Analyzing data with ChatGPT

Source: OpenAI Blog

OpenAI Academy

Explore, analyze, and turn data into clear insights and actions.

ChatGPT can help you move from raw data to useful insights with minimal setup. You can upload a CSV or Excel file, paste in a table, or connect a data source (if supported in your workspace), then start asking questions in plain language.

Instead of building formulas, pivot tables, or dashboards for every question, you can quickly explore data, clean up tables, generate simple visualizations, and extract key takeaways in a format that’s easy to share.

It’s especially useful early in the process—when you’re still figuring out what’s in the data, identifying anomalies, and deciding where to dig deeper. It also helps translate findings into summaries others can review and act on.

How to get the most out of ChatGPT for data analysis

• Start with the decision you’re trying to support. A simple frame is: “I’m trying to decide ___, based on ___.” This tells ChatGPT what “done” looks like and keeps the analysis focused.
• Provide your data along with any critical context—definitions, timeframe, and what key columns represent. You can provide data via file upload, or by using a connected app.
• Ask for an approach, not just an answer. For example, request an exploratory data analysis (EDA) summary followed by hypotheses to test. This leads to more structured and reliable results than jumping straight to conclusions.
• If visuals would help, request them explicitly—specify what to plot, how to segment, and any must‑haves like axis labels or units.
• Ask for reusable outputs such as a clean final table or a short executive summary that translates findings into action.

Example prompts

1. Summarize key insights from a Shopify sales dataset

Task
Analyze this data and summarize key insights.

Context
Use the sample dataset from our Shopify store (last 30 days).

Expected output

• Structured summary of key insights, including what stands out across channels and products.
• Identification of underperforming areas (e.g., low‑converting channels).
• 4–6 prioritized observations.
• 5 specific follow‑up analyses or questions to investigate next.

2. Review and analyze a sales funnel

Task
Review and analyze our sales funnel data.

Context
Use the data from [Campaign name] from [connected analytics app].

Expected output
Produce clearly separated sections:

1. Key observed patterns in the funnel
2. Hypotheses explaining those patterns (e.g., onboarding as primary driver)
3. Recommended experiments or tests

Insights should be ranked by business impact, with emphasis on conversion bottlenecks and leverage points.

3. Identify process inefficiencies using ticket data

Task
Identify issues or inefficiencies in a process using data.

Context
Review the attached current process document, as well as the support team ticket data CSV.

Expected output

• Prioritized list of operational issues and bottlenecks (e.g., escalation delays, repeat ticket drivers), each supported by data signals.
• Clear reasoning for why each issue matters.
• Recommended areas for immediate improvement or investigation, grouped into quick wins vs deeper fixes.

Tips for trustworthy analysis

• Define “good” up front. Include the success metric you care about, the timeframe, and the groups or segments you want to compare.
• Request transparency. Ask ChatGPT to show how it arrived at its conclusions, including assumptions, formulas, and quick checks for missing data or unusual spikes.
• Set ground rules. For example, tell it not to treat correlations as causation, to flag any data limitations, and to highlight anything that looks off.
• Validate results. Before sharing or deciding, spot‑verify a couple of key numbers to ensure everything adds up.