Retail Planby RetailNorthstar

How to calculate weeks of supply

Weeks of supply is the number of weeks the current inventory would last at the expected rate of sale: on-hand units or dollars divided by average weekly sales, trailing against recent history or forward against the sales plan. 2,400 units selling 300 a week is 8.0 weeks of supply. The same figure is called weeks of cover, and its forward version forward cover. The division is the easy part; which weeks go into the denominator, which basis both sides are counted in, and whether on-order stock is in the numerator decide what the answer means.

Run your own numbers in the weeks of supply calculator; the weekly grid the figure lives on is explained in what is a WSSI and read week by week in how to read a WSSI; the receipt budget it feeds is open to buy.

All worked figures are illustrative — not benchmarks, and not drawn from any brand.

The short version
Divide closing on-hand stock by the average weekly sales for the same grain and the same basis. Use the forward plan as the denominator wherever a plan exists and keep the trailing figure beside it as the check on the forecast. Keep on-order stock out of on hand and report it as a separate projected figure. Read a replenished item against a floor set by lead time, review period and safety stock, and a seasonal item against a ceiling set by the trading weeks left plus the exit stock. Compute it by size for anything sold in a size run.
Definition — Weeks of supply (WOS)
Weeks of supply converts a stock position into time: how many weeks of selling the stock on hand represents at a stated weekly rate. It is read against a floor for replenished items and a ceiling for seasonal ones, and a figure without its denominator, its basis and its grain is a number without a meaning.
Weeks of supply = on-hand inventory ÷ average weekly sales (trailing actuals or forward plan); projected = (on hand + on order) ÷ average weekly sales
Used by: Merchandise planners, allocators, replenishment and trading teams — weekly in-season, and pre-season as the cover target that sets closing stock
Related: Weeks of cover, forward cover, stock-to-sales ratio, inventory turnover, sell-through, WSSI, open-to-buy

Already know the method and just need the number? The weeks of supply calculator takes current inventory, average weekly sales and optional incoming receipts and returns weeks of supply with and without the receipts.

Use the Weeks of Supply Calculator →

One numerator, two denominators, three bases

The numerator is stock on hand at the end of a week, for one grain — a SKU in a door, a style in a channel, a category across the business. The denominator is a weekly rate of sale, and there are two honest ways to get one:

The result is in weeks whatever the basis, and the basis is the third decision: units over units a week, retail dollars over retail sales a week, or cost dollars over cost of sales a week, never one over another. Four weeks is the usual window for n, long enough to smooth a promotional week and short enough to follow a trend; six or eight suit slower, steadier items, and the weeks left in the selling window cap the forward n near the exit.

Weeks of supply is the reciprocal view of a rate. Where sell-through asks what share of the stock sold, inventory turn how many times it was replaced in a year and the stock-to-sales ratio how heavily a period opened against its own demand, weeks of supply asks the trading question: at this rate, when does this run out? The calculator runs the headline form with an optional receipts line for the projected figure.

Forward against trailing, and why planners read the forward figure

A trailing figure answers a historical question — how long would this stock last if the next weeks sell like the last ones — and that is the right question only when the next weeks will. For a replenished basic in mid-season it usually is, and the trailing figure is cheap: it needs no forecast, only the sales ledger. For anything with a season curve it is the wrong question, because the next four weeks are the one thing a seasonal item is guaranteed not to share with the last four. Going into a peak the trailing rate is below the rate ahead, so cover reads long and comfortable while the stock is about to be swallowed; coming out of a peak the trailing rate is above the rate ahead, so cover reads short while the stock is about to become an overstock. The error has the wrong sign at both ends of the season, which is why the forward cover line on a WSSI divides by the planned weeks ahead and not the actual weeks behind.

