What are the use cases for this prompt?

This prompt can be used for: Analyzing User Feedback.

What is the difficulty level and which AI tool should I use?

This prompt is intermediate level. It works with gemini.

Are there variants of this prompt?

Yes, 3 variants are available: Beginner, Intermediate, Advanced.

📊Analyse de donnéesIntermediateGemini

Prompt Gemini for Analyzing User Feedback

Analyzing user feedback is a strategic lever to improve a product, service, or customer experience. However, manually processing hundreds or thousands of reviews takes considerable time and exposes you to interpretation biases. Gemini, Google's AI model, excels in natural language processing and automates this analysis with remarkable accuracy. Thanks to its ability to understand context, detect emotional nuances, and categorize recurring themes, Gemini transforms raw text data into actionable insights. Whether you manage customer reviews on an e-commerce site, responses to an NPS survey, comments on social networks, or support tickets, a well-structured prompt will allow you to extract key trends, identify priority friction points, and quantify the overall sentiment of your audience. This guide provides an optimized prompt to fully leverage Gemini's analytical capabilities on your user feedback, with variants adapted to each expertise level.

Paste in your AI

Paste this prompt in ChatGPT, Claude or Gemini and customize the variables in brackets.

You are a senior UX analyst specialized in leveraging user feedback. I will provide you with a set of customer feedbacks. Analyze them following this structured methodology:

Sentiment Analysis: Classify each feedback as positive, negative, neutral, or mixed. Calculate the percentage distribution.
Thematic Categorization: Identify the 5 to 8 recurring themes (e.g., UX/ergonomics, price, performance, customer support, missing features). For each theme, indicate the number of mentions and the dominant sentiment.
Critical Friction Points: List the 3 most urgent issues to resolve, ranked by frequency and intensity of dissatisfaction. For each, quote a representative verbatim.
Strengths to Leverage: Identify the 3 most appreciated elements with supporting verbatims.
Weak Signals: Spot emerging trends mentioned by few users but potentially important.
Prioritized Recommendations: Propose 5 concrete actions ranked by estimated impact (high/medium/low) and implementation effort.

Format your response with tables where relevant. Use emojis for sentiment readability (✅ positive, ❌ negative, ➖ neutral, 🔄 mixed).

Here are the feedbacks to analyze:
[PASTE_YOUR_FEEDBACKS_HERE]

Personalize this prompt with Léa

Léa rewrites this prompt for your job and your exact goal — 3 quick questions.

Why this prompt works

This prompt uses role-playing by assigning Gemini an expert role, which enhances the quality and depth of the analysis. The 6-step methodology structures the model's reasoning sequentially (chain-of-thought), avoiding superficial responses. The explicit request for verbatims, percentages, and tables forces concrete and verifiable outputs rather than generalities.

Use Cases

Analyzing User Feedback

Variants

Expected Output

You will obtain a structured analysis report including a quantified overview of sentiment, a thematic mapping of feedback, and a prioritized list of corrective actions. The deliverable is directly usable in product meetings or executive committees, with quantified data and direct quotes that anchor each insight in the actual voice of users.

Frequently Asked Questions

How many feedbacks can I analyze at once with Gemini?

Gemini 1.5 Pro has a context window of up to 1 million tokens, allowing you to process several thousand feedback items in a single prompt. In practice, for very large volumes (over 2,000 feedbacks), it is recommended to group them into thematic or chronological batches for a more accurate analysis. The Gemini 1.5 Flash model also has a very large window, but its analyses will be less nuanced. Remember to structure your feedbacks (one per line or in CSV format) to facilitate processing.

How should I prepare my feedbacks before submitting them to Gemini for optimal analysis?

The quality of the analysis directly depends on the quality of the input data. Clean your feedback by removing exact duplicates, spam, and empty responses. If possible, add contextual metadata like the date, source (email, App Store, NPS survey), user segment, or the rating given. Present them in a structured format: CSV, table, or a numbered list. Avoid mixing different languages in the same batch—Gemini handles multilingual input, but the analysis will be more consistent if the feedbacks are in the same language.

Can Gemini detect sarcasm and cultural nuances in feedbacks?

Gemini handles explicit sarcasm and common ironic phrases in English fairly well (e.g., 'Great update, I love waiting 30 seconds for every click'). However, subtle sarcasm, very localized cultural references, or implicit humor can sometimes be misinterpreted. To improve detection, you can add a specific instruction in your prompt: 'Pay close attention to sarcasm and ironic expressions—when in doubt, flag the feedback as ambiguous rather than misclassifying it.' Reviewing feedbacks classified as positive with a low confidence score is a good verification practice.

Improve this prompt

Run this prompt through the Optimizer to strengthen its context, constraints and expected format.

Improve this prompt with the Optimizer

Comments

LéaAI
Pour des résultats fiables, ajoute le contexte de ton produit (type d’utilisateurs, objectifs) et segmente tes feedbacks par cohorte avant analyse. Précise aussi le volume attendu : sur un petit échantillon, demande à Gemini de distinguer les thèmes réellement présents des inférences, sinon il risque de forcer des catégories. Enfin, testez le prompt avec une dizaine de retours variés pour ajuster le format des tableaux.