NPS & Customer Satisfaction Survey Analyzer
Analyze NPS and CSAT survey results you've already collected into clear findings.
What is the NPS & Customer Satisfaction Survey Analyzer?
The NPS & Customer Satisfaction Survey Analyzer is a free AI skill that interprets results from an NPS or CSAT survey you have already fielded and collected, built for food and beverage insights teams who have the scores and verbatims in hand and need the story behind them. You paste in the scores, the distribution, and any open-ended comments; it returns a clear read on what's driving promoters versus detractors, the recurring themes clustered from the verbatims, the gap between what respondents say and what the score alone suggests, and a prioritized list of the actions the data actually supports. It is built for teams sitting on completed survey data, not designing the instrument that collects it. Because it works only from what you paste in, it never fabricates a driver theme that isn't in your actual comments. Cross-checking the driver themes against live food and beverage conversation data helps confirm whether a complaint pattern in your survey shows up more broadly too.
Who it's for
- Insights analysts interpreting a completed NPS or CSAT wave
- Consumer research leads reporting satisfaction results to leadership
- Customer experience teams prioritizing fixes from survey feedback
- Brand teams diagnosing a score drop after a product or service change
What you get
- A promoter-versus-detractor driver analysis
- Recurring themes clustered from your open-ended comments
- A say-versus-score gap read where scores and comments diverge
- A prioritized action list the data actually supports
- A verbatim-quote shortlist illustrating each major theme
- A flag on themes that rest on thin comment volume
How to use it
- 1Paste the prompt into ChatGPT, Claude or Gemini.
- 2Paste your NPS/CSAT scores, distribution and verbatims below it.
- 3Answer its clarifying questions about the survey wave and business question.
- 4Ask it to compare this wave against a prior one if you have it.
See it in action
Example input
Pasting results from our Q2 NPS survey for fictional 'Harbor & Vine' meal-kit subscriptions: NPS score 22 (down from 31 in Q1), 480 responses. Common detractor comments mention 'recipes repeat too often' and 'delivery box arrived warm/melted ice pack.' Promoter comments mention 'love the recipe variety' and 'easy weeknight dinners.' We need to know what to prioritize before Q3.
Example output
NPS analysis — Harbor & Vine meal-kit subscriptions, Q2 (score 22, down from 31 in Q1, n=480)
Headline
The nine-point drop is not a single new problem — it's a delivery-reliability issue emerging on top of an existing, slower-burning recipe-variety complaint. Treat these as two separate priorities with different urgency, not one blended "customer satisfaction" issue.
Driver analysis: promoters vs detractors
Promoter comments cluster tightly around recipe variety and weeknight convenience — "love the recipe variety" and "easy weeknight dinners" are the same core promise being delivered well for this group. This tells you the core product concept is working when it works.
Detractor comments cluster around two distinct issues that should not be merged: recipe repetition ("recipes repeat too often") and delivery condition ("box arrived warm/melted ice pack"). Notably, one of your promoter themes — recipe variety — is the mirror image of your top detractor complaint. This is a real tension worth naming directly: some customers experience strong variety and others experience repetition, which suggests either a segmentation issue (some customer cohorts get repeated recipes more than others) or a tenure effect (longer-subscribed customers run out of new recipes faster). This needs a cross-tab against subscription length before you conclude the variety is inconsistent versus simply thinning out over time for loyal subscribers.
The say-versus-score gap
The delivery-condition complaint ("melted ice pack") is a newer-sounding, more urgent-sounding issue than the recipe complaint, and is the kind of operational failure that tends to hit NPS scores harder and faster than a slow-burning content complaint — worth checking if the Q2 score drop correlates with a specific week, shipping lane, or region where a delivery or packaging issue may have occurred, rather than being evenly spread across the whole response base.
Recurring themes, ranked by apparent priority
1. Delivery condition / cold-chain failure — likely the sharper, more recent driver of the score drop; investigate operationally first since this is a fixable logistics issue, not a product-design issue.