Forward cover has its own weakness, and it is the mirror image: it is only as honest as the forecast under it. When actuals run behind plan and nobody reforecasts, forward cover keeps reading to plan while stock builds, and the trailing figure — which has started to rise — is the only line on the report telling the truth. The practical answer is not to choose but to carry both and read the gap between them as a measurement of the forecast: trailing above forward says the plan expects acceleration that history has not yet shown; forward above trailing says the plan expects a slowdown. Either can be right. A gap that widens week after week without a reforecast is the one that cannot be.

Which figure is the operating one follows from the lifecycle. A replenished program with a weekly review reads trailing cover against its floor, because the rate is stable and the floor was set from it; a seasonal item reads forward cover against the weeks left, because the rate is the plan’s curve; a clearance item reads forward cover against the exit date at the markdown rate of sale, which is a forecast by definition. The reforecast is the step that keeps the forward denominator honest, and the second worked example below shows how far the two figures separate on one stock line when nothing is wrong except the shape of the demand.

Units, retail dollars or cost dollars — and never mixed

A single style is counted in units, because units are what sit on the shelf, what the size run is made of and what a customer buys. A category or a department is counted in dollars, because its styles carry different prices and 1,000 units of belts and 1,000 units of coats are not the same stock position. Dollars come in two flavours. Retail dollars value stock at the ticket and sales at what rang through the till, which ties cover to the merchandise plan, the WSSI and the open-to-buy, all of which are usually held at retail. Cost dollars value stock at what it cost to own and sales at cost of goods sold, which answers the working-capital question — how many weeks of cash are sitting on the shelf — and is the basis GMROI and inventory turn share.

While the cost complement is constant, the three bases give the same answer: stock at cost is stock at retail times the complement, cost of sales is sales times the complement, and the complement cancels in the division. The third worked example shows an outerwear category reading 6.00 weeks at retail and 6.00 at cost at the open, and 5.78 on both at the week-four close. The bases part the moment something moves one side without the other, and markdowns are that thing: a markdown reduces the retail value of the stock still on hand without changing its cost or its unit count, so from the first reduction retail cover reads lower than cost cover and lower than unit cover. In the example, week six reads 3.71 at retail against 4.29 at cost. Neither is wrong; they describe different quantities, and the unit figure for the style underneath is the one that says how many weeks of garments are left.

Mixing the bases produces a number that is wrong by the complement and looks like nothing in particular. Stock at cost over sales at retail on the same category reads 2.70 weeks; stock at retail over cost of sales reads 13.33. Both come from figures a planning report already carries side by side, which is why the basis is written on the report rather than assumed, and why a department figure built from style figures has to be summed in one basis before dividing — the pitfalls section works through what happens when it is averaged instead. The retail inventory method is how a pool of stock held at retail is converted to cost when its styles carry different markups.

By week, by period, by day

The unit of the answer is a week because the unit of the trading cycle is a week: sales are read Monday, the WSSI closes on a fixed day, and a reorder placed this week lands a known number of weeks out. A figure in days is the same arithmetic multiplied by seven — 7.0 weeks is 49 days — and belongs where lead times are quoted in days, in a distribution centre or on a replenishment system set to daily review. A figure in months or periods is a different measure wearing similar clothes: a stock-to-sales ratio of 2.0 is eight weeks of cover on a four-week period and ten on a five-week period, and a 4-5-4 calendar makes every third period a different length from the two before it.

Weekly cover and period cover also describe different moments. Weeks of supply is a point-in-time figure — the closing position of one week over a rate — while the stock-to-sales ratio is a period figure, the opening position of a period over the whole period’s sales, and inventory turn is a period average over a year. The three agree on an item whose stock is flat and whose sales are steady, and disagree on everything else, because a point and an average of points are not the same number on a build or a sell-down. The GMROI guide works through why a two-point average misreads a seasonal build; the same reasoning is why a weekly cover line on a WSSI and a monthly stock-to-sales ratio on a merchandise plan can both be right and still differ.

The practical rule: compute weeks of supply from the weekly ledger on the week the report closes, convert to days only at the point where the decision is in days, and convert a period ratio to weeks only after checking how many weeks the period holds. When a report mixes a cover figure in weeks for one category with a period ratio for another, the two are compared only after both are expressed as weeks of forward sales.

