Best Perplexity Prompts for Business Data

“Analyse this spreadsheet” sounds decisive until you discover it has treated a blank as a zero and counted the subtotal as another sale. Start with a smaller job and tell it what a row means. That one sentence has rescued more spreadsheets than any amount of enthusiastic charting.

Start with the data you can use

Perplexity supports file analysis on eligible accounts, including CSV uploads. Check your plan and workplace permissions first. For these exercises, use a fictional or approved sample. Remove irrelevant columns and supply the units, reporting period and meaning of one row. Use a clean export with a single header row if possible.

Read the infographic as text

1. Name the decision: Say what will change.
2. Check the input: Inspect missing values.
3. Define the metric: Agree units and scope.
4. Show the calculation: Show counts and formula.
5. Reconcile the totals: Check in the spreadsheet.
6. Write the next action: Give it an owner.

One row in this approved sample represents [unit]. Check for missing values, duplicate [ID], inconsistent date formats and mixed units. Report counts and five example rows. Keep blanks distinct from zeros. Suggest corrections and leave the source unchanged.

Calculate monthly conversion rate using qualified enquiries as the denominator and converted enquiries as the numerator. Group by enquiry month. Show counts and formula for each month, list exclusions and flag any zero denominator. Reconcile the counts to the supplied file.

Compare [metric] across [period A] and [period B]. Show absolute and percentage changes separately. Break down the change by [segment] using the same definitions. Flag tiny groups, partial periods and missing data. Separate observed patterns from possible explanations.

Find UK benchmarks for [metric] in [industry] for [date range]. Prefer original reports. Show sample size, population, period and metric definition beside each figure. Explain which figures are comparable with our definition [definition]. Keep private company figures out of the search.

Our proposed interpretation is [claim]. Identify alternative explanations in this approved data. For each, say what you can test now and what additional data you would need. Look for seasonality, missing periods and changes in the mix of customers. Avoid claims of causation.

Use only the verified figures below [figures and sources]. Write a one-page brief for [audience]. Give the decision, the two findings that matter, their limits and a next step with an owner. Mark unsupported points as open questions.