Supplier Stock

The problem

There are SKUs you don't hold in your warehouse but can still sell: the supplier has them and ships them. On a marketplace, though, only what you publish exists — and you publish what Odoo says you have.

So there are two ways out, and both are bad. Either you don't list them, and stop selling product you could have delivered. Or you list them with a made-up quantity, and sooner or later you sell something the supplier no longer had.

And on a marketplace that second one is not a stock discrepancy: it is a cancellation, a penalty and a hit to seller reputation — the thing your position in the channel's search results depends on.

How it works

The supplier's list is uploaded with Odoo's importer. Each upload is a dated snapshot: re-uploading corrects the figure rather than duplicating it.

From there comes the published figure:

• What's available is what the supplier declared minus what has sold since that snapshot's date. Not since midnight today: if the supplier hasn't restocked, yesterday's sales stay deducted.
• Only confirmed orders count. A quotation nobody closes doesn't hold pieces.
• It rounds down. It never publishes an extra unit over a decimal.

The confidence factor. Suppliers overstate. Rather than arguing over the list, a percentage gets published: at 80%, out of 100 pieces you offer 80. It gets tuned by who actually delivers, supplier by supplier, without ever having that conversation.

The snapshot expires. Each supplier has its validity in days. The one who stopped sending their list stops selling by inertia, instead of going on offering stock from three months ago.

And the result is added to your own stock in the same field the connector already reads, so the marketplace simply sees more availability. The integration on the other side isn't touched. Each row remembers how much this module contributed, so republishing recalculates instead of accumulating.

What it does not do

A SKU sourced from both sides gets deducted twice. If a product has its own stock and the supplier's, the sale is subtracted from both: from your own stock the usual way, and again from the declared quantity. The effect is conservative — it never oversells — but it can pull a product from sale that was actually there. It is designed for SKUs that come only from the supplier.

There is a window where the contribution disappears. The connector deletes and rebuilds its sync table every four hours; this module restores its share every hour. In between, for up to an hour, the channel sees only your own stock: it under-publishes, never over. There is a button to close that window by hand.

It does not buy or order anything. It publishes availability. When the sale comes in, fulfilling it is still an errand with the supplier.

It does not verify what the supplier declares. It trusts the file, corrected by the factor. Against a supplier who overstates badly, the tool is to lower their percentage, not to detect it.

The code has to match. A supplier SKU that doesn't exist in your catalogue is left pending and publishes nothing; it gets resolved by creating the product or fixing the code.

What changes

The catalogue you offer stops being limited by what fits in your warehouse. You start selling what your supplier holds — but with a brake: netting out what already sold, publishing a percentage rather than the full figure, and dropping whatever stopped being confirmed.

Which is the difference between widening your catalogue and widening your cancellation rate.