What Is the Order-Up-To Level Model?
The order-up-to level model is a periodic review method of replenishing inventory. Instead of watching every item continuously and ordering when it falls to a trigger point, you check stock at fixed intervals, say every Monday or on the first of each month, and order whatever quantity brings the item back up to a target level. That target is the order-up-to level, often written as S.
The idea is simple: the timing of orders is fixed, and the quantity varies. If demand was heavy since the last review, you order more. If it was quiet, you order less, or nothing at all.
In inventory theory this is often called an (R, S) policy, where R is the review period and S is the order-up-to level. It is one of the most widely used replenishment methods in distribution and retail, because it fits the way many businesses actually buy: weekly orders to the same supplier, monthly container shipments, or scheduled deliveries to branches.
This guide explains how the model works, how to calculate the order-up-to level with a worked example, how to choose the review period and service level, how the model compares with a reorder point system, and how to run it in an ERP.
How the Model Works
Every review follows the same three steps:
- Check the inventory position. This is not just stock on the shelf. Inventory position = stock on hand + stock already on order but not yet received − quantities committed to customers or backordered.
- Compare it with the order-up-to level for that item.
- Order the difference: order quantity = S − inventory position.
Using inventory position rather than physical stock is essential. If you ignore what is already on order, you will order the same goods twice. If you ignore what is already promised to customers, you will think you have more than you do.
Calculating the Order-Up-To Level
The order-up-to level must cover demand for two periods back to back: the time until the next review (R) and the supplier lead time after that review (L). The logic is that once you place an order today, the next chance to order is at the next review, and that next order will itself take L days to arrive. So today's order must protect you for R + L days.
The standard formula is:
S = average daily demand × (R + L) + safety stock
And safety stock, when demand varies but lead time is reasonably stable, is commonly calculated as:
Safety stock = z × standard deviation of daily demand × √(R + L)
where z is the service factor for the probability of not running out that you choose. Common values:
| Target cycle service level | z value |
|---|---|
| 90% | 1.28 |
| 95% | 1.65 |
| 98% | 2.05 |
| 99% | 2.33 |
If lead time itself varies a lot, which is common with imported goods, you need a fuller formula that includes lead time variability. Our guide to safety stock covers those versions.
A Worked Example
Consider a hypothetical distributor in Jeddah selling a fast-moving bottled water carton to grocery stores. From its sales history:
- Average daily demand: 40 cartons
- Standard deviation of daily demand: 12 cartons
- Review period (R): 7 days, every Sunday
- Supplier lead time (L): 10 days, so each order is still in transit at the next review
- Target service level: 95%, so z = 1.65
Step 1, expected demand over R + L: 40 × (7 + 10) = 680 cartons.
Step 2, safety stock: 1.65 × 12 × √17 = 1.65 × 12 × 4.12 ≈ 82 cartons.
Step 3, order-up-to level: S = 680 + 82 = 762, rounded to 760 cartons.
At Sunday's review the warehouse has 180 cartons on hand, 260 cartons on last week's purchase order still in transit, and no backorders. Inventory position is 180 + 260 = 440. The order is 760 − 440 = 320 cartons.
Four weeks of reviews
| Week | On hand | On order | Inventory position | Order placed |
|---|---|---|---|---|
| 1 | 180 | 260 | 440 | 320 |
| 2 | 170 | 320 | 490 | 270 |
| 3 | 120 | 270 | 390 | 370 |
| 4 | 230 | 370 | 600 | 160 |
The pattern shows the model working: week 3 followed a heavy week, so the order was large; week 4 followed a slow week, so the order was less than half of week 3's. The day of ordering never changed. Only the quantity did.
Choosing the Review Period
The review period is the biggest design decision. Shorter periods mean less safety stock and less average inventory, but more orders, more receiving work and possibly higher freight costs. Longer periods do the reverse.
Using the same water carton, compare weekly and fortnightly reviews:
| Measure | Weekly review (R = 7) | Fortnightly review (R = 14) |
|---|---|---|
| R + L | 17 days | 24 days |
| Safety stock | About 82 cartons | About 97 cartons |
| Order-up-to level | About 762 | About 1,057 |
| Average order size | About 280 | About 560 |
| Approximate average stock (safety stock + half the average order) | About 222 | About 377 |
| Orders per year | 52 | 26 |
Fortnightly review holds roughly 70% more average stock for this item, but halves the number of orders. Which is better depends on what an order costs you versus what holding a carton costs you. That trade-off is the same one behind the economic order quantity, and EOQ can be used to suggest a sensible review interval: divide EOQ by average demand to get the number of days between orders.
Practical constraints often decide the review period: the supplier's delivery days, minimum order values, container sizes or a branch delivery schedule.
When the Order-Up-To Model Fits Best
- Many items from one supplier. Reviewing all of a supplier's items on the same day lets you combine them in one order, meet minimum order values and fill trucks or containers.
- Scheduled deliveries. If the supplier delivers only on Tuesdays, or the central warehouse supplies branches on a fixed route, periodic review matches reality.
- Branch replenishment. Many distributors replenish branches from a central warehouse to an order-up-to level each week.
