AI Prompts to Turn Raw Data into Clear Answers
Teams across Southeast Asia spend hours staring at spreadsheets, dashboards, and reports trying to work out what the numbers are actually saying. AI tools like ChatGPT, Claude, and Gemini can speed this up by reading your figures and explaining them in plain language. The prompts below are grouped by task. Copy one, replace the bracketed variables with your own details, paste in your data or describe your dataset, then drop the whole thing into your AI tool of choice.
One important caution. AI can misread numbers, mix up columns, or invent figures that sound plausible, so always sanity check the output against your source data before you act on it or share it. Never paste confidential, sensitive, or personal data into a public AI tool. When in doubt, remove names and account numbers, or describe the shape of the dataset instead of pasting it.
Data Interpretation Prompts
1. Explain what a table of numbers shows
Use this when you have a table or spreadsheet and you want a plain-language read on what it means and where the action is.
You are a data analyst. Read the table below and explain in plain language what it shows. Identify the top 3 trends, the highest and lowest values, and anything that stands out. Avoid jargon and keep it to short, clear sentences a non-technical reader can follow. State any assumptions you made about the columns. Data: [PASTE_DATA]2. Compare two time periods
Use this when you need to know what changed between two periods, such as this month versus last month or this year versus last year.
Compare these two sets of figures for [METRIC, e.g. monthly sales] covering [PERIOD_1] and [PERIOD_2]. Tell me what went up, what went down, and the percentage change for each item. Highlight the 3 biggest movers and suggest one likely reason for each, clearly marked as a hypothesis to verify. Data: [PASTE_DATA]3. Find anomalies and outliers
Use this when you want a second pair of eyes to flag unusual values worth investigating before you trust the dataset.
Review the dataset below and flag any anomalies or outliers, such as values that are unusually high or low, sudden spikes or drops, or numbers that look inconsistent with the rest. For each one, explain why it stands out and what I should check to confirm whether it is real or a data error. Do not guess at causes you cannot support. Data: [PASTE_DATA]Chart Explanation Prompts
4. Describe a chart in plain language for a slide
Use this when you have a chart and need a clear caption or talking point that a general audience will understand on a slide.
I have a [CHART_TYPE, e.g. line chart] that shows [WHAT_IT_MEASURES] over [TIME_OR_CATEGORY]. Here is the underlying data. Write a 2 to 3 sentence plain-language description I can put on a presentation slide. Say what the chart shows, the main pattern, and the single most important takeaway for a [AUDIENCE, e.g. management] audience. Data: [PASTE_DATA]5. Suggest the right chart for my data
Use this when you are not sure how to visualise a dataset and want a recommendation before you build the chart.
I want to visualise the data below for a [AUDIENCE] audience to show [MESSAGE_I_WANT_TO_GET_ACROSS]. Recommend the best chart type and explain why. Tell me which columns go on each axis, what to use for colour or grouping, and one tip to keep it easy to read. Suggest a clear title for the chart. Data: [PASTE_DATA]Report Summary Prompts
6. Turn a long report into an executive summary
Use this when you have a long report and need a tight summary that busy leaders will actually read.
Summarise the report below into an executive summary for [AUDIENCE, e.g. senior leadership]. Keep it under [WORD_COUNT] words. Cover the key findings, the numbers that matter most, and the main recommendations. Use short paragraphs or bullet points and lead with the most important point. Report: [PASTE_REPORT]7. Pull the key numbers out of a report
Use this when you only need the headline figures and metrics from a dense document, not the full narrative.
Read the report below and extract the key numbers into a simple list or table. For each figure include the metric name, the value, the period it covers, and how it compares to the previous period if stated. Only include numbers that actually appear in the text. Do not estimate or fill in gaps. Report: [PASTE_REPORT]8. Rewrite a report summary for a non-technical reader
Use this when a summary is too technical and you need a version a general audience can understand without losing the meaning.
Rewrite the summary below so a non-technical reader can understand it. Replace jargon and acronyms with plain words, explain what each key number means in practical terms, and keep the tone [TONE, e.g. clear and friendly]. Keep all the facts and figures accurate to the original. Summary: [PASTE_REPORT]Insight Extraction Prompts
9. Extract the 3 most important insights and actions
Use this when you want the dataset boiled down to a few insights with clear next steps you can act on.
Analyse the data below and give me the 3 most important insights. For each insight, state the finding in one sentence, the evidence from the data that supports it, and one recommended action. Rank them by impact, most important first. Flag any insight where the data is too limited to be confident. Data: [PASTE_DATA]10. Generate questions the data raises
Use this when you want to dig deeper and need a list of smart follow-up questions to guide your next round of analysis.
Look at the data below and list the 5 most useful follow-up questions I should investigate next. For each question, explain why it matters and what extra data or breakdown would help answer it. Focus on questions that could change a decision about [TOPIC_OR_GOAL]. Data: [PASTE_DATA]Browse More Prompts
These data analysis prompts pair well with our business and writing collections. For more ready-to-use templates across every workflow, browse the full library at PromptCraft Asia. Every prompt is free, uses fill-in-the-blank variables, and works on ChatGPT, Claude, and Gemini. No signup required.