AI Prompts for Gemini for Analysis
20 tested prompts across 4 stages. Works with ChatGPT, Claude, and Gemini.
Most people try to use AI for Gemini for Analysis with a single vague prompt and get generic results. This guide takes a different approach: 4 targeted stages, from Frame the analysis through Communicate findings, each with a prompt that gives the AI exactly the context it needs. Use Gemini to analyze information, evaluate options, and surface insights that drive better decisions. Works with ChatGPT, Claude, and Gemini.
Stage 1
Frame the analysis
Start here to define the question your analysis needs to answer before touching the data.
Define the question
I need to analyze [DESCRIBE THE SITUATION]. Help me clarify the core question and break it into 3-5 sub-questions I need to answer to reach a decision.
Choose an analysis framework
I need to analyze [DESCRIBE THE SITUATION]. What framework fits best: SWOT, Porter's Five Forces, cost-benefit, risk matrix, or root cause analysis? Explain your recommendation.
Identify what I need to know
To make a decision about [DESCRIBE DECISION], what information do I need to gather? List the key facts, data points, and unknowns in priority order.
Spot potential biases
I'm analyzing [DESCRIBE SITUATION]. What cognitive biases might affect my thinking? How do I check for confirmation bias, sunk cost fallacy, and overconfidence in this specific situation?
Structure the analysis
Create an analysis structure for [DESCRIBE SITUATION]. Define: the objective, key questions, data needed, methodology, and decision criteria.
Stage 2
Run the analysis
These prompts help you execute the analysis and surface the insights that matter.
SWOT analysis
Run a SWOT analysis for [COMPANY / PRODUCT / INITIATIVE]: [DESCRIBE IT]. Give 4+ items per quadrant and identify which strengths and opportunities to double down on.
Evaluate options
I have these options: [LIST OPTIONS]. Evaluate each against these criteria: [LIST CRITERIA]. Build a decision matrix and recommend the best option with reasoning.
Cost-benefit analysis
Do a cost-benefit analysis for [DECISION OR INVESTMENT]. List all costs (direct, indirect, opportunity costs) and all benefits (tangible and intangible). What is the net assessment?
5 Whys root cause
I have this problem: [DESCRIBE PROBLEM]. Use the 5 Whys technique to help me find the root cause. Start with the symptom and work down to the underlying cause.
Risk analysis
Create a risk analysis for [PLAN OR DECISION]. For each risk: likelihood, impact, and mitigation approach. Focus on high-likelihood, high-impact risks first.
Stage 3
Interpret findings
Use these prompts to make sense of your results and draw the right conclusions.
Synthesize insights
I have these findings: [LIST FINDINGS]. Synthesize them into 3-5 key insights that directly answer my original question: [STATE THE QUESTION].
Identify implications
Given these analysis findings: [LIST FINDINGS], what are the implications? What should I do, change, stop, or start based on this information?
Challenge the conclusion
I've reached this conclusion: [DESCRIBE CONCLUSION]. Challenge it. What evidence contradicts it? What assumptions might I be wrong about?
Find the blind spots
Review my analysis: [DESCRIBE IT]. What important factors, perspectives, or data am I missing? What blind spots do you see?
Write a recommendation
Based on this analysis: [SUMMARIZE FINDINGS], write a clear recommendation with: the recommended action, 3 key reasons, risks, and immediate next steps.
Stage 4
Communicate findings
These prompts help you present your findings in a way that drives decisions.
Write an executive summary
Write an executive summary of this analysis: [DESCRIBE OR PASTE FINDINGS]. Cover: the question analyzed, the key finding, and the recommendation. Under 200 words.
Prepare the presentation
Create a structure for presenting this analysis to [AUDIENCE]. How do I sequence: context, problem, analysis, options, recommendation, and next steps?
Prepare for questions
I'm presenting this analysis to [AUDIENCE]: [DESCRIBE]. What are the toughest 5-7 questions they will ask? Write concise answers.
Translate for non-technical audiences
Translate these analysis findings for a non-technical audience: [PASTE FINDINGS]. Remove jargon, use plain language, and focus on the business implications.
Write a one-pager
Summarize this analysis as a one-page briefing: [DESCRIBE THE ANALYSIS]. Include: context, key findings, recommendation, and implications.
Frequently asked questions
What analysis tasks is Gemini best suited for?+
Strategic analysis (SWOT, competitive landscape), decision support (evaluating options, cost-benefit), data interpretation, and problem diagnosis. It is most useful when you provide structured input and ask it to apply a specific framework.
How does Gemini compare to other AI tools for analysis?+
Gemini has the advantage of integrating with Google Workspace, making it useful for analyzing documents, spreadsheets, and information from Google services. For data-heavy analysis, it can work with structured data you paste directly.
Can Gemini analyze my business data?+
You can paste data samples or summaries into Gemini and it will analyze patterns, identify trends, and suggest interpretations. For large datasets, use it to write analysis code or interpret summary statistics you have already computed.
How do I avoid getting biased analysis from Gemini?+
Provide balanced input, explicitly ask Gemini to challenge your assumptions, and request counter-arguments to your preferred conclusion. Use AI analysis as a structured thinking partner, not as an authority.
What is the best way to use Gemini for competitive analysis?+
Provide information about your market, competitors, and positioning, then ask Gemini to apply frameworks like Porter's Five Forces or competitive matrix analysis. For current competitive data, pair Gemini with Perplexity or direct research.
Other prompt packages you might like
AI Prompts for ChatGPT for Analysis
Use ChatGPT to structure complex analysis, evaluate options, and surface insights to make smarter decisions..
See prompts
AI Prompts for Claude for Data Analysis
Data analysts spend hours on the wrong things: cleaning messy data, writing boilerplate code, and translating findings into language that non-technical stakeholders understand.
See promptsAI Prompts for Perplexity for Research
Researchers often struggle to find comprehensive information quickly.
See prompts