- Limited staff time. Reviewing on a schedule is easier to manage than continuous monitoring when transactions are not recorded in real time.
- Medium and lower value items where some extra stock is acceptable in exchange for simpler ordering.
When it fits less well
- Very expensive items where extra safety stock is costly.
- Critical spare parts where a stockout stops production; continuous review with a reorder point reacts faster.
- Items with very lumpy, infrequent demand, where averages and standard deviations are not meaningful.
Order-Up-To Level Versus Reorder Point
| Aspect | Order-up-to level (R, S) | Reorder point (s, Q) |
|---|---|---|
| Review | Periodic, at fixed intervals | Continuous, after every transaction |
| Order timing | Fixed | Variable, when stock hits the reorder point |
| Order quantity | Variable, up to S | Fixed, usually EOQ or a standard pack |
| Safety stock covers | Review period plus lead time | Lead time only |
| Safety stock needed | Higher | Lower |
| Order consolidation | Easy | Harder, orders trigger at different times |
| Data needed | Accurate stock at each review | Accurate stock at all times |
There is also a hybrid, the (s, S) or min-max policy: at each review, order up to S only if the inventory position has fallen below a minimum s. This avoids tiny orders while keeping the periodic rhythm. For the continuous alternative in detail, see our article on the reorder point.
Setting Different Levels for Different Items
A single service level for all items wastes money. Use ABC classification to differentiate:
- A items (high value, often around 20% of items carrying most of the value): shorter review periods, carefully calculated levels, service targets chosen deliberately.
- B items: standard weekly or fortnightly review, 95% service level as a starting point.
- C items: longer review periods and simpler rules, since the holding cost of extra stock is small.
Our guide to ABC analysis explains how to classify your items.
Pack Sizes, Minimum Orders and Rounding
Real orders rarely match the calculated quantity exactly. Suppliers sell in cartons, pallets or minimum order values, and the model has to live with that. A few practical rules:
- Round up to the pack size for A and B items, so you do not shave safety stock. If the calculation says 318 and the pack is 24, order 336.
- Apply a minimum order quantity through the min-max variant: if the calculated order is smaller than the supplier minimum, either skip the order (when inventory position is still above the minimum s) or round up to the minimum.
- Balance across the supplier's range. If a supplier requires a SAR 10,000 minimum per order and the calculated lines add up to SAR 8,200, top up the C items with the longest cover rather than the A items, so extra stock lands where holding cost is lowest.
Setting Up the Model Step by Step
- Group items by supplier and delivery day so each review produces one consolidated order per supplier.
- Pull 12 months of daily or weekly sales per item and calculate average demand and its standard deviation. Remove days when the item was out of stock.
- Measure actual lead time from purchase order date to receipt for each supplier over the same period.
- Choose service levels by ABC class.
- Calculate S per item and enter it in the system as the maximum or target level.
- Run a pilot on one supplier for eight weeks, track stockouts and average stock, then extend.
Saudi Considerations
- Seasonal demand. Ramadan, Eid and summer change demand sharply for many products. Recalculate S ahead of each season using the same period from previous years, remembering the Hijri calendar shifts about eleven days earlier each year.
- Imported goods. Lead time for imports includes shipping, customs clearance and conformity checks. Use actual measured lead times and consider the lead time variability version of the safety stock formula.
- Branch networks. Weekly branch replenishment from a central warehouse in Riyadh, Jeddah or Dammam is a natural fit for order-up-to levels per branch.
Common Mistakes
- Ordering to S from physical stock only, ignoring open purchase orders, which doubles orders.
- Setting S once and never updating it. Demand changes; levels should be recalculated at least quarterly and before seasons.
- Covering lead time only. Safety stock in a periodic system must cover R + L, not just L.
- Skipping reviews. A missed review stretches the effective review period and causes stockouts.
- Using averages that include stockout periods. If you ran out for a week, recorded sales understate real demand for that week.
Running the Model in an ERP
The model depends on an accurate inventory position at every review, which means stock on hand, open purchase orders and customer commitments all in one place. In a spreadsheet this is the hardest part. In an ERP it is a report. Tranquil's inventory management runs alongside purchasing and sales in one cloud system, with central inventory across locations and multiple units of measure, so the figures needed for each review come from the same records as your orders and receipts.
Frequently Asked Questions
What is the difference between order-up-to level and maximum stock level?
They are often used interchangeably. Strictly, the order-up-to level is the target inventory position you order up to at each review, including stock on order. Physical stock on the shelf usually stays below it.
How often should the order-up-to level be recalculated?
At least quarterly for most items, monthly for fast movers, and always before major seasonal peaks such as Ramadan.
Is the order-up-to model better than a reorder point?
Neither is better in general. The order-up-to model suits scheduled ordering and combining items from one supplier. A reorder point reacts faster and needs less safety stock, which suits expensive or critical items.
What service level should I choose?
Start with 95% for most items, raise it for items where a stockout loses customers, and lower it for low-value items where some risk is acceptable.
Can Tranquil help with periodic replenishment?
Tranquil keeps stock, purchase orders and sales in one system, giving you the inventory position each review needs. Contact the Tranquil team to discuss your replenishment setup.