What Does Analyze Mean?
To analyze something means to examine its parts, relationships, evidence, patterns, or implications in order to understand it more deeply.
For example, instead of simply asking AI to summarize a sales report, you could ask it to identify the main causes of a decline in sales and explain the evidence supporting each possible cause.
Analysis Is Different From Summarization
A summary focuses on the important information contained in a source.
Analysis goes further by examining relationships, patterns, implications, causes, or possible explanations.
For example:
Summary: Sales declined by 12 percent in the second quarter.
Analysis: The second-quarter sales decline was concentrated in two regions, which may indicate that the problem is regional rather than company-wide.
The second statement interprets the information rather than merely repeating it.
What Does Compare Mean?
To compare means to examine two or more things using common criteria.
A useful comparison should not simply describe each option separately. It should identify meaningful similarities, differences, strengths, weaknesses, or trade-offs.
Start With the Purpose
Before asking AI to analyze or compare something, decide what you want to learn.
For example:
Compare these two software options to determine which is more suitable for a small business.
The purpose tells the AI what the comparison is intended to accomplish.
Define the Criteria
A comparison becomes more useful when you specify the criteria.
For example:
Compare the two options based on cost, ease of use, available features, security, and customer support.
This gives the AI a consistent framework.
Use the Same Criteria for Each Option
When comparing alternatives, use the same criteria for all of them.
If one option is evaluated for cost, security, and support while another is evaluated only for features, the comparison is uneven.
A consistent set of criteria produces a more useful comparison.
Ask for Similarities and Differences
If you want a balanced comparison, ask explicitly for both similarities and differences.
For example:
Identify three important similarities and five important differences between these approaches.
This prevents the response from focusing exclusively on one side.
Ask for Advantages and Disadvantages
When evaluating alternatives, it can be useful to request advantages and disadvantages for each option.
For example:
For each option, identify the three main advantages and two important disadvantages.
Ask for Trade-Offs
A trade-off occurs when improving one characteristic may require accepting a disadvantage somewhere else.
For example, one solution may be cheaper but less flexible, while another may provide more features but cost more.
Ask AI to identify these trade-offs explicitly.
Define What Matters Most
Not every comparison criterion has equal importance.
For example:
Prioritize reliability and security over price.
This tells the AI which factors should receive greater attention.
Use Ranking When Appropriate
If you want AI to rank several alternatives, explain the basis for the ranking.
For example:
Rank these five options from most suitable to least suitable based on cost, reliability, and ease of implementation.
The criteria make the ranking more meaningful.
Ask for a Recommendation Carefully
You can ask AI to recommend an option after performing a comparison.
For example:
After comparing the options, recommend the one that best meets the stated requirements and explain the main reasons.
The recommendation should be connected to the criteria rather than simply presented as an unexplained preference.
Separate Evidence From Interpretation
When analyzing information, distinguish between what is directly supported by the source and what is an interpretation.
For example:
For each conclusion, identify the evidence from the source and then provide the interpretation.
This makes the analysis easier to evaluate.
Ask AI to Identify Evidence
When working with supplied material, ask the AI to connect conclusions to evidence.
For example:
Identify the three most significant findings and cite the relevant evidence from the supplied report for each one.
This is stronger than simply asking for three findings.
Limit the Source of Analysis
If the analysis should be based only on information you provide, state that clearly.
For example:
Base the analysis only on the information contained in the attached report. Do not introduce outside assumptions.
Be Careful With Causation
AI may identify a relationship between two events without having enough evidence to establish that one caused the other.
Instead of asking:
What caused the decline?
you may ask:
What factors in the supplied information could be associated with the decline?
This wording avoids assuming that a cause has already been established.
Ask for Possible Explanations
When the evidence is incomplete, ask for possible explanations rather than definite causes.
For example:
Identify three plausible explanations supported by the available information and clearly distinguish them from established facts.
Ask AI to Identify Uncertainty
Good analysis should recognize when the available information is insufficient.
For example:
Identify any conclusions that cannot be established confidently from the information provided.
Comparing Two Options
A simple comparison prompt could be:
Compare Option A and Option B based on cost, implementation time, reliability, and ease of use. Identify the main advantage and disadvantage of each option and recommend one based on reliability.
This defines the options, criteria, output, and priority.
Comparing Multiple Options
When comparing several alternatives, a table can make the result easier to evaluate.
For example:
Compare all four options in a table using cost, implementation time, reliability, and scalability as columns. Then identify the strongest option overall.
Use a Consistent Scale
For some comparisons, a consistent rating scale can be useful.
For example:
Rate each option from 1 to 5 for cost efficiency, reliability, ease of use, and scalability. Explain each rating briefly.
The scale creates a common basis for comparison.
Be Careful With Numerical Scores
Scores can create an appearance of precision even when the underlying information is subjective or incomplete.
If you ask AI to assign scores, explain what each score should represent.
For example:
Use 1 for very weak and 5 for very strong. Base ratings only on evidence provided in the source material.
Weighted Criteria
Some decisions require certain criteria to matter more than others.
For example:
Evaluate reliability as twice as important as cost when comparing the options.
This creates a weighted evaluation.
Ask for a Decision Matrix
A decision matrix can organize several criteria and alternatives.
For example:
Create a decision matrix comparing the four options across cost, reliability, implementation time, and scalability. Include a final overall assessment.
Analyze Text
AI can analyze written material for patterns, themes, claims, or issues.
