Syntha Lab RU

Replenishment and stock cover

When and how much to reorder so bestsellers do not run out.

For retail Planning and buying

Open interactively →

About the topic

Replenishment answers the question “when and how much to reorder”. Behind it lies a simple conflict: if you order too late or too little, popular items run out and you lose sales; if you order too early or too much, cash settles in stock. The topic “Replenishment and stock” teaches you to find the balance between these two mistakes so that it is deliberate, not accidental. It is used by the buyer, the logistician, the operations director and the warehouse analyst.

The calculation rests on four quantities: average demand, its variability, lead time and service level. Average demand is how much you usually sell a week. Variability shows how much demand jumps from week to week: at the same average, an unstable item needs more stock. Lead time is how long it takes from order to goods appearing in the warehouse, and it may also vary. Service level is the probability of not running out within an order cycle: 95% means in 5 of 100 cycles you accept a shortage.

Safety stock protects against demand fluctuation during the lead time. Illustrative example: average demand 40 units a week, weekly variability 12 units, lead time 3 weeks, service level 95% (factor about 1.65). Variability over three weeks does not triple but grows by the square root of three: 12 x 1.73 = about 21. Safety stock = 1.65 x 21 = about 34 units. Reorder point = demand during lead time + safety stock = 40 x 3 + 34 = 154 units: when the balance including what is on order falls to 154, it is time to order.

There are three easy ways to go wrong. First: ignoring goods in transit, so the order is duplicated. Second: taking average demand over a period with promotions, which overstates the reorder point. Third: setting a high service level for all items: each extra bit of service costs more, so for popular items 95–98% may be justified, while for seasonal novelties and rare sizes 80–90% and fast reorder are better.

For seasonal novelties the logic is different: the first batch is ordered with room for a reorder, and after the first weeks of sales you decide how much to reorder while the factory can still deliver on time. Rare items (extreme sizes, expensive accessories) sell in bursts and ordinary formulas describe them badly; intermittent-demand methods suit them.

Where to start: take sales history without promotions, calculate average demand and variability, find the real lead time for each source, set service level by item class and compare the reorder point with the current balance.

Key ideas

Reorder point

The reorder point reacts most strongly to lead time. Demand 40 a week, variability 12, service 95%. A 3-week lead time gives 154, while 4 weeks gives 160 + 1.65 x 12 x 2 = about 160 + 40 = 200. One extra week raises the point by almost a third and cash in stock grows although demand did not change. So lead time is checked against actual deliveries, not against the contract.

Safety stock and service level

Safety stock grows non-linearly: moving from 90% to 99% service almost doubles it (factor 1.28 against 2.33). So service is set by the item’s value, not “99 everywhere”.

Lead time variability

If a factory delivers sometimes in 3 and sometimes in 5 weeks, safety stock must cover both demand variability and lead time variability. Ignoring it understates stock. Calculate lead time for each supplier separately.

Reorder from early sales

If 600 of 2,000 units of a novelty sell in two weeks (30%), the season forecast is built from that pace with an adjustment for the season curve. A reorder costs more than the first batch (urgency), so the decision is weighed against the leftover risk.

Intermittent demand

If an item sells once every 3–4 weeks in 1–2 units, an ordinary average misleads. Croston’s method estimates the size of a sale and the interval between sales separately. For such items the goal is not high service but minimal stock for one or two sales.

The rate of sale is cleaned of stock-outs and promotions

On illustrative data the base rate is 12.0 units a week, the promotion uplift 1.67; safety stock from the forecast error 279 units; the reserve of a new item 200 units.

Formula: Rate = sales ÷ in-stock days × 7

What to do: Take in-stock days from stock balances and count stock from the forecast error at a long lead time.

How to work through it

  1. Clean sales history of promotions and out-of-stock periods: otherwise demand is distorted.
  2. Calculate average demand and weekly variability for each item or group.
  3. Determine the real lead time and its variability for each source.
  4. Set service level by item class: runners, regular, rare, seasonal.
  5. Calculate safety stock and reorder point and compare with the balance including goods in transit.
  6. Determine the order size accounting for minimum quantities and volume discounts.
  7. For seasonal novelties provide a reserve for reorder and a decision date.
  8. Monthly, compare actual service (share of demand fulfilled) with the target.

