When Three Tabs Tell Three Stories

The Company

Quinta Distribuidora imports artisan olive oils, tinned fish, and regional wines from small Portuguese producers and sells to independent restaurants, hotel groups, and delicatessens across the Iberian peninsula. Twelve people. Founded in 2009 by two brothers who are still on the road most weeks. The warehouse is in Mafra, the sales office is in Porto, and the admin lives, uncomfortably, in both.

The Pain

Inês runs customer service. On any given morning she's fielding a call from a hotel buyer in Bilbao asking whether the smoked sardine tins are back in stock, a restaurant group in Lisbon chasing a late pallet, and a delicatessen in Madrid wanting to know when the new harvest olive oil arrives.

The answers live in three places. Stock quantities are in Odoo, their ERP. Confirmed supplier delivery dates are in a shared Google Sheet the purchasing manager updates when he remembers. Open customer orders are in a second Google Sheet that Inês maintains herself. None of them reliably agree. Odoo shows units on hand but not units already promised to other customers. The purchasing sheet is usually a few days out of date. So Inês spends the first part of every morning piecing the picture together before she can say anything confident to a caller. When she gets it wrong — and it happens — there's a refund, an apology, or an order that doesn't come back.

The Move

Every Friday afternoon, Inês exports two things: the open-orders sheet from Odoo as a CSV, and the purchasing Google Sheet as a downloaded spreadsheet. She pastes both into a Claude conversation alongside a short prompt she wrote once and saved in a Notes document.

The prompt asks Claude to read both files and produce a plain list: which SKUs are oversold against confirmed stock, which supplier deliveries are running late relative to promised customer dates, and which customers have an open order touching either problem. It takes Claude about forty seconds. Inês reads the output on Monday morning before the phones ring, and she already knows where the trouble is.

No integration. No automation. One export, one paste, one standing question. It took her an afternoon to get working.

The blind spot

Everyone assumes the problem is the tooling. It's not. Nobody checks these numbers until a customer's already chasing them. Doing it on a fixed schedule, before anything goes wrong, isn't a technical change. It's just a decision to do it weekly instead of reactively.

The pattern

The same move fits anywhere data lives in more than one place and someone spends their morning manually piecing it together:

• A recruiter copying candidate status from an ATS into a hiring-manager spreadsheet every Monday
• An account manager reconciling invoiced amounts in Xero against project hours logged in Harvest before client calls
• A logistics coordinator cross-checking carrier confirmation emails against an internal shipment tracker before the daily standup