What was the result?
vendor-reportedTed Taylor, an independent franchisee of three co-branded Schlotzsky's and Cinnabon restaurants in New Mexico, says monthly cash reconciliation across the three stores used to take 3 hours and now takes 5 to 10 minutes. He also cut 4 hours of manual work from every payroll cycle, and annual menu repricing from 5 days to 1 to 2 hours.
Ted Taylor runs three co-branded Schlotzsky's and Cinnabon restaurants in New Mexico. In an OpenAI case study dated Oct 20, 2026, he describes using ChatGPT and ChatGPT Work on pricing, payroll, and cash. For monthly cash reconciliation he uploads each store's cash transaction data and bank records; ChatGPT Work compares transactions with the deposits. Payroll inputs are summarized the same way, which he says cuts 4 hours of manual work from every payroll cycle. For the annual menu reprice he uses ChatGPT Work to research local competitor pricing and recommend in-store and delivery prices; it now takes 1 to 2 hours instead of 5 days. He still reviews the output and makes adjustments himself. The first time he used it for pricing, he also ran the analysis by hand.
Who did this?
How do I try it?
- Editorial adaptation: export one month of cash transactions and the matching bank deposits for a single location, and sit with last month's reconciliation so you can check the result.
- Give both files to ChatGPT (or ChatGPT Work, as named in the case study) and ask it to match transactions to deposits and list anything that does not line up.
- Check every unmatched item against the bank statement yourself before you change the books.
- Once the monthly match is reliable, try the same review habit on payroll inputs, and keep a person on the final payroll file.
- For an annual menu reprice, give it current prices, costs and competitor notes and treat its spreadsheet as a first draft, not a price list to publish.
Editorial adaptation, not Ted Taylor's exact prompt. Attached are this month's cash transactions for [location] and the bank deposits for the same period. Match them by date and amount. List: 1) deposits with no matching cash transactions, 2) cash transactions with no matching deposit, 3) amounts that differ, with both figures. Do not change, delete or post anything in my books. I will review every item and make the changes myself.
Make it run every week
A play works once. Turn it into a routine in three small steps:
- Run it once and check the result yourself.
- Give your AI a schedule. Tell it when to run this play (every week, or every month) and to message you with what it found before you act.
- Keep a log. One line per run: what was checked, anything flagged, what you did.
What should I watch out for?
Vendor-reported in an OpenAI case study dated Oct 20, 2026, with quotes attributed to Ted Taylor. The 3 hours to 5 to 10 minutes, 4 hours off every payroll, and 5 days to 1 to 2 hours figures are his reported results, not an independent audit. A line about $50,000 to $100,000 a year in free cash flow is his estimate of what better scheduling could eventually do, not money already freed; it is not used as a result here. ChatGPT Work is the product name used in that article; a person should still review prices, payroll and books before anything is filed or published. Build it with Page: a monthly cash close page that logs what matched, what was flagged, and what you changed.
Did it work for you?
PageMotor keeps this site free.