Where to start

For a fast-selling item or group start with the reorder point and safety stock; for a seasonal novelty with a top-up from early sales; then check whether moving stock between stores pays back. Calculations of the topic in order: “Reorder point and safety stock”, “Reorder from early sales”, “Safety stock by fill rate and lead-time variability”, “Intermittent demand: Croston and SBA for rare sizes and accessories”, “Initial allocation of a new item without history”, “Safety stock from forecast error”, “Normalized rate of sale”. See also (their primary topic is another one): “Does moving stock between stores pay back”, “First order and reserve for reorders”, “Transfers between points: surplus, need and the gain against a discount”, “Replenishment by size: what and how much to send to stores by size”, “EOQ: economic order quantity with a minimum lot and volume discounts”.

Calculations in this topic

  • Reorder point and safety stock

    Calculates the stock level at which to reorder, how much safety stock to hold and how much to order now for an item or group.

    At what level to reorder, how much to hold as safety and how much to order now.

  • Reorder from early sales

    Builds a season forecast from the first weeks of sales and the cumulative curve of past seasons, estimates the reorder needed (low, central and high) and calculates the profit of the reorder lot including the rush surcharge and the minimum lot.

    How much to reorder after the first weeks of sales and whether it is worth it at the rush surcharge.

  • Safety stock by fill rate and lead-time variability

    Counts safety stock in two ways: by cycle service level (the probability of not running out) and by fill rate (what part of demand is met from the shelf), with demand and lead-time variability; shows reorder points and the money in stock.

    How much safety stock to hold to meet the required share of demand and how that differs from protecting every order cycle.

  • Intermittent demand: Croston and SBA for rare sizes and accessories

    From a series of sales with many zero periods estimates the expected demand per period by the Croston method or its SBA correction, sets the demand class (smooth, erratic, intermittent, lumpy), compares the accuracy with simple smoothing and shows the expected demand over the lead time.

    What demand to expect per week for an item that sells rarely and why that is better than ordinary smoothing.

  • Initial allocation of a new item without history

    Spreads the first batch of a new item across stores of three grades with a central reserve for top-ups: the grade weight, the minimum display, rounding by the largest remainder. Shows the allocation per store and the actual reserve.

    How much to give each store and how much to keep in the central reserve for top-ups.

  • Safety stock from forecast error

    Counts safety stock from the forecast accuracy rather than the demand scatter: the weekly error from WAPE, the accumulation of the error over the lead time and review with the correlation of weeks, the service level. Shows the bounds for independent and fully linked errors.

    What safety stock is needed to cover the forecast error over the lead time and the review period at the chosen service level.

  • Normalized rate of sale

    Cleans the rate of sale: divides weekly sales by in-stock days, excludes promotion weeks and weeks with few in-stock days, counts the base rate, the promotion uplift, the cover of stock and the XYZ stability class.

    What the base rate of sale of an item is, how much a promotion raises it and for how many weeks the stock will last.

  • Does moving stock between stores pay back

    Identifies donors and recipients by cover to the season end, the pieces to move, the gain per piece (with sale probabilities), the total result including fixed costs and the break-even transport cost.

    Whether moving surplus from one store to another pays after the season starts.

  • First order and reserve for reorders

    Calculates what share of the plan to order at once and how much capacity to keep for a reorder after the first sales. Compares the expected profit, leftover and lost sales of a two-stage order with a single large order. The logic is our own: demand is given by scenarios, early sales remove part of the uncertainty, a reorder costs more and is limited by capacity.

    What share of the plan to order at once and how much capacity to keep for a reorder after the first sales.

  • Transfers between points: surplus, need and the gain against a discount

    From the stock and selling speed of each point (a store or a warehouse) finds where there is a surplus above the required cover and where a shortage, picks transfers with lot multiples and compares the cost of moving with the discount that would have to be given to clear the surplus on the spot.

    Where to move surplus and whether moving pays better than a discount on the spot.

  • Replenishment by size: what and how much to send to stores by size

    From the stock, the goods in transit and the sales of each “store × size” cell calculates the target stock in weeks of cover with a minimum display, the need rounded up to the pack multiple and distributes the limited warehouse across sizes: cells with the lowest cover first. Shows how much to send and what the warehouse lacks.

    What and how much to send to stores by size from a limited warehouse.

  • EOQ: economic order quantity with a minimum lot and volume discounts

    Finds the order size at which ordering and holding costs together are lowest, tests volume price breaks, respects the minimum lot and the pack multiple, and shows annual costs and the saving against your size.

    In what lots to order so that ordering and holding together cost least, and whether a volume discount pays for a larger lot.

