StockSweep StockSweep | Help Center

Transfer Opportunities Explained

How StockSweep decides that dead stock at one location should move to another instead of being discounted, and what the confidence score is based on.

Last updated: September 23, 2026

Contents
Show all sections (16) ▼

A product can be dead at one location and selling steadily at another. Discounting it is the wrong answer in that case — you would take a markdown on inventory that has a full-price buyer a shipment away. Transfer Opportunities finds those pairs and tells you which units to move where.

Estimated time: 4 minutes

What you'll accomplish

  • Understand when a transfer is suggested instead of a discount
  • Know exactly how suggested quantity and confidence are calculated
  • Know what the suggestion does and does not account for
  • Know why a suggestion expires

Requirements

  • Plan: Enterprise (multi-location transfers).
  • At least two locations with sales history, synced and analysable.
  • Found at Dead Stock → Transfers.

When a suggestion is created

The engine only looks at unresolved dead-stock rows already flagged as selling at other locations. For each one it scans every analysable location belonging to you and, among the peers that sold the variant in the last 30 days, picks the one with the highest sales velocity as the target — not simply the first eligible one. In short: the source has stock it cannot move, the target is the fastest-selling peer for that same variant.

If no peer location is selling the variant, no transfer is suggested — and the product stays a normal discount candidate.

How the numbers are calculated

The arithmetic is deliberately simple and worth knowing before you act on it:

  • Target need = (daily velocity × 30) − target inventory, with a floor of 1 unit. The 30 days is a coverage horizon: the suggestion aims to top the target up to roughly a month of cover, not to move everything.
  • Suggested quantity = the smaller of source inventory and target need. It never suggests shipping stock you do not have, and never suggests overshooting the target's month of cover.
  • Potential value = suggested quantity × unit price.

The confidence score

Confidence starts at 50 and is adjusted by how convincing the target's demand is:

  • Velocity bonus — up to +30, reaching the cap at 2 units/day or faster.
  • Recency bonus — +15 if the target sold the variant within the last 7 days.
  • Empty-shelf bonus — +5 when the target currently has zero stock, which is the strongest "ship now" signal there is.

The result is clamped to 0–100. A score in the 90s means a fast-moving target that sold this week and is out of stock; a score near 50 means a peer that sells the variant, but slowly and not recently.

What the suggestion does not include

Stated plainly, because it changes whether a transfer is worth it:

  • Shipping cost is not modelled. The potential value is the retail value of the units moved, not a net figure after freight. For low-value or bulky items, check that the freight does not exceed the markdown you were trying to avoid.
  • Lead time is not modelled. If the transfer takes three weeks and the target's 30-day cover is the whole point, that matters.
  • It does not move anything. A suggestion is a recommendation. Creating the actual transfer or stock movement stays in your hands and in your commerce platform.

Why a suggestion disappeared

Pending suggestions carry an expiry and are marked expired nightly once it passes. This is deliberate: a suggestion is a snapshot of two inventory positions and one velocity reading, and all three drift. An old suggestion is more likely to be wrong than useful, so it is retired rather than left to age on the page.

A suggestion will also stop being relevant if the source sells through, the target restocks, or the dead-stock row itself resolves.

Verify it is working

  • Open Dead Stock → Transfers. Each row names the source location, the target location, the suggested quantity and the confidence.
  • Cross-check one row against Dead Stock → By Location — the source should show the stock, the target should show the sales.

Troubleshooting

No transfers are suggested at all

Most common causes, in order: only one analysable location; no dead-stock row is flagged as selling elsewhere; or the peer location's sales history has not synced. Transfers need per-location sales, not just per-location inventory.

The suggested quantity looks too small

That is the 30-day coverage horizon working as designed. It tops the target up to about a month of cover rather than rebalancing the whole position, so you are not shipping stock that will sit dead at the second location instead of the first.

FAQ

Does a transfer suggestion stop the product being discounted?

No. The product stays on your Dead Stock list and can be discounted normally. The transfer is an alternative, not a lock.

Can I transfer to a location that has never sold the item?

The engine will not suggest it — demand at the target is the entire basis of the recommendation. You can still move stock yourself for any reason.

Does the confidence score account for shipping cost?

No. Confidence measures how likely the target is to sell the units, nothing about the cost of getting them there.