Syntha Lab RU

Seasonal buying: critical ratio and the optimal quantity under normal demand

Finds the buying quantity of a seasonal item that maximizes expected profit when demand is given by a mean and a spread: the critical ratio, the stock-out probability, expected leftovers and lost sales, the gain against your plan.

Forecast and budget For whom: Buyers and category managers before the first order of a seasonal category.

Calculate on your data →

At a glance

What question does it answer?
How much of a seasonal item to buy so that expected profit is highest when demand is uncertain, and what that gives against your plan.
What data is needed?
Price, cost and the leftover price at season end.
How to read the result?
Example on the default values: the critical ratio is 79.4%, the optimal buy 3,657 units against mean demand of 3,000; expected profit RUB 7.33 million, for the plan of 3,000 units RUB 7.01 million, a gain of RUB 310,328; the stock-out probability at the optimum is 20.6%.

Why and how to reason

A seasonal item is bought once, and demand is unknown. Order little and you lose sales; order much and goods remain that you sell below cost. The question is on which side the error is dearer. The cost of shortage Cu is the margin per unit: 4,900 − 2,200 = 2,700 (plus an extra loss on shortage, if any). The cost of excess Co is the loss on leftovers: 2,200 − 1,500 = 700. The optimum is where the probability of not reaching a shortage equals Cu ÷ (Cu + Co) = 2,700 ÷ 3,400 = 79.4%. This is the critical ratio. With mean demand of 3,000 and spread of 800 we need to cover 79.4% probability, that is z = 0.82 and a buy of 3,000 + 0.82 × 800 = 3,657. Your plan of 3,000 covers only 50%: the probability of shortage is 50%, and expected profit is lower by 310 thousand.

When to use it

It is meant for seasonal and fashion buying where a repeat order is impossible or too slow. Do not use it for continuing goods where top-up is possible: there the reorder point applies. Assumptions: demand is normal (with a large spread a negative part is possible), the price of leftovers is known, the error of the mean forecast is small. With strongly skewed demand use Monte Carlo. It works well as a first step, and the result is then tested with Monte Carlo on your shape of demand.

How to read the result

The key number is the critical ratio (79%). It says: erring on the excess side is cheaper than on the shortage side, so buy above the mean. The lower the price of leftovers, the lower the ratio and the more cautious the buy. Expected leftovers (749) and lost sales (93) show the price of the decision in units. The gain against your plan (310 thousand) is a guide, not a promise: it depends on assumptions about demand. Check whether you hit the budget or factory capacity, in which case the optimum must be constrained. It is worth writing the conclusion in one line: “we buy X because shortage is Y times dearer than excess”: the decision is then easier to defend.

Common mistakes

Typical trap: taking the price of leftovers above the real one, not entering a loss beyond margin on shortage (reputation, customers lost), understating demand spread, taking past demand as the mean, ignoring that nothing can be topped up in season. Take the spread from past seasons of similar models. Do not take the spread from weekly data if the buy is for the season: the spread of a season is different.

What data is needed?

  • Price, cost and the leftover price at season end.
  • Mean demand and its spread from past seasons.
  • Your planned buy for comparison.

Inputs

FieldUnitValue in the exampleMeaning
Selling price excl. VAT₽4,900The average selling price over the season. The example value is illustrative.
Unit cost₽2,200Full cost of a unit at the warehouse: goods, delivery, duties. The example value is illustrative.
Leftover price at season end₽1,500What leftovers really sell for after price reductions and clearance costs. If above cost, leftovers do no harm and the quantity is unbounded.
Loss beyond margin on a shortage₽0An extra loss per unit of unmet demand: a lost customer, a retailer fine. Zero if only the lost margin counts.
Mean demandunits3,000Expected sales over the season at full price and discounts. Footwear, accessories and children’s have their own values: take the category data.
Demand spread (std. deviation)units800An estimate from past seasons of this category; for a new category take wider. The example value is illustrative.
Your planned buyunits3,000Your planned buy in units: the optimal quantity is compared with it to show the gain in expected profit.

How it is calculated

StepFormula
Cu, CoCu = price − cost + shortage loss; Co = cost − leftover price
Critical ratioCR = Cu ÷ (Cu + Co)
Optimal quantityQ* = μ + z × σ, z = Φ⁻¹(CR)
Lost salesσ × (φ(z) − z × (1 − Φ(z))), z = (Q − μ) ÷ σ
Profitprice × sold + leftover price × leftover − cost × Q − loss × shortage

Worked numbers

Default values from the app; in the interactive version you enter your own.

Critical ratio
79.4 %Margin ÷ (margin + loss on a leftover): the probability with which demand should be covered.
Optimal buying quantity
3,657 units
Expected profit at the optimum
7.33 m ₽
Your plan: expected profit
7.01 m ₽
Optimum gain against the plan
310,328 ₽
Stock-out probability at the optimum
20.6 %
Expected leftover at the optimum
749 units
Expected lost sales at the optimum
93 units
Share of demand met at the optimum
96.9 %
Your plan: stock-out probability
50 %
Your plan: expected leftover
319 units
Expected profit by buying quantity
Line chart: Expected profit: 2.69m → 6.42m2m3m4m5m6m7m8m1,0002,0003,0004,0005,000optimumplanExpected profit, 1000: 2.69mExpected profit, 5400: 6.42m₽Buying quantity, units
Data as a table
Buying quantity, unitsExpected profit
1,0002.69m
1,1833.18m
1,3673.67m
1,5504.15m
1,7334.61m
1,9175.06m
2,1005.49m
2,2835.89m
2,4676.25m
2,6506.56m
2,8336.82m
3,0177.03m
3,2007.18m
3,3837.28m
3,5677.32m
3,7507.32m
3,9337.28m
4,1177.22m
4,3007.13m
4,4837.03m
4,6676.91m
4,8506.8m
5,0336.67m
5,2176.55m
5,4006.42m
At the optimum: sold, leftover, lost sales
Bar chart: Sold 2,907, Leftover 749, Lost 92.701,0002,0003,000SoldSoldUnits: 2,9072,907LeftoverLeftoverUnits: 749749LostLostUnits: 92.792.7units
Data as a table
MeasureUnits
Sold2,907
Leftover749
Lost92.7

Demand is your estimate: the wider the spread, the further the optimum departs from the mean. For footwear and children’s count by size and the size core, for accessories by article. Leftovers may be reordered or moved: the calculation does not allow for that.

Nuances and limits

  • Buying at mean demand: with a high margin the optimum is above the mean, with a low one below.
  • Taking the leftover price at full price and not at the real clearance price.
  • Estimating the spread from one season: an estimate from one season is unreliable.

Terms used

Calculate on your data →