How do you run inter-store transfers, and when do they pay off?
Who it is forGrocery chain dispatch and allocation teams, area and store operations, stock control, fresh category teams
In a grocery chain a transfer is the store-level correction of a warehouse allocation. Used well, it saves yoghurt approaching its use-by date from disposal and keeps a promotional product from becoming permanent dead stock; used badly, it moves the same case back and forth between two stores, produces nothing but trip and handling cost, and leaves an unexplained variance at the count. This guide covers the calculation behind a transfer decision, how the sending and receiving stores are chosen, why the rule differs between fresh and ambient goods, and where the record breaks at each step from request to stock posting. The calendar and vehicle fill of routine warehouse-to-store deliveries are the subject of the store delivery schedule guide.
When does an inter-store transfer pay off?
In the inventory management literature, stock movement between locations at the same echelon is called a lateral transshipment. A review published in the European Journal of Operational Research in 2011 separates two uses: proactive transfers that redistribute stock at predetermined times, and reactive transfers made to meet demand that stock on hand cannot cover. A grocery chain has both: the weekly rebalancing round is proactive, asking the neighbouring store for goods because the milk shelf emptied in the evening is reactive.
Four situations justify a transfer in grocery. The first is shelf life: if a product approaching its use-by date will not sell at the sending store, moving it to a store where it will sell prevents a disposal loss. The second is promotional and in-and-out products: an item allocated once for the period and never replenished from the warehouse can only move from a store that is not selling it to one that is through a transfer. The third is a product the warehouse does not have: if the supplier is late, another store is the receiving store's only source. The fourth is a store closure or refit; the stock is distributed to surrounding stores.
The situation in which a transfer creates cost is just as clear. If the warehouse will close the receiving store's gap within a few days anyway and the goods at the sending store will sell at full price, the transfer only changes the timing and adds trip, handling and recording cost. The right move then is an extra warehouse dispatch or bringing the next delivery forward.
| Trigger | Type | Sender → receiver | Decision rule |
|---|---|---|---|
| Fresh product approaching its use-by date | Proactive | Store that will not sell it within its selling window → fast-selling store | Markdown and disposal loss avoided plus gross profit recovered exceeds the cost |
| Promotional or in-and-out product | Proactive | Store not selling it → store selling it | Time left in the promotion clearly longer than the transfer time |
| Product not available in the warehouse | Reactive | Store with surplus → store out of stock | Gross profit lost until supply arrives exceeds the cost |
| Store closure or refit | Proactive | Closing store → surrounding stores | Stock split by remaining shelf life and receiver capacity |
| Ambient grocery with warehouse replenishment due within days | Not made | — | An extra warehouse dispatch or an earlier delivery is cheaper |
How are transfer gain and break-even quantity calculated?
A transfer decision compares three items. The first is gross profit recovered at the receiving store: the demand the receiver could not have met until its next delivery had the goods not arrived. Not all of that demand is lost, because some shoppers who find an empty shelf buy another brand or another size. According to the global out-of-stock study published by GMA, FMI and CIES in 2002, an average of 31 per cent of consumers facing an out-of-stock go to another store and 9 per cent do not buy at all; 40 per cent can therefore be used as the lower bound for the lost sales share.
The second is the loss avoided at the sending store: the loss on goods that would have been marked down or disposed of at the sender had they not been transferred. Goods the sender would sell at full price do not enter this calculation, because those goods should not be transferred in the first place. The third is the transfer cost, which has two parts: a fixed part incurred once per transfer (vehicle share, paperwork, receiving count) and a part that varies per unit (picking, packing, putting on the shelf).
The fixed part sets the break-even quantity. A transfer added to the return leg of an existing route has a low fixed cost and pays off even for a few units; a transfer that needs its own vehicle has a high fixed cost and only makes sense in large quantities. That is why the same product, between the same two stores, can pay off one day and lose money the next.
