Regional FMCG distributor cleared €240k of slow stock before peak season

A 120-person regional FMCG distributor ran Odoo for eight years across two warehouses and 2,400 active SKUs. FMCG inventory management there was a quarterly debate about slow movers that never ended in a clear-down. After connecting Odoo read-only to Flowra, purchasing had a ranked stop-buy and clear-down list the next morning, and cleared €240,000 of slow stock before the summer peak.

The situation

The operations director owned availability and working capital for the whole catalog: beverages, snacks, personal care, and household lines with shelf-life windows between four months and two years. Two warehouses, twelve buyers and merchandisers, and a peak season that starts in late May.

Monthly reviews mixed Odoo pivot tables, buyer memory, and supplier promotion calendars. Everyone knew slow stock existed. Nobody could say which 80 lines mattered, in what order, or how much cash they tied up. By the time the list was agreed, seasonal demand had shifted again and the debate restarted.

The cost was not abstract. Slow lines occupied pallet positions needed for peak-season inbound, and the credit line was being used to fund stock that had not moved since the previous winter.

What the data showed, and didn't say

Odoo held complete sales orders, deliveries, returns, and purchase receipts for eight years. The standard inventory valuation report listed quantities and value by product category. The stock forecast view showed incoming and outgoing moves per product.

What no report answered was a ranked decision: which SKUs to stop buying, which to discount, which to return to the supplier, and in what order, given current cover and a declining sales velocity. The signal was in the transaction history. It was simply never assembled into one list with evidence attached.

Connecting Flowra

IT created a read-only Odoo user and an API key one afternoon. Flowra connected over XML-RPC to Odoo 16, ingested sales, purchase, and stock-move history, and reconstructed a stock trajectory per SKU from net transaction flow. No opening stock count was needed and nothing was written back to Odoo.

The first product risk report was ready the next morning. From then on Flowra re-scored the catalog nightly, with webhook-triggered syncs when Odoo records changed. The Odoo connector was the only integration work involved.

What Flowra surfaced

• A stop-buy list of 89 SKUs in the high and critical risk tiers: rising forward cover combined with a falling 4-week versus 12-week velocity trend.
• A clear-down list ranked by cash trapped above the cover target, with the estimated annual carrying cost for each line.
• A dead-stock onset watchlist: 31 lines still selling and still being restocked that were likely to go silent for eight consecutive weeks before the season ended.
• Evidence on every line: velocity trend, reconstructed cover, last movement date, and the weighted components behind the risk score.

What the team did

Purchasing approved the stop-buy list in the first review, overriding 11 lines where a supplier rebate depended on annual volume. Each override was logged with a reason, which later exposed three lead times in Odoo that were six weeks out of date.

Sales took the top 40 clear-down lines and ran a targeted promotion with the accounts whose RFM segment showed rising frequency on those categories. Merchandising handled returns to suppliers on the 12 lines where return terms made recovery better than discounting.

The weekly review became a daily 15-minute action review in Slack. Flowra posted the lines whose tier changed overnight, the buyer approved or adjusted, and the Monday digest gave the operations director cash released against the plan. The decisions stayed with people. The assembly of evidence moved to the assistant.

Results

"We were sitting on three months of slow stock we couldn't see in one view. Flowra flagged it in the first analysis. We cleared it before the season ended."

— Operations Director, regional FMCG distributor

What made it work

• Separate the two problems. Velocity tiers kept clearance actions away from core lines, so the clear-down never touched the top movers. The distinction is explained in dead stock versus slow-moving inventory.
• Rank by cash, not by SKU count. Forty lines carried most of the trapped cash. The method is in the overstock cash trap.
• Log the overrides. Eleven documented exceptions surfaced stale master data that no report would have caught. Every recommendation carried the fact, the forecast, and the hypotheses, so disagreeing took one line, not a meeting.