AI prompts for NPS, CSAT, and SUS: score analysis, follow-up coding, and CX programs

These prompts help UX researchers and CX teams use AI tools to analyze standardized metric data — from coding thousands of open-ended NPS responses to calculating SUS scores and designing measurement programs. Replace [bracketed placeholders] with your specifics.

Prompt 1: Code NPS follow-up responses by promoter/detractor group

I have NPS survey data with open-ended follow-up responses. The question was: "What is the primary reason for your score?"

Data format: each row has [respondent_id, nps_score (0-10), follow_up_text]

[Paste or attach the data]

Please:
1. Group respondents into Promoters (9-10), Passives (7-8), and Detractors (0-6)
2. For each group separately, identify the top 5-8 themes in the follow-up responses
3. Create a codebook: theme name, description, 2-3 example quotes, frequency count
4. Highlight themes that appear in one group but not others (what Promoters love that Detractors don't mention, and vice versa)
5. Write an executive summary (5-7 sentences) covering the most important finding from each group
6. Suggest 3-5 actions the product team could take based on the Detractor themes

Prompt 2: Calculate and interpret SUS scores

I administered the System Usability Scale (SUS) to [number] participants after [context: a usability test / a product evaluation / a periodic assessment].

Here are the raw SUS responses (10 items, each rated 1-5):
[Paste the data — columns: participant_id, Q1 through Q10]

Please:
1. Calculate the SUS score for each participant using the standard formula (odd items: rating minus 1; even items: 5 minus rating; sum × 2.5)
2. Calculate the overall mean SUS score with 95% confidence interval
3. Interpret the score using the adjective rating scale (Bangor et al., 2009): below 50 = not acceptable, 50-70 = marginal, above 70 = acceptable; and the letter grade scale: below 51 = F, 51-68 = D, 68-80 = C, 80-90 = B, above 90 = A
4. Compare to the industry average of 68
5. If there are segment variables in the data, calculate SUS scores per segment and test for significant differences
6. Identify which individual SUS items have the lowest scores — these point to specific usability dimensions that need attention

Prompt 3: Build a multi-metric CX measurement program

I want to set up a customer experience measurement program for [product name/type].

Context:
- Product type: [B2B SaaS / e-commerce / mobile app / etc.]
- Current measurement: [what, if anything, is being tracked now]
- Key stakeholder questions: [what do executives/product team want to know?]
- User journey stages: [list the key touchpoints: awareness, onboarding, active use, support, renewal]

Please design a measurement program that includes:
1. Which instrument (NPS, CSAT, SUS, CES, UMUX-Lite) to deploy at each journey stage, and why
2. The exact trigger for each survey (when and where it appears)
3. Frequency caps to prevent survey fatigue
4. Target sample sizes per instrument per period
5. Benchmark sources for each instrument (where to find industry comparisons)
6. A reporting cadence and format recommendation
7. A dashboard mockup description (what charts and metrics to display)