For example:
Analyze these customer comments and identify the five most common complaints. Group comments referring to the same underlying issue.
Analyze Data
When data is provided, specify what should be examined.
For example:
Analyze the monthly sales data. Identify the largest increases and decreases, calculate the percentage changes, and identify any consistent patterns.
Analyze a Business Report
A business analysis prompt might be:
Analyze the attached quarterly report. Identify the three strongest areas of performance and three areas of concern. For each point, provide supporting evidence and explain its potential business significance.
Analyze a Proposal
You can ask AI to evaluate a proposal using defined criteria.
For example:
Evaluate this proposal based on cost, feasibility, implementation risk, and expected benefit. Identify the strongest argument and the largest weakness.
Compare Writing Options
AI can compare different versions of written content.
For example:
Compare these two versions of the email for clarity, professionalism, and friendliness. Identify which version performs better on each criterion.
Compare Different Approaches
When several approaches could solve the same problem, ask AI to compare their trade-offs.
For example:
Compare these three approaches based on implementation effort, expected benefit, risk, and long-term scalability.
Ask for a Balanced Analysis
If you want to avoid a one-sided response, request a balanced assessment.
For example:
Present the strongest arguments in favor of and against each option before making a recommendation.
Ask for Counterarguments
For important decisions, asking for counterarguments can reveal weaknesses.
For example:
After presenting the recommendation, identify the strongest argument against it.
This encourages a more critical evaluation.
Ask What Could Change the Recommendation
A recommendation may depend on assumptions or priorities.
Ask AI to identify the conditions that could change the result.
For example:
Identify the circumstances under which the second option would become preferable.
Do Not Ask for Analysis Without Defining the Subject
A request such as Analyze this provides very little direction.
A better request identifies what should be examined.
For example:
Analyze this proposal for financial risks, implementation challenges, and assumptions that may require further verification.
Do Not Ask for Comparison Without Criteria
Which one is better? is often too vague.
Better:
Which option is better for a small business with a limited budget and a need for simple implementation?
The criteria are now connected to the actual decision.
Define the Decision Context
The same option may be better in one situation and worse in another.
Provide the relevant context.
For example:
Compare these two solutions for a small company with a limited technical team and a requirement for implementation within 30 days.
Use Output Formats That Support Analysis
Useful formats include:
- Comparison tables.
- Decision matrices.
- Ranked lists.
- Advantages and disadvantages.
- Evidence and interpretation sections.
- Recommendation summaries.
Combine Analysis With Constraints
You can control the scope of the analysis.
For example:
Analyze only the financial risks mentioned in the report and identify the five most significant ones.
This combines an analytical task with scope and quantity constraints.
Combine Analysis With Structured Steps
Complex analysis can be divided into stages.
- Identify the relevant information.
- Group related information.
- Compare the groups.
- Identify significant differences.
- Summarize the implications.
This approach can make the workflow easier to evaluate.
Ask AI to Challenge Its Initial Conclusion
For important decisions, you can ask AI to reconsider an initial recommendation.
For example:
Review the recommendation again and identify any evidence that could support a different conclusion.
This can reveal weaknesses or overlooked factors.
Ask for Missing Information
If the available information is insufficient, ask AI what additional information would improve the analysis.
For example:
Identify the five pieces of additional information that would most improve this comparison.
Review AI-Generated Analysis
Analysis should not automatically be treated as fact.
Check:
- Are the facts accurate?
- Are conclusions supported by evidence?
- Are assumptions clearly identified?
- Are alternatives considered fairly?
- Are important factors missing?
- Is the recommendation consistent with the stated criteria?
Improve a Weak Analysis
If the analysis is too general, add specific criteria.
If it is too long, define the required scope.
If it is one-sided, request arguments for and against.
If the recommendation is unsupported, ask for evidence and reasoning behind the recommendation.
A Complete Comparison Prompt
Consider this prompt:
Compare Option A, Option B, and Option C for a small business with a limited budget and a small technical team. Evaluate each option based on cost, implementation time, reliability, ease of use, and scalability. Rate each criterion from 1 to 5 using only the information provided. Identify the main trade-offs and recommend the option that provides the best balance of reliability and affordability. Present the comparison in a table followed by a short recommendation.
This prompt provides context, criteria, constraints, evaluation rules, output format, and a decision objective.
A Complete Analysis Prompt
For example:
Analyze the attached customer feedback. First group similar complaints. Then rank the groups by frequency. Identify the three most significant issues and provide supporting evidence for each. Distinguish observed facts from possible explanations. Finally recommend one practical action for each issue. Present the final findings in a table.
This creates a structured analytical workflow.
Analysis and Comparison Are Decision Tools
The purpose of analysis and comparison is usually not simply to produce more text.
The goal is to understand information better, identify meaningful differences, and support a decision or action.
What You Should Remember
- Analysis examines information more deeply than a simple summary.
- Comparison works best when the same criteria are applied to each option.
- Define the purpose and decision context before asking for an analysis.
- Specify the criteria that matter.
- Ask for evidence when conclusions are important.
- Separate facts from interpretations and assumptions.
- Ask for trade-offs, advantages, disadvantages, and counterarguments when useful.
- Use tables or decision matrices for structured comparisons.
- Do not assume that an AI-generated recommendation is automatically correct.
- Ask for missing information when the available evidence is insufficient.
What Comes Next?
The next lesson will focus on Iterative Prompting and how to improve an AI interaction through successive rounds of refinement.