Cash Reconciliation from 3 Hours to 5 to 10 Minutes: A Franchisee's ChatGPT Work Playbook

Hi, it's Shii-chan!

OpenAI News openai.com

Today I'm covering a customer story from OpenAI's News page about a restaurant owner using ChatGPT Work.

What was announced?

OpenAI's News page features Ted Taylor, a former 25-year semiconductor industry veteran who is now an independent franchisee running three co-branded Schlotzsky's and Cinnabon restaurants in New Mexico.

He uses ChatGPT and ChatGPT Work to speed up pricing, labor, and back-office tasks. The story reports a 95% cut in repricing time and 4 hours saved on each payroll run.

What changed in practice

Menu repricing

Annual repricing weighs recipe costs, item popularity, local competitor prices, portion sizes, and the different economics of in-store and delivery orders.

• The process dropped from about 5 days to 1 to 2 hours
• He first has ChatGPT Work research pricing targets and strategy, then provides a categorized item list with current in-store and delivery prices and the increase he wants overall and for specific items
• Adding item-level sales history and supplier prices lets the tool estimate sales impact and optimize prices against cost-of-goods targets
• The output is a spreadsheet of recommended in-store and delivery prices
• When he also ran the analysis manually, ChatGPT Work's local comparison caught increases that would have made some items uncompetitive. He now uses it for first-pass recommendations

Labor management

• With read-only access to sales and labor data, ChatGPT Work compares scheduled, budgeted, and actual hours against forecast and actual sales over the past 12 weeks. Reports flag overtime, overlong shifts, and coverage gaps for manager review
• For scheduling, he gave it a roster (role, training, availability, pay) and staffing needs at different sales levels. Comparing with his manager's schedule revealed missing rules about busy periods, which were added to its instructions. It now produces a full schedule that the manager reviews, adjusts, and enters into the scheduling system
• Estimated impact: cutting hourly labor costs by 1 to 2 percentage points of sales across all three restaurants, adding roughly $50,000 to $100,000 in annual free cash flow

Back office

• Cash reconciliation: the monthly check of each store's cash transactions against bank deposits dropped from 3 hours to 5 to 10 minutes
• Payroll: summarizing payroll inputs eliminated 4 hours of manual work per cycle
• HR policies: a bodily-fluid incident procedure required by New Mexico was drafted in about 30 seconds. A social media policy draft was reviewed by an outside HR firm, which made no changes
• Request portal: he built a web portal where employees submit supply and repair requests, with alerts to a dedicated email address and status tracking

Dive Deep

Ted's approach is simple:

• Give ChatGPT the same data and guidelines he would use himself
• Compare its output with his own work
• Expand its use as his confidence grows
• Always review results before acting

He values what the time makes possible, such as analysis he would otherwise skip and more time for customers, employees, and growth. Since this is a customer story, the numbers are based on his own reports.

Wrap-up

• Repricing: about 5 days to 1 to 2 hours (95% less)
• Cash reconciliation: 3 hours to 5 to 10 minutes; payroll: 4 hours saved per run
• Better scheduling could add roughly $50,000 to $100,000 in annual cash flow
• The method: compare against your own work, then widen use gradually

This one is for small business owners handling the back office themselves!