2. Recipe repetition — likely a longer-running issue that a single quarter's dip didn't create; needs a content/menu-planning fix, not a quick patch, but should not be deprioritized just because it's slower-moving.
3. Recipe variety (as a positive) — confirm what's working here so the fix to repetition doesn't accidentally undermine it.
What this data alone cannot confirm
Whether the delivery complaints are concentrated in a specific time window, region, or shipping partner — the verbatims tell you the theme exists, not its distribution. Requesting a time-stamped or region-tagged breakout of the delivery complaints specifically would let you determine whether this is a systemic cold-chain problem or an isolated incident inflating the Q2 average.
Priority actions the data supports
1. Pull a time and region breakout on delivery-condition complaints before Q3 to determine if this is systemic or isolated — this is the fastest lever if it's operational.
2. Cross-tab the repetition complaint against subscriber tenure to see if it's a new-menu-planning gap or a long-tenure fatigue issue.
3. Hold a targeted qualitative follow-up (a handful of interviews) with detractors who mentioned delivery condition specifically, since 480 survey responses give you the theme but not the full story of what went wrong.
Comment volume flag
The verbatims quoted in your input are illustrative examples, not necessarily the full distribution — before presenting "recipes repeat too often" or the ice-pack issue as the two dominant themes to leadership, confirm what share of the 480 total detractor comments actually mention each, since a vivid quote can outweigh its real frequency in memory.
Want me to draft the Q3 leadership readout based on this analysis?The prompt
Here's the start of the prompt. Download the free bundle for the full, ready-to-paste version — plus the installable Claude Skill and Custom-GPT instructions.
# Role You are a senior customer insights analyst who interprets NPS and CSAT survey results that have already been collected. You separate score-level patterns from comment-level themes, and never let one vivid quote stand in for the full distribution of feedback. # Context I'll provide - Survey type and scores: [NPS OR CSAT SCORE(S), DISTRIBUTION, SAMPLE SIZE] - Prior-wave comparison, if available: [PRIOR SCORE / WAVE] - Verbatim comments (paste a representative set): [VERBATIMS] - Business question: [BUSINESS QUESTION] - Audience for the analysis: [AUDIENCE] # Your task 1. If the scores, verbatims, or business question are missing or vague, ask up to 3 clarifying questions BEFORE writing anything.
Frequently asked questions
- What does an NPS and customer satisfaction survey analysis include?
- It includes a read on what's driving promoter and detractor scores, the recurring themes clustered from open-ended comments, where the numeric score and the verbatim feedback tell different parts of the story, and a prioritized list of actions the data actually supports. This skill produces that analysis from NPS or CSAT results and comments you have already collected.
- How is this different from the Survey Questionnaire Designer skill?
- The Survey Questionnaire Designer builds the survey instrument itself — the questions, scales, and structure — before any data exists. This skill starts after the survey has already been fielded: you paste in the scores and comments you collected, and it analyzes what they show. Use the questionnaire designer to build the NPS or CSAT survey, and this skill once responses are in hand and you need the findings, not the instrument.
- Which AI models can run this prompt?
- Any capable chat model — ChatGPT, Claude, or Google Gemini. The prompt is model-agnostic, so paste your results directly into a chat, save the prompt as a Custom GPT, or store it as a reusable skill so every survey wave gets analyzed with the same driver-and-theme discipline.
- Will it invent findings if my verbatim sample is small?
- No. It works only from the scores and comments you paste in, and if your comment sample is thin, it will flag that the resulting themes are directional rather than confirmed, rather than presenting a handful of quotes as a definitive pattern. It structures the interpretation approach and clusters what's actually there — it does not fabricate additional responses or invent a theme absent from your data.
Related skills
Assortment Gap Finder
Find the SKUs your shelf is missing before a buyer does.
Get it freeBrand Health Report Translator
Turn brand tracker scores into a plain-English story and a next-actions list.
Get it freeBrand Positioning Statement Builder
Nail the positioning your whole brand can build on.
Get it freeWant the live data behind sharper outputs?
These skills get better with real-time F&B intelligence. See what Tastewise can do for your team.