How to Use AI to Analyze Your Sales in Excel Without Knowing Formulas

Almost every business has an Excel file with the last few years of sales. Sometimes several: one per month, one per location, one built by someone who no longer works there. They hold valuable information, like which products are growing, which customers stopped buying or which months sales drop, but almost nobody gets it out because doing so takes formulas, pivot tables and hours nobody has.

AI tools changed that. Today you can upload a spreadsheet to an assistant like ChatGPT, Claude or Copilot in Excel and ask it questions in plain language, as if you were asking an analyst: what were the ten best-selling products this year, which customers bought less than last year, which month did sales drop. The AI writes the formulas, builds the tables and makes the charts.

What can you analyze in Excel with AI without knowing formulas?

• Sales by product, customer, salesperson or location, compared against the previous month or year.
• Customers who stopped buying: the ones who ordered every month and have gone 60 or 90 days without an order.
• Seasonality: which months sales rise and fall, to plan purchasing and staffing.
• Products that are growing and products that are dying, sorted by how much they changed.
• Average ticket and how it changes by channel or salesperson.
• Data cleanup: unifying customer names written five different ways, removing duplicates and splitting badly entered columns.

How do you ask the AI so the analysis comes out right?

The quality of the answer depends on the quality of the question. Before asking for the analysis, explain to the AI what the file contains: what each column means, what currency the amounts are in, whether they include VAT and what period they cover. That context prevents half the mistakes.

Then ask one thing at a time and have it show how it got the result. A good example: "This file has sales for 2025 and 2026. The Amount column is in pesos without VAT. Tell me which 15 customers bought less in 2026 than in 2025, how much each one dropped in pesos and as a percentage, and show me the table." That works far better than "analyze my sales," which produces a generic summary that doesn't help you decide anything.

AI knows nothing about your business that isn't in your data or in your question. You provide the context.

What mistakes should you watch for when analyzing data with AI?

• Dirty data: if the same customer appears as "Abarrotes López," "Abarrotes Lopez" and "López," the AI counts them as three. Ask it to clean the data first and show you the changes.
• Numbers that don't add up: always check a total against a figure you already know, like the monthly sales your accountant reported. If it doesn't match, something was misread.
• Overreaching conclusions: the AI might say sales dropped because of the season when in fact a salesperson left. You confirm the whys.
• Sensitive information: before uploading a file, check what it contains. Remove anything the analysis doesn't need, like customer phone numbers or tax IDs, and use business versions when the information is sensitive.

Who on your team should learn this?

Anyone who spends hours building reports by hand today: the admin person, the sales manager, whoever handles purchasing or inventory. They don't need to know how to code or master advanced Excel; they need to learn to prepare the file, ask good questions and check that the result makes sense. With that, a report that took an afternoon gets done in twenty minutes.

If you want your team to learn this with your business's real files instead of textbook examples, a hands-on AI course can focus on exactly this: analyzing sales, purchasing and inventory with the tools they already use, so analysis stops depending on the one person who knew how to build pivot tables.