Weeks of supply against weeks of cover, stock-to-sales, inventory turn and sell-through

Five measures put stock against sales, and each converts to the others under stated conditions. The identities are useful as checks — a report whose turns and cover do not reconcile has computed them on different bases or different windows — and as translations between teams that speak different ones: a finance team in turns, a merchandising team in cover, a planning team in stock-to-sales.

Formula
Trailing weeks of supply
Closing stock ÷ average of recent weekly sales
Forward weeks of cover
Closing stock ÷ average of planned weekly sales ahead
Stock-to-sales ratio
Opening stock ÷ the same period’s sales
Inventory turnover
Cost of goods sold ÷ average inventory at cost
Sell-through
Units sold ÷ (units sold + units on hand)
Basis
Trailing weeks of supply
Units, retail or cost — matched
Forward weeks of cover
Units or retail, matched
Stock-to-sales ratio
Retail, usually
Inventory turnover
Cost
Sell-through
Units
Window
Trailing weeks of supply
A week’s close, over recent weeks
Forward weeks of cover
A week’s close, over weeks ahead
Stock-to-sales ratio
A period (month or 4-5-4 period)
Inventory turnover
A year, on an average
Sell-through
A week, or season to date
Identity to weeks of supply
Trailing weeks of supply
Itself
Forward weeks of cover
Same numerator; differs in the denominator
Stock-to-sales ratio
Ratio × weeks in the period: 2.0 on a four-week period = 8.0
Inventory turnover
52 ÷ turns, on average inventory: 6.5 turns = 8.0 weeks
Sell-through
(1 − sell-through) ÷ sell-through for one week: 20% = 4.0 weeks
What it misses
Trailing weeks of supply
Where demand is heading
Forward weeks of cover
Whether the forecast is right
Stock-to-sales ratio
How long the stock lasts
Inventory turnover
Timing inside the year
Sell-through
Price, cost and margin
Main decision
Trailing weeks of supply
Replenishment floor, reorder timing
Forward weeks of cover
Chase, markdown, overstock week
Stock-to-sales ratio
Pre-season stock targets by period
Inventory turnover
Working-capital budget, GMROI
Sell-through
Chase, hold, markdown

Two of the identities carry a condition. 52 ÷ turns gives the cover the average inventory represented over the year, which equals the current figure only if stock is flat; the inventory turnover calculator returns the turn half. The sell-through identity holds for a single week’s sell-through at that week’s rate — a week that sells 100 of 500 available is 20% sell-through and 400 ÷ 100 = 4.0 weeks of supply — and not for a season-to-date figure, which has receipts inside it. The stock-to-sales ratio calculator and the sell-through calculator run those two; the sell-through guide covers the base that figure divides by. Where two of the five disagree after conversion, the sales rate is changing, which is the one thing a single figure cannot show.

On hand against on hand plus on order: the projected figure

On hand is what can be sold this week. On order is what has been bought and not yet received — a purchase order raised, cartons at the vendor, a container on the water, a trailer in the yard. The two answer different questions and the error is folding them together. Weeks of supply on hand says when the shelf empties if nothing lands; projected weeks of supply says what the position becomes if everything does, and a reorder decision needs both: the first to know how urgent the gap is, the second to know whether an order already placed closes it. The illustrative basic reads 7.00 weeks on hand at the week-one close, 2,100 units over 300 a week, and 13.00 projected with two deliveries of 900 still to land, 3,900 over 300.

The projected figure is only as good as two things behind it: the delivery dates, and the horizon. An order due in week three belongs in a four-week projection; an order due in week nine does not, because by week nine the stock on hand today is gone and the figure is describing a different position. The lead-time guide sets out how variable a delivery date is in practice and how that variability becomes a buffer, and a projected cover built on the vendor’s promised date rather than the vendor’s demonstrated date inherits every slip. Where lead times are long — home and furniture on ocean freight, imported basics on a twelve-week cycle — the projected figure is the one the reorder is placed on, because the on-hand figure will always say reorder too late; where lead times are short, the on-hand figure does most of the work.

