← All use cases

Use case B20 · Business

Coffee shop franchisee uses ChatGPT to find weak food sales and design a profitable promotion

Yagnesh Mehta, owner of a PJ's Coffee branch (via The Baltimore Banner) ·

The short version

What they did

Mehta used ChatGPT to analyze his cafe’s sales data, which showed food sales were underperforming. It then helped him design a Mother’s Day promotion ($17.99 for beignets and two coffees) priced to keep some profit, and a flyer for it.

Result (as claimed)

The article says he “used the technology to analyze sales data that helped him determine that PJ’s food sales were underperforming and decide what to do about it.” The promotion offered a discount while maintaining some profit, “so we’re not giving away food.” The AI-generated flyer drew criticism when he posted it to a local Facebook group: “With an AI mom?” one commenter wrote.

Prompt

I run a coffee shop and want to understand weak food sales before planning a promotion for [holiday]. My last three months of sales are [attach item sales, units and prices]. Item costs and fees are [paste], and franchise pricing limits are [paste or say none].

1. Compare food categories with the rest of our sales. Show the figures behind any category you call underperforming.
2. Suggest three food-and-coffee bundles for the holiday. Show regular price, proposed price, customer discount, cost and margin for each.
3. Explain the trade-offs in a short paragraph so I can choose one.

Return tables with clear calculations and plain words. Don’t propose a bundle below its supplied cost. If costs or fees are missing, ask before calling a promotion profitable. These are ideas to test, not a sales forecast.

Do it yourself

These steps are my suggested way in, not Yagnesh Mehta’s setup.

What you need

  • Three months of sales by item from your point-of-sale system.
  • Your food cost for each item.
  • A chat assistant that can read spreadsheets, such as ChatGPT or Claude.

Start simple

  1. Export sales by item and add a column for food cost.
  2. Upload the file to a chat with the Prompt and name the holiday or season.
  3. Ask which categories are lagging and why it thinks so.
  4. Pick one of its bundle ideas and ask for the price, discount and margin in a table.
  5. Run the promotion for a set number of days.

Take it further

  • Compare sales during the promotion with the same days before it.
  • Ask for flyer and social post copy. Use a photo of your real food, not an AI picture. The AI flyer here drew public criticism.
  • Repeat for each slow category.

Check your work

  • Check the margin math yourself. Price minus food cost and fees must still be positive.
  • Check the totals it reports against your point-of-sale report.
  • If you’re a franchisee, check pricing and marketing rules with your franchisor first.

Before you rely on it

  • This is general information, not financial, tax or accounting advice for your situation.
  • AI tools make arithmetic mistakes and misread statements, and they sound just as sure when they do. Check every number against your own records before you act on it.
  • For decisions with tax, loan, insurance or retirement consequences, talk to a qualified professional who knows your situation.

Source checked. The link, author, date and quote were confirmed against the source. The result itself has not been reproduced. Prompt adapted for this library, inspired by the case. Business size: Not stated. Results are quoted as the source claimed them and never improved. Spot an error or a broken link? Email charlie@cbarmorecpa.com.

More in Finance for owners

New cases go out every Saturday in Inside AI.