- Gross profit recovered at receiver
min(Transfer quantity, Receiver gap) × Lost sales share × Unit gross profit- Loss avoided at sender
Units going to markdown × Markdown loss per unit + Units going to disposal × Unit cost- Transfer cost
Fixed cost + Transfer quantity × Variable cost per unit- Break-even quantity
Fixed cost ÷ (Gain per unit − Variable cost per unit)Gain per unit = (Gross profit recovered at receiver + Loss avoided at sender) ÷ Transfer quantity
Using a lost sales share above 40 per cent is only justified by your own data: if sales of similar products do not rise during empty-shelf hours, shoppers are not substituting. Measuring lost sales is covered in the out-of-stock guide.
How are the sending and receiving stores chosen?
The measure for choosing the sender is days of cover, but not on its own: it is compared with the remaining selling window. If days of cover exceed the selling window, the difference is surplus stock and a transfer candidate. Only that surplus is taken from the sender; taking goods the sender could sell itself moves the receiver's gap to the sender and triggers a second transfer in the opposite direction a few days later.
The measure for choosing the receiver is the gap until its next delivery: sales rate multiplied by the days until the delivery, less stock on hand and in transit. If several stores have a gap, the one with the higher sales rate, on the route and with enough shelf capacity comes first. In fresh food the shelf life remaining on arrival must be longer than the time the receiver needs to sell that quantity; otherwise disposal is simply moved from one store to another.
Sales rate is measured as the average of recent days, but days spent out of stock are excluded. A store whose shelf was empty appears to sell slowly, and the system mistakenly treats it as a sender.
- Days of cover
Stock on hand ÷ Average daily sales- Sender surplus
Stock on hand − Average daily sales × Remaining selling window (days)- Receiver gap
Average daily sales × Days until next delivery − Stock on hand − Stock in transit- Transfer quantity
min(Sender surplus, Receiver gap, Receiver's free shelf and stockroom capacity)The result is rounded down to whole cases; if it is below the break-even quantity, no transfer is raised.
- The same product does not return to the sender within the same period; a no-return rule sits in the system
- The minimum stock the sender needs to keep visible on the shelf is not broken by a transfer
- Until the transfer arrives, the receiver does not get a separate warehouse dispatch for the same product; the same gap is not closed twice
- For a promotional product the sender is not chosen before the promotion's sales curve in that store has been read; slow first days do not always mean surplus stock
Why do transfer rules differ between fresh food and ambient grocery?
In fresh food the value of a transfer falls quickly over time. Under Article 27 of the Turkish Food Codex Regulation on Food Labelling and Consumer Information, food past its use-by date is deemed unsafe food, and Article 21 of Law No. 5996 states that unsafe food may not be placed on the market. The limit of a fresh transfer is therefore drawn by the calendar: if the shelf life remaining on arrival is not enough for the receiver to sell the goods, no transfer is made and the markdown is applied at the sender.
In fresh food the transport condition is also part of the cost. Chilled goods travel in their own temperature regime on a transfer just as on a warehouse-to-store delivery; a short transfer in an unrefrigerated vehicle puts at risk the very goods it is trying to save. Fresh transfers are therefore planned on the return leg of the refrigerated route serving those stores.
In ambient grocery the picture is reversed. Shelf life is measured in months, the sender's surplus eventually sells at full price and warehouse replenishment usually arrives within a few days. Because the loss avoided at the sender approaches zero, the whole case for the transfer rests on lost sales at the receiver; the example below shows this.
| Criterion | Fresh food (dairy, deli, fruit and vegetables) | Ambient grocery and long-life products |
|---|---|---|
| Main reason for the transfer | Preventing disposal and markdown loss | Preventing lost sales of a product the warehouse does not have |
| What determines the sender | Gap between remaining selling window and days of cover | Days of cover running into weeks and age of dead stock |
| What determines the receiver | Shelf life remaining on arrival and sales rate | Gap until the next warehouse delivery |
| Transport | In its own temperature regime, on the return leg of the refrigerated route | In the spare volume of any route |
| Time limit | Hours and days | Weeks |
| Right decision in most cases | Transfer if the route fits, otherwise markdown at the sender | Extra warehouse dispatch; transfer only if the warehouse is out |
The transfer process: six steps from request to stock posting
The physical work of a transfer is short; the recording work is long. Between the moment goods leave the sender and the moment they are scanned at the receiver, the product is on neither store's shelf and must sit in a separate status in the system: in-transit stock. If in-transit stock is not defined, the goods appear either in both stores or in neither.
Under Article 230 of the Turkish Tax Procedure Law, goods moved between a taxpayer's several workplaces and branches must also be covered by a dispatch note issued by the sender, and the note states where and to whom the goods are sent. The dispatch note is both the legal document of the transfer and the list against which the receiver's goods receipt is checked.
- RequestThe transfer suggestion is generated from the sender's surplus and the receiver's gap, or a store raises a request. The request carries product, quantity, sender, receiver and a reason code (use-by date, promotion, out of stock at warehouse, closure).
- ApprovalThe dispatch planner checks the break-even quantity and the route, then approves, rejects or changes the quantity on the line. No transfer leaves a store without approval.
- PickingThe sender picks by scanning. If the picked quantity is below the approved quantity, the difference is recorded at the moment of picking; a quantity written on paper and keyed in later is the first source of variance.
- Dispatch and transportThe dispatch entry is made at the same time as the dispatch note; the goods leave the sender's stock and move into in-transit stock. Fresh goods travel in a vehicle running their own temperature regime.
- Goods receiptThe receiver counts every item by scanning and checks it against the dispatch note. Short, over or damaged items are recorded with a reason code at receipt; nothing is received by copying the quantity on the note.
- Stock postingThe received quantity moves from in-transit stock into the receiver's sellable stock. Any receiving difference goes back to the sender and the dispatch team the same day; a transfer that stays in transit beyond the set time raises an alert.
A transfer is not a purchase; the goods move at their cost at the sender. Systems that post a transfer as a goods receipt at a new cost distort the average cost at the receiver; this is covered in the weighted average cost guide.
Where do stock record errors in transfers come from?
A transfer is the movement that changes the records of two stores at once, and it needs both sides to complete their part. When an error stays on one side, the count shows two opposite variances in two stores: a negative one in one, a positive one in the other. That pair is a sign of a recording error, not a loss, and resembles the barcode mix-up pair described in the stock count variance guide.
The common lesson of the table is that a transfer variance should not wait for count day. If dispatch and receipt are compared on the same day, the variance is found across two stores and a single dispatch note while the cause is still remembered; if it is found at the count, it blends into the two stores' monthly variance and the cause is lost.
| Error | At the sender's count | At the receiver's count | Prevention |
|---|---|---|---|
| Goods sent, no dispatch entry made | Negative variance | Positive variance | No vehicle is loaded without a dispatch entry |
| Dispatch entered, receiver did not post the receipt | No variance | Positive variance; goods stuck in in-transit stock | In-transit age is monitored, with an alert when the time limit is exceeded |
| Short-picked, dispatch note issued for the approved quantity and receiver received from the note | Positive variance | Negative variance | Picking and receiving by scanning; quantities are not typed in |
| Receiver scanned a different variant (1.5 litre instead of 1 litre) | No variance | Positive in one variant, negative in the other | A barcode that does not match the dispatch note line triggers a warning at receipt |
| The same transfer received twice | No variance | Negative variance | A dispatch note number can be received only once |
| Dispatch of a cancelled transfer not reversed | Positive variance | No variance; in-transit stock inflates | Cancellation reverses the dispatch entry automatically |
| Damaged item received but not written off | No variance | Negative variance | Damage is written off as shrinkage with a reason code at receipt |
Which measures track transfers?
The transfer scorecard answers two questions: did the transferred goods actually sell, and did the record match on both sides? The first is measured by post-transfer sell-through, the second by the receiving difference and time in transit. The return transfer rate shows the health of the decision rule: the same product coming back within a short time is proof that the wrong sender was chosen.
The ratio of transfer cost to value recovered measures the transfer policy as a whole. If it approaches 100 per cent, transfers are barely covering themselves; the break-even quantity is then raised or dedicated-vehicle transfers are left to exceptions.
In Mindpera the transfer suggestion is produced in Mind Send: a branch holding dead stock is matched with a branch selling fast, the transfer cost is included in the suggestion and the line does not go to dispatch without the planner's approval; if the suggestion was overridden, who did it, when and why is recorded. In Mind Stock shelf, in-transit and warehouse stock are shown separately; because receipts, issues and count movements are compared daily, a transfer variance shows up by product and store without waiting for count day.
- Post-transfer sell-through
Transferred units sold at the receiver within the selling window ÷ Units transferred × 100- Transfer receiving difference rate
|Units dispatched − Units received| ÷ Units dispatched × 100- Average time in transit
Σ (Receipt time − Dispatch time) ÷ Number of transfers- Return transfer rate
Transfer lines returned to the sender within a set number of days ÷ Total transfer lines × 100- Cost to value recovered
Total transfer cost ÷ (Gross profit recovered + Loss avoided) × 100
If transfer, dispatch and goods receipt data are scattered between the ERP and the warehouse system, our team carries out the setup; the movements are connected into a single data model.
Frequently asked questions
- When should an inter-store transfer not be made?
- If the warehouse will close the receiving store's gap within a few days and the goods at the sending store will sell at full price, a transfer only changes the timing; an extra warehouse dispatch or an earlier delivery is cheaper. In fresh food, no transfer is made if the shelf life remaining on arrival is not enough for the receiver to sell the goods; the markdown is applied at the sender. Below the break-even quantity, no transfer is raised.
- Does a transfer of goods between branches need a dispatch note?
- Yes. Under Article 230 of the Turkish Tax Procedure Law, goods moved between a taxpayer's several workplaces and branches must also be covered by a dispatch note issued by the sender, stating where and to whom the goods are sent. Price and value information is not required on this note. The note is also the list against which the receiver's scanned goods receipt is checked.
- Should a product approaching its use-by date be sent to another store?
- Yes, if the shelf life remaining on arrival is longer than the time the receiving store needs to sell that quantity. Thirty units of yoghurt sent to a store selling 16 a day take about two days to sell, so they must arrive with at least that much selling time. Because food past its use-by date is deemed unsafe, a transfer without enough time merely moves the disposal; mark it down at the sender instead.
- How do you find the break-even quantity for a transfer?
- Divide the fixed cost of the transfer by the net gain per unit. The net gain per unit is the gross profit recovered at the receiver plus the markdown and disposal loss avoided at the sender, divided by the quantity, less the variable cost per unit. In the yoghurt example in this guide the net gain per unit is ₺78.2; break-even is 8 units on a return leg with ₺600 fixed cost and 39 units on a dedicated ₺3,000 vehicle.
- How is transfer stock still in transit handled at a stock count?
- In-transit stock is counted at neither the sender nor the receiver. When the count starts, every transfer in transit is listed separately and matched to its dispatch note; a transfer arriving during the count is held in a separate area without being received. A transfer that has stayed in transit beyond the set time is closed before the count; otherwise it shows up as a positive variance at the receiver and an unexplained variance across the two stores.
- Which measures are used to track inter-store transfers?
- Five: post-transfer sell-through, the transfer receiving difference rate, average time in transit, the return transfer rate and the ratio of transfer cost to value recovered. The first shows whether the transfer actually brought sales, the second and third whether the record matched in both stores, and the last two the health of the decision rule.
Sources
- European Journal of Operational Research, vol. 210, no. 2, pp. 125-136: Inventory models with lateral transshipments: A review — 2011
- Food Marketing Institute, New Industry Study Explores Root Causes of Global Out-Of-Stock Dilemma (GMA, FMI and CIES study) — 2002-10-14
- Tax Procedure Law No. 213 (consolidated text), Art. 230/5 (Turkish) — 1961-01-10
- Official Gazette, 26 January 2017, No. 29960 (repeat issue): Turkish Food Codex Regulation on Food Labelling and Consumer Information, Art. 27 (Turkish) — 2017-01-26
- Law No. 5996 on Veterinary Services, Plant Health, Food and Feed (consolidated text), Art. 21 (Turkish) — 2010-06-13