The same split applies to the WSSI’s intake row. Forward cover on the WSSI is normally computed on closing stock after the week’s intake has landed, so a receipt phased into week four lifts cover from week four onward and not before. That is the projected logic made weekly: each future week’s cover already assumes the intake planned into it. Reading the cover line forward from the current week is therefore reading projected cover, and a receipt that slips a week shows as a dip in cover in the week it was due. The WSSI calculator runs the chain, and the intake term covers what counts as a receipt and when.

Which weeks go into the rate, and which stay out

The denominator is an average, and an average is a choice about which weeks count. The window comes first. Four weeks is the working default because it is long enough to absorb one odd week and short enough to follow a trend; a basic with steady demand can take six or eight, which smooths the figure and slows its response; a fast-moving fashion item in a short window may need two or three, because by the fourth week back the item was in a different phase of its life. The test is whether the weeks in the window describe the weeks ahead — the same test that separates trailing from forward, applied inside the trailing figure. A longer window is not a more accurate rate; it is a rate for a longer-lived item.

Then the exclusions. An out-of-stock week sells nothing and says nothing about demand, so it leaves the trailing window, or the rate records how little an empty shelf sold and the cover reads long for the item that most needs it to read short — the pitfalls section quantifies this. A promotional week sells above the rate for a reason that will not repeat, and either leaves the window or is normalised to the rate the promotion lifted; keeping it in makes cover read short for a fortnight after the event and the chase that follows lands into the post-promotion trough. A partial week — a launch midweek, a store closed for a refit — is annualised to a full week or dropped. And a week before a markdown is a week at a different price: for an item now trading at a reduced ticket, the pre-markdown weeks describe a slower rate than the one it is selling at, and the forward rate at the new price is the honest divisor.

Finally the item with no history. A new style in its first week has one trailing data point, which is why the second worked example reads 11.50 weeks in week one — 4,600 units over a single week of 400 — and why the figure means nothing. Until three or four weeks of actuals exist, the denominator is the plan, or a like-for-like style’s curve from last season scaled to this season’s buy, and the first weeks of actuals are read against it as a forecast-accuracy question before they are allowed into the rate. The comp-sales convention — like for like, same weeks, same doors — is the discipline that keeps a trailing rate honest when the store count or the channel mix changed inside the window.

Seven steps, in order

Step five is the division. Steps one to four decide what gets divided, and steps six and seven decide what the answer is for; a correct division of the wrong stock over the wrong weeks is the usual way a cover report misleads.

  1. 1

    Fix the grain, the basis and the week

    Decide what the figure describes — one SKU in one door, a style across a channel, a category across the business — and whether it is counted in units, retail dollars or cost dollars. Numerator and denominator share all three. Use the same week definition the sales ledger uses.

  2. 2

    Take on-hand stock at the end of the week

    Closing stock for the grain, available to sell: owned, in the location, not reserved, not damaged. On-order and in-transit units are kept separate for the projected figure in step six, never folded into on hand.

  3. 3

    Pick the denominator: trailing or forward

    Trailing: the average of the last four (or six, or eight) weeks of actual sales. Forward: the average of the next four weeks of the sales plan or forecast, or of the weeks left in the window where fewer remain. Write which one the report uses on the report.

  4. 4

    Compute the average weekly rate

    Sum the weeks in the window and divide by the number of weeks. Exclude a week in which the item was out of stock from a trailing average, or the average records how little an empty shelf sold rather than how fast the item sells.

  5. 5

    Divide stock by the weekly rate

    Weeks of supply = closing stock ÷ average weekly sales. 2,400 units selling 300 a week is 8.0 weeks. Multiply by seven for days of supply.