Self-check questions

  • Is demand cleaned of promotions and out-of-stock periods?
  • Is the lead time taken from actuals, not the contract?
  • Is lead time variability accounted for, not only the average?
  • Is the service level set by item class?
  • Does the balance compared with the reorder point include goods in transit?
  • Is an intermittent-demand method applied to rare items?
  • For marked goods, is time for codes included in the lead time?
  • Does the order size account for minimum quantities?
  • Is there a rule for who decides on a reorder and by which date?

Knowledge check

Multiple-choice questions. The correct answer and the explanation follow each question.

  1. What does the Croston method smooth separately when an item sells rarely?

    • A. The size of a sale and the interval between sales: the forecast is their ratio
    • B. Seasonality and trend
    • C. Demand of stores and demand of warehouses
    • D. The mean and the safety stock

    Correct answer: A. The size of a sale and the interval between sales: the forecast is their ratio. For rare sales, the size of one sale and how often it happens are estimated separately. With a size of 2 units and an interval of 4 weeks the forecast is 2 / 4 = 0.5 units a week. Seasonality and trend are smoothed separately by another method (Holt-Winters). Splitting by stores and warehouses is a level of detail, not a smoothing method. The method does not estimate safety stock: that is computed separately from demand variation and lead time.

  2. Demand is 50 units a week, variation 10, lead time 3 weeks, review period 1 week, service level 95% (z = 1.65). On hand 120 units and 60 in transit. How much should be ordered up to the target level?

    • A. 33 units
    • B. 113 units
    • C. 233 units
    • D. 53 units

    Correct answer: D. 53 units. Safety stock = 1.65 × 10 × √(3 + 1) = 33. Target level = 50 × (3 + 1) + 33 = 233. On hand plus on order is 120 + 60 = 180, so order 233 − 180 = 53 units. 113 comes from ignoring the 60 units in transit (233 − 120). 233 is the target level itself without subtracting stock. 33 is only the safety stock.

  3. Demand variation over the protection period is 40 units. How many extra units of safety stock does raising service from 95% (z = 1.65) to 99% (z = 2.33) require, and what does it cost at a unit cost of RUB 1,500?

    • A. 4 units, RUB 6,000
    • B. 93 units, RUB 139,800
    • C. 27.2 units, RUB 40,800
    • D. 66 units, RUB 99,000

    Correct answer: C. 27.2 units, RUB 40,800. Safety stock = z × variation: at 95% it is 1.65 × 40 = 66 units, at 99% it is 2.33 × 40 = 93.2 units. The increase is 27.2 units, 27.2 × 1,500 = RUB 40,800. 4 units and RUB 6,000 come from reading 4 points of service as 4 units. 93 units and RUB 139,800 are the whole stock at 99%, not the increase. 66 units and RUB 99,000 are the stock already held at 95%.

  4. Annual demand is 1,800 units, placing one order costs RUB 2,000, holding one unit costs RUB 20 a year. What is the economic order quantity (EOQ)?

    • A. 424 units
    • B. 600 units
    • C. 300 units
    • D. 1,200 units

    Correct answer: B. 600 units. EOQ = √(2 × demand × cost per order / holding) = √(2 × 1,800 × 2,000 / 20) = √360,000 = 600 units, that is three orders a year. 424 comes from losing the factor of 2 under the root (√180,000). 300 is the average stock with an order of 600 (half a batch), not the order size. 1,200 comes from doubling 600 instead of taking the root.

  5. A rush reorder: cost with the surcharge is RUB 1,800, the selling price is RUB 3,000, unsold leftovers go for RUB 900. What share of the reordered batch must sell at full price for the reorder to pay back?

    • A. 42.9%
    • B. 60%
    • C. 30%
    • D. 50%

    Correct answer: A. 42.9%. At break-even share p: p × 3,000 + (1 − p) × 900 = 1,800, so p = (1,800 − 900) / (3,000 − 900) = 900 / 2,100 ≈ 42.9%. 60% is 1,800 / 3,000: the leftover sold at 900 is not accounted for. 30% is 900 / 3,000, the ratio of the clearance price to the price. 50% is “half” without a calculation: in fact the threshold depends on three prices.

  6. Which weeks are excluded from the base rate?

    • A. All even ones
    • B. The first two
    • C. None
    • D. With a promotion and with few in-stock days

    Correct answer: D. With a promotion and with few in-stock days. They distort the rate up and down. Covered in: The rate of sale is cleaned of stock-outs and promotions.

