Montgomery County Food Bank cuts overtime after Claude-assisted shift changes, CEO says

Montgomery County Food Bank changed its work shifts after Claude analyzed routing and delivery requirements, and biweekly driver and warehouse overtime fell from 99 hours to 11. In a case study published by Anthropic on October 26, CEO Scott Burns credits that analysis for much of the reduction—one of several customer-reported gains across purchasing, reporting and planning.

The Texas nonprofit says its 46 employees and more than 10,000 unique volunteers distributed over 12 million meals in the last 12 months. Its account describes two different uses for AI: making established work less costly, and taking on analysis the team would otherwise have sent to consultants. The results are customer claims published by the tool’s supplier, not independently verified findings.

Changing schedules and shopping habits

The overtime problem began with accommodating delivery requests. Serving 96 partner agencies, the food bank accepted every requested delivery time, and drivers and operations staff accumulated extra hours. Claude examined routing and shift patterns. The team then staggered start times so the same few drivers no longer bore the overtime burden. Burns attributes the subsequent reduction in large part to that shift analysis.

Purchasing offered a different lever. The sourcing team noticed that partner agencies were buying supplemental food at grocery-store retail prices. Using Claude in Excel, staff helped one agency compare its purchases with the food bank’s at-cost prices from retail partners. That agency switched all its food buying to the food bank and now spends 40% less to serve the same neighbors. The figure concerns that agency, not the entire partner network.

Faster reports versus work not previously attempted

The most dramatic speed claim involves a recurring spreadsheet. Data Analytics Manager Stephanie Arend says a monthly report that took a week to assemble now takes 30 seconds using Claude Fable and Claude in Excel. That is a comparison for one reporting task, rather than a measure of how much faster the organization works overall.

Burns describes the larger benefit differently: analysis the team would not have attempted before. He downloaded public Form 990 filings—the financial statements nonprofits publish—from local peers and similarly sized food banks nationwide. Claude compared government and donor funding, pounds distributed, meals and fundraising efficiency. The exercise gave the board a comparison beyond the food bank’s own year-over-year growth.

The food bank says that benchmarking would otherwise have required an outside consulting firm. Burns now also compares fundraising events to decide where staff effort returns the most. Unlike the monthly spreadsheet, this work has no reported before-and-after completion time. Its stated value is expanding the questions a small team can investigate, not simply finishing an existing assignment sooner.

A selective rollout, with more planning ahead

About 14 staff members have paid Claude seats. Marketing and Communications Manager Jay Martin reserves them for people who will use tools such as Claude Cowork in depth, and shows teams how to build projects containing their own documents. Programs, Sourcing and Operations staff also use Claude in Microsoft applications for slide decks, dashboards and one-page documents.

The next planned steps combine physical logistics with further analysis. The food bank is adding a GPS routing system this year. Burns also plans to use Claude Fable on a strategic plan he aims to finish by year end, seeking an estimated return for each fundraising and program effort so he can direct resources toward reaching more neighbors.