Electronics B2B distributor unified reorder logic across three warehouses

An electronics B2B distributor stocked fast-moving components and a long tail of accessories across three warehouses in three countries, all on Odoo 17. Each site ran its own min-max table. Global availability suffered anyway, and planners spent hours every week moving stock after a stockout had already happened. Seven years of movement history, read once through a single connector, gave the team the multi-warehouse inventory management they had been rebuilding by hand every Monday.

The situation

Around 9,500 active SKUs, three warehouses (Netherlands, Poland, Spain), and a planning team of four. Roughly 300 SKUs carried customer SLA commitments with contracted OEM accounts. A stockout on one of those lines meant a penalty or, worse, a lost design-in.

Odoo proposed replenishment per company and per warehouse. It did not know, or did not say, that the Polish site already held eleven weeks of cover on the same global SKU that Spain was about to reorder from the manufacturer with a six-week lead time.

The legacy min-max fields had been set at go-live, three years earlier. Nobody had the time to review 9,500 × 3 parameters, so they stayed.

What the data showed, and didn't say

Seven years of receipts, sales and inter-company transfers held the true global velocity per SKU and the real transfer lead time between sites (two to four days, versus six weeks from the supplier). The data also showed a recurring pattern: central stock high, regional velocity spiking, then an emergency air-freight shipment a fortnight later.

What the ERP never combined was site-level days of supply with global cover. Every warehouse was an island in the replenishment logic, even though the trucks between them ran twice a week.

Connecting Flowra

IT created a read-only Odoo user with access to the three companies and connected it through the native XML-RPC connector. Webhook-triggered syncs pick up new moves during the day; the full re-scoring runs nightly or on demand. No write permission was granted, and none was needed.

Flowra ingested the full history and reconstructed a stock trajectory per SKU per site from the net transaction flow. The first global risk report was ready the following morning, under 24 hours after access was granted.

What Flowra surfaced

• A single global reorder queue, ranked by stockout risk and revenue at stake, with a ship-from suggestion per line: transfer from a sister site or order from the supplier.
• Relocate-stock proposals on more than 60 SKUs where one site sat in the overstock tier (forward cover above eight weeks) while another was inside one supplier lead time of running out.
• Stockout risk windows mapped to SLA tiers, so contracted OEM lines were checked on a 14-day horizon and long-tail accessories on 60 days.
• A handful of lines marked unknown where the transfer history was incomplete. Flowra withheld a recommendation on those rather than guess, and listed what was missing.

What the team did

Every Monday, the head of planning opened the ranked queue and approved transfers first, then the reduced list of supplier POs. Approved actions were exported to the WMS pick-and-transfer workflow. Flowra did not write to Odoo; the planners kept the purchase order and transfer creation exactly where it had always been.

Site managers received their own view in Slack: only the lines affecting their warehouse, with the proposed action and the evidence. When the Spanish site manager disagreed with a transfer because a local promotion was about to start, the override and its reason were logged and the forecast hypothesis was corrected for the next run.

By the third week, the Monday balance spreadsheet, four tabs and roughly five hours of work, was retired.

Results

"Odoo knew how much stock we had in every warehouse. It just never told Valencia that Poznań had it. Now the transfer is the first thing on the list, and the supplier order is the exception."

— Head of Planning, electronics B2B distributor

What made it work

• One queue, not three. Ranking every site's needs in one list, by risk and revenue at stake, is what makes transfer-first possible. Per-site min-max rules can never see the sister warehouse. Stockout risk windows explains how the horizons were matched to SLA tiers.
• Lead time from history. The three-day transfer route versus the 42-day supplier route only matters if both numbers come from real receipts, not from the supplier card. See when to reorder: days of cover vs supplier lead time.
• Site managers kept the decision. Flowra routes the signal to the right role and logs the override. Adoption came from planners seeing their corrections change the next recommendation.