  7. What is safety stock counted from at a long lead time?

    • A. The price of the item
    • B. The forecast error over the lead time
    • C. The number of stores
    • D. The warehouse area

    Correct answer: B. The forecast error over the lead time. The error accumulates over the term with the correlation of weeks. Covered in: The rate of sale is cleaned of stock-outs and promotions.

Common mistakes

  • Mistake: One high service level for all

    What it costs: Cash goes into low-value stock

    How to avoid: Service by class.

  • Mistake: Ignoring goods in transit

    What it costs: The order is duplicated and overload appears

    How to avoid: Count the balance including what is on order.

  • Mistake: Taking demand with promotions

    What it costs: Reorder point overstated, stock grows

    How to avoid: Clean the history.

  • Mistake: One lead time for all

    What it costs: Shortages at unreliable suppliers

    How to avoid: Lead time and variability by source.

  • Mistake: Applying the formula to rare items

    What it costs: A zero or absurd result

    How to avoid: Intermittent-demand methods.

  • Spreading a new item entirely at once

    What it looks like (illustrative figures): Nothing left to top up by sales.

Russia and the EAEU

Lead times in Russia depend strongly on the source: for a Russian factory they are usually weeks; for imports from China and other countries, transport, customs clearance and certification are added to production, and the total is often measured in months. Calculate the reorder point separately for each source and keep lead time variability explicit: for imports it is noticeably higher.

For goods subject to marking in the Chestny ZNAK system (clothing, footwear and other categories under Government Resolution No. 1956), add time for ordering and receiving codes and for putting goods into circulation: without a code the goods cannot be sold. Electronic document exchange (UPD through an EDI operator) speeds up receiving, but an error in code data stops acceptance.

For the price of stock use cost without VAT; the 22% VAT rate from 01.01.2026 affects payment to the supplier but not the reorder point. If you sell through a marketplace using the platform’s warehouse, the replenishment time includes delivery to the platform’s warehouse and acceptance: take actual lead times from your own experience, because acceptance rules and tariffs change.

The logic is the same; lead times and documents differ. Within the EAEU duties on goods in free circulation are not charged and VAT is paid at the country of destination with an import declaration (Treaty on the EAEU, Annex 18), so the lead time from Belarus, Kazakhstan, Armenia or Kyrgyzstan consists of production, transport and documents, not of a customs queue at the border. As a guide, check: the time is usually shorter than imports from third countries, but variability remains. For supplies to Kazakhstan mandatory marking is being introduced in stages from 01.12.2026 (medium reliability): include time for codes in the reorder point.

Categories and audiences

Apparel
In apparel, popular basics (T-shirts, jeans) are replenished by formula, while seasonal trends are ordered in a small run and reordered after the first weeks. Delivery time and the seasonal window decide whether the reorder is in time: if lead time is 6 weeks and the window 8, replenishing is too late.
Footwear
In footwear replenishment is by size, not by style: the order is calculated for each size, and safety stock for extreme sizes is almost zero. Demand per size is intermittent, so Croston’s method is more useful than a simple average. The minimum quantity is often a carton.
Accessories
Accessories sell in small quantities and rarely, many items have intermittent demand. Replenishment works well: shorter lead time, small batches, and a lost sale due to absence costs more because the shopper often does not wait. Gifting peaks require stock in advance, before the shortage.
Kids
In children’s wear replenishment is limited by the season window: a reorder after 1 September is pointless, the child is already dressed. High service is set for basic school items and held cautiously for the rest. Sizes do not carry over, so safety stock at the extremes is minimal.
Men
The men’s basics (shirts, jeans, socks) are convenient to hand to automatic replenishment: the style sells for years and a formula with history works. But for trousers and jeans replenishment goes by “size x length” cells, and the reorder point is calculated for each cell, not for the SKU: in total the stock may look sufficient while the fast-moving cells are empty. The risk remains in outerwear, where the window is short.
Women
For women’s novelties the reorder point is set per color-style, not per style, and the formula helps little here: fabric for a reorder is ordered in advance and the window is short. The reorder decision is taken from sales of the first two or three weeks. The safety stock formula is left for basics (lingerie, basic knitwear) with service of about 95%, and for a novelty you calculate how much can be reordered while the factory can still deliver.

Questions this topic answers

  • At what stock level to reorder?
  • How much safety stock to hold?
  • How much to order now?
  • What service level is justified for fast sellers and for seasonal items?

Key measures and terms

Study first

What next

Next step: “Working with stock: transfers, top-up, defects, prices, order entry”. Related: “Assortment and sizes”, “Buying budget and cash”, “Size runs, stores and range in season”.

Documents on this topic

Open interactively →