  6. 6

    Add on order for the projected figure

    Projected weeks of supply = (on hand + on order due inside the horizon) ÷ average weekly sales. Report it beside the on-hand figure, not instead of it: the first says what can be sold now, the second what the position becomes if every delivery lands.

  7. 7

    Read it against the floor or the ceiling for its lifecycle

    A replenished item is read against a floor set by lead time, review period and safety stock. A seasonal item is read against a ceiling set by the trading weeks left plus the exit stock the plan intends to carry out. Then act on the row that caused the gap: intake, allocation or price.

A replenished basic over six weeks, in units

A core tee in one colour, one channel, counted in units. It opens with 2,400 on hand; the last four weeks sold 310, 290, 305 and 295, which sum to 1,200 and average 300; the forward plan is flat at 300 a week. It is on a three-week lead time, reviewed weekly, with one week of safety stock, so its cover floor is 3 + 1 + 1 = 5 weeks, and it is replenished in orders of 900 landing in weeks three and six. Actual sales over the six weeks are 300, 310, 290, 300, 320 and 280, which sum to 1,800 — exactly the 1,800 received — so the position closes where it opened.

Hypothetical six-week run for one replenished basic in units: opening stock, receipts, sales, closing stock, the trailing four-week average rate, and weeks of supply on the trailing average and on the flat forward plan.
LineW1W2W3W4W5W6Six weeks
Opening stock2,4002,1001,7902,4002,1001,780—
Receipts00900009001,800
Sales3003102903003202801,800
Closing stock2,1001,7902,4002,1001,7802,400—
Trailing four-week average297.5302.5298.75300.0305.0297.5—
Trailing weeks of supply7.065.928.037.005.848.07—
Forward weeks of supply (plan 300)7.005.978.007.005.938.00—

Week one closes at 2,100. The trailing window is now 290, 305, 295 and 300 — the oldest week has dropped off and the new one has joined — which sums to 1,190 and averages 297.5, so trailing weeks of supply is 2,100 ÷ 297.5 = 7.06; forward is 2,100 ÷ 300 = 7.00. Week two closes at 1,790 over a trailing average of 302.5: 5.92 trailing, 5.97 forward. The first 900 lands in week three and the position closes at 1,790 + 900 − 290 = 2,400, 8.03 weeks trailing and 8.00 forward. On-hand cover oscillates between 5.84 and 8.07 weeks, and the position never approaches the five-week floor, which is what a replenishment program on a three-week order cycle should look like. The floor is tested on the position — on hand plus on order — not on the on-hand line: the order for week six is placed at the week-three close, when on hand reads 8.00 and on order is 0, so the position is 8.00, above the floor; on hand then troughs at 1,780 in week five — 5.84 trailing, 5.93 forward — with 900 still to land, so the position is 2,680, or 8.93 weeks on the plan rate. The on-hand line is read for the stock-out week, the week it would run out before a delivery; the position is read against the floor.

The two denominators never read more than 0.10 weeks apart — 5.84 against 5.93 in week five is the widest gap — because demand is flat and the plan is right; this is the case in which the trailing figure is a perfectly good operating number, and the cheap one. The projected figure tells the other half of the story: at the week-one close, 2,100 on hand plus 1,800 on order over 300 is 13.00 weeks, and the gap between 7.00 and 13.00 is two deliveries that have to arrive on their dates. Had the week-three delivery slipped a week, week three would have closed at 1,500 on hand — 5.00 weeks — while the position still read 8.00 because the 900 was owed, and a second slip would have taken week four to 1,200 on hand, 4.00 weeks: that is the on-hand line doing its job, saying how long the stock lasts if the delivery keeps slipping. The replenishment guide turns this floor into a reorder point and a pack-rounded order quantity; the replenishment term covers the vocabulary.

A seasonal build, where forward and trailing diverge

A fashion dress with a ten-week selling window and no replenishment after the second drop. The sales plan is 400, 600, 800, 1,000, 1,100, 1,000, 900, 700, 500 and 300 units — 7,300 in total — with a launch drop of 5,000 in week one and a second drop of 2,600 in week four, 7,600 received, so the plan exits with 300 units for the outlet. To isolate the effect of the denominator, actual sales are held exactly to plan: every figure below comes from one stock line, and the only thing that differs between the two cover rows is which weeks are in the divisor. The trailing average uses up to four weeks of actuals (one in week one, two in week two, and so on); the forward average uses the next four weeks of plan, or the weeks left where fewer remain.

Hypothetical ten-week seasonal window for one dress in units with actuals held to plan: receipts, sales, closing stock, the trailing average and trailing weeks of supply, the forward average and forward weeks of supply, the trading weeks left and the planned exit stock expressed in forward weeks.
LineW1W2W3W4W5W6W7W8W9W10Window
Receipts5,000002,6000000007,600
Sales (plan, held as actual)4006008001,0001,1001,0009007005003007,300
Closing stock4,6004,0003,2004,8003,7002,7001,8001,100600300—
Trailing average (up to four weeks)4005006007008759751,000925775600—
Trailing weeks of supply11.508.005.336.864.232.771.801.190.770.50—
Forward average (next four weeks or fewer)8759751,000925775600500400300——
Forward weeks of supply5.264.103.205.194.774.503.602.752.00exit—
Trading weeks left9876543210—
Exit stock in forward weeks0.340.310.300.320.390.500.600.751.00——

Read week three. The dress closes at 3,200 with 800 sold. The trailing average is (400 + 600 + 800) ÷ 3 = 600, so trailing weeks of supply reads 5.33; the forward average is (1,000 + 1,100 + 1,000 + 900) ÷ 4 = 1,000, so forward cover reads 3.20. Two weeks apart on the same 3,200 units, and the forward figure is the one that matters: the second drop of 2,600 is due in week four, and if it slipped two weeks the dress would sell 1,000 in week four and 1,100 in week five and close week five at 1,100 units — 1.42 weeks of forward cover (1,100 ÷ 775) with the drop still not landed. The forward read of 3.20 at the week-three close said so; the trailing read of 5.33 said there were two more weeks than there were. A planner reading the trailing line would not chase the delivery; a planner reading the forward line would.

Now read week six. The dress closes at 2,700 after the peak. The trailing average is (800 + 1,000 + 1,100 + 1,000) ÷ 4 = 975, so trailing reads 2.77 weeks — a stock-out signal, on its face. The forward average is (900 + 700 + 500 + 300) ÷ 4 = 600, so forward reads 4.50 against four trading weeks left. The planned exit of 300 units is 300 ÷ 600 = 0.50 forward weeks, and four weeks left plus 0.50 is exactly 4.50: the dress is precisely on plan, carrying out the stock it intended to. The trailing figure has manufactured a shortage out of a sell-down, and a chase placed on it would land units the window cannot sell. From week six to week nine the forward figure equals the trading weeks left plus the exit allowance every week — 3.60, 2.75, 2.00 — which is what a plan trading exactly to itself looks like on a cover line, and the test for an overstock week is forward cover above that sum, as the WSSI walkthrough sets out.

Three things follow. First, the trailing figure is not wrong about the past; it is wrong about the future, and on a seasonal item the future is the question. Second, the forward figure here is exact because the actuals were held to plan — in a live season the gap between the two rows is the reforecast signal, and the reforecast guide is where it goes. Third, the receipt timing drives everything: the receipt-flow guide and the seasonal safety-stock guide cover how the second drop’s week is chosen and how much buffer sits around it.

In dollars: an outerwear category at retail and at cost

An outerwear category across one channel, in thousands of dollars, over an eight-week window. It opens with 420 of stock at retail. The sales plan is 60, 70, 80, 70, 60, 50, 40 and 30 — 460 in total — with one receipt of 120 at retail in week three and markdowns of 20 in week six and 15 in week seven, 35 in total, so the chain closes at 420 + 120 − 460 − 35 = 45 of exit stock at retail. The category runs a 0.45 cost complement on its opening stock and its receipt — a 55% initial markup — so stock at cost opens at 420 × 0.45 = 189, the receipt is 54 at cost, and each week’s cost of sales is 0.45 × its retail sales. Markdowns reduce the retail value of the stock and leave its cost alone. Forward cover uses the next four weeks of plan, or the weeks left.

Hypothetical eight-week window for one outerwear category in thousands of dollars: sales, receipts and markdowns at retail, closing stock at retail with the forward average and forward weeks of supply at retail, closing stock at cost with the forward average and forward weeks of supply at cost, and the trading weeks left plus the exit allowance each week.
LineOpenW1W2W3W4W5W6W7W8Window
Sales at retail—6070807060504030460
Receipts at retail—0012000000120
Markdowns—000002015035
Closing stock at retail4203602903302602001307545—
Forward average at retail7070655545403530——
Forward weeks of supply at retail6.005.144.466.005.785.003.712.50exit—
Closing stock at cost (0.45)189162130.5148.51179067.549.536—
Forward average at cost31.531.529.2524.7520.2518.015.7513.5——
Forward weeks of supply at cost6.005.144.466.005.785.004.293.67exit—
Trading weeks left plus exit allowance—7.646.695.825.004.133.292.50——

At the open, 420 at retail over a forward average of (60 + 70 + 80 + 70) ÷ 4 = 70 is 6.00 weeks; 189 at cost over 0.45 × 70 = 31.5 is also 6.00. The two bases agree to the second decimal for five weeks — 5.14, 4.46, 6.00, 5.78 and 5.00 on both lines — because the complement is the same on every dollar of stock and every dollar of sales, and cancels. Mix them and the figure is wrong by that complement: 189 over 70 reads 2.70 weeks, and 420 over 31.5 reads 13.33. Both look like cover figures; neither is one.

Week six is where the bases part. The category sells 50 and marks down 20, closing at 200 − 50 − 20 = 130 at retail; at cost it closes at 90 − 22.5 = 67.5, because the markdown took retail value off the stock and left its cost where it was. Forward cover is 130 ÷ 35 = 3.71 at retail and 67.5 ÷ 15.75 = 4.29 at cost, and the complement on the stock still on hand has risen from 0.45 to 67.5 ÷ 130 = 0.52. Both figures are right about what they measure: there are 4.29 weeks of cost sitting on the shelf, and the retail value of it covers 3.71 weeks of planned sales at plan prices. The lesson is not that one basis is correct but that after a markdown the retail figure understates how many weeks of garments remain, and the unit figure for each style underneath is the one that answers that question.

The last row shows what the retail cover line is for. From week four the forward window is the whole of the window left, and trading weeks left plus the exit allowance (45 over the forward average) is 5.00, 4.13, 3.29 and 2.50. Forward cover at retail sits above that line by 0.78, 0.88 and 0.43 weeks in weeks four to six and exactly on it in week seven — and those excess weeks, multiplied back by the forward average, are 35, 35 and 15 at retail: the markdowns still to be taken. The arithmetic is the identity closing stock − (sales left + exit) = markdowns left, read in weeks. A cover line above the weeks-left-plus-exit line is a markdown the plan has already written, whether or not anyone has written it down, which is the subject of the markdown plan guide and the markdown calculator.

Size-level weeks of supply and the broken-run trap

A style sold in a size run is five or more stock positions reported as one, and the one figure is the stock-weighted average of the five. That average hides the only two things that matter about a size run: which size runs out first, and which sizes will still be there at the exit. The illustrative style below holds 1,200 units selling 250 a week — 4.8 weeks of supply at the style level — on a four-week lead time.

Hypothetical style in five sizes: units on hand, weekly sales and weeks of supply by size, with the style total.
SizeOn handWeekly salesWeeks of supply
XS