Allocating a stock among stores: by demand and limits against the proportional
Allocates the warehouse stock among stores: first the display minimum, then one pack at a time to the store with the lowest cover while the cover is below the target and there is room. Compares with allocation proportional to demand: cover, the number of stores below target and the expected lost sales over the horizon.
At a glance
- What question does it answer?
- How much to send to each store so that stock cover evens out and how much better that is than allocation proportional to demand.
- What data is needed?
- The warehouse stock, the shipping multiple, the target cover in weeks.
- How to read the result?
- Example on the default values (650 units, six stores): 650 units sent; the average cover rose from 0.3 to 2.9 weeks; lost sales 278 units against 284 for the proportional allocation. Example on the default values: Sent to stores: 650 units; Left in the warehouse: 0 units; Need to reach the target cover: 928 units; Average cover before: 0.3 weeks. The figures are illustrative: enter your own to see your result.
Why and how to reason
The warehouse stock has to be distributed to stores, and it is easy to err: proportionally to demand seems fair, but ignores that some stores already have stock, others have goods in transit, and a third need a minimum for display. The calculation builds the distribution by rules. First a display minimum for each store. Then by pack to the store with the lowest coverage (stock ÷ weekly demand), while coverage is below target and there is room. Coverage counts goods in transit. For example, the store “Center” with demand of 52 a week and stock of 18 is covered for 0.35 weeks, to reach the target 4 weeks it needs 4 × 52 − 18 = 190 units. All six stores need 928 units to reach target while the warehouse has 650, so everything is distributed. Lost sales over 4 weeks are 278 units against 284 with proportional distribution, a gain of 6 units or 9 thousand of margin.
When to use it
Reach for it when distributing a limited batch of a novelty, a leftover or a clearance across a network of stores with different demand and stock. Do not use it if stock exceeds the need: then everyone gets enough. Assumptions: weekly demand is stable, the dispatch multiple (pack) is fixed, the store maximum is limited by space, goods in transit arrive before the end of the horizon. It is also useful for seasonal novelties, where the first dispatch determines the sales of the first weeks.
How to read the result
Look at the minimum coverage after distribution (2.9 weeks at a target of 4) and the number of stores below target (6 of 6): there is simply too little stock, and the rule merely evens out the shortage. Lost sales with and without distribution (278 against 928) show the value of distributing, and the gain against proportional (6 units) is modest: here the fact of distributing matters more than the method. If the gain is large, there is a strong imbalance of stock between stores. Decision: if stock is not enough, think of transfers between stores or a new order. Check whether there are stores where distribution hit the maximum: that is a signal that the space limit hampers sales.
Common mistakes
Typical trap: not entering goods in transit, setting a zero display minimum, taking last season’s demand for a new store, setting a maximum below the minimum, ignoring the dispatch multiple. For a new store without history take the demand of a similar one. Do not forget to update stock and goods in transit before each distribution: the calculation goes stale within days.
What data is needed?
- The warehouse stock, the shipping multiple, the target cover in weeks.
- Weekly demand, stock and goods in transit by store.
- The display minimum and the capacity.
Inputs
| Field | Unit | Value in the example | Meaning |
|---|---|---|---|
| Warehouse stock to allocate | units | 650 | How many units of this SKU or group can be sent to stores now. |
| Target cover | weeks | 4 | How many weeks of sales a store’s stock must cover after sending; the allocation follows this target. The example value is illustrative. |
| Loss horizon | weeks | 4 | Over how many weeks the lost sales are counted if the stock is not enough. |
| Shipping multiple | units | 2 | A pack or a pair: sending in multiples of this number. Footwear, accessories and children’s have their own values: take the category data. |
| Margin per unit | ₽ | 1,500 | For valuing the losses in money. Empty: losses in units only. |
How it is calculated
| Step | Formula |
|---|---|
| Minimum | first the display minimum, stores in descending order of demand |
| Cover | (stock + in transit + sent) ÷ weekly demand |
| Step | a pack to the store with the lowest cover below the target where there is room |
| Losses | max(0, demand × horizon − stock after sending) |
Worked numbers
Default values from the app; in the interactive version you enter your own.
- Sent to stores
- 650 units
- Left in the warehouse
- 0 unitsNot sent because the cover reached the target, there is no room or the rest is smaller than a pack.
- Need to reach the target cover
- 928 units
- Average cover before
- 0.3 weeks
- Average cover after
- 2.9 weeks
- Lowest cover after
- 2.9 weeks
- Stores below the target after
- 6
- Lost sales with no allocation
- 928 units
- Lost sales with this allocation
- 278 units
- Lost sales with allocation proportional to demand
- 284 units
- Gain against proportional, units
- 6
- Gain against proportional, margin
- 9,000 ₽
- Before
- After
Data as a table
| Measure | Before | After |
|---|---|---|
| Center | 0.4 | 2.9 |
| Mega | 0.3 | 2.9 |
| High street | 0.4 | 2.9 |
| Outlet | 0.1 | 2.9 |
| Suburb | 0.6 | 2.9 |
| New mall | 0.1 | 2.9 |
- This allocation
- Proportional to demand
Data as a table
| Measure | This allocation | Proportional to demand |
|---|---|---|
| Center | 132 | 134 |
| Mega | 162 | 164 |
| High street | 96 | 98 |
| Outlet | 124 | 116 |
| Suburb | 52 | 56 |
| New mall | 84 | 76 |
Allocation by store
| Store | Demand a week | Stock with transit | Send | Cover after, weeks | Proportional | Lost, this | Lost, proportional |
|---|---|---|---|---|---|---|---|
| Center | 52 | 18 | 132 | 2.9 | 134 | 58 | 56 |
| Mega | 64 | 22 | 162 | 2.9 | 164 | 72 | 70 |
| High street | 38 | 14 | 96 | 2.9 | 98 | 42 | 40 |
| Outlet | 45 | 6 | 124 | 2.9 | 116 | 50 | 58 |
| Suburb | 22 | 12 | 52 | 2.9 | 56 | 24 | 20 |
| New mall | 30 | 4 | 84 | 2.9 | 76 | 32 | 40 |
Demand is a forecast: for stores where the item was out of stock past sales are understated. The allocation equalises cover and not sales: a store with high conversion on a size may deserve more, you add that rule by hand. For footwear and children’s allocate by size, for accessories by article.
Nuances and limits
- Taking the past sales of a store where the item was out as demand.
- Allocating by style and not by size: the shortage in sizes is not visible.
- Forgetting goods in transit: a store gets too much.