scale preserved
2. Follow-up included: Open-ended “why” question administered with every NPS survey
3. Adequate sample: ≥20 per measurement point; ≥50 for reliable estimates
4. Benchmark-contextualised: Score reported with industry/competitor comparison
5. Longitudinal tracking: A3.3 NPS measured at ≥3 time points (trend, not snapshot)
6. Actionable analysis: Promoter and Detractor themes identified with specific improvement recommendations

Theoretical Foundation

Seminal references:

• : Introduced NPS in the Harvard Business Review, demonstrating that the single question “How likely are you to recommend [product] to a friend or colleague?” correlates more strongly with revenue growth than any other satisfaction measure. While subsequent research has debated whether NPS is truly the “one number you need to grow,” its simplicity and widespread adoption make it the most benchmarkable loyalty metric available.
• : Expanded the NPS framework with the closed-loop follow-up process: after scoring, ask Detractors “What would we need to change?” and Promoters “What do you value most?”—transforming a score into actionable feedback.

Contemporary references:

• : Provided statistical guidance for NPS in UX research: sample size requirements (minimum 20 for directional signal, 50+ for reliable comparison), confidence intervals, and the relationship between NPS and task-based performance metrics.
• : Positioned NPS as a product team's “north star” metric for product-market fit, with ≥40 indicating strong fit and <0 indicating fundamental value proposition problems.

The NPS Question and Classification

The question (exactly as administered):

“On a scale of 0 to 10, how likely are you to recommend [product name] to a friend or colleague?”

Classification:

NPS formula:

Range: -100 (all Detractors) to +100 (all Promoters). Passives are excluded from the calculation but included in the denominator.

Challenges and Solutions

Challenge 1: Small Sample Volatility

Symptoms: NPS swings from +40 (Week~2) to +15 (Week~4) with only 25 respondents per wave.

Solutions: Report confidence intervals. With n = 25, NPS 30 has 95% CI of approximately 20—the true NPS could be 10–50. Aggregate across waves for more stable estimates. Minimum 50 respondents for reliable point estimates.

Challenge 2: Cultural Bias

Symptoms: Users in some cultures rarely give 9–10 scores (“10 means perfect; nothing is perfect”), deflating NPS. Users in other cultures rarely give low scores (social desirability), inflating NPS.

Solutions: Compare NPS within the same population over time (trend matters more than absolute level). If pilot spans multiple markets, segment NPS by market and interpret against market-specific benchmarks.

Challenge 3: NPS Without Follow-Up

Symptoms: Team reports “NPS is 35” with no context. Decision-makers ask “What does that mean? What should we do?”

Solutions: Always include the follow-up question. Report NPS alongside: (a) benchmark comparison, (b) Promoter/Detractor theme analysis, (c) trend direction, (d) correlation with retention data. NPS without context is a vanity metric.

Relationship to Other Methods

NPS receives input from:

• Task-Based Usability Testing (the referenced method): User's task experience shapes their NPS response in A3.2
• A3.3 Pilot Experience: Sustained usage shapes longitudinal NPS in A3.3

NPS provides input to:

• A3.4 Evaluation: NPS is a key desirability-lens metric and leading indicator for viability
• Cohort Analysis (the referenced method): NPS segmented by cohort reveals whether newer users are more or less satisfied
• A3.5 Decision: NPS trend (rising, stable, declining) is direct evidence for go/no-go

NPS is complemented by:

• SUS (the referenced method SUS measures usability; NPS measures loyalty—both needed for complete picture
• User Interviews (the referenced method): NPS follow-up themes often warrant deeper exploration through structured interviews
• Customer Success Management (the referenced method): Detractor identification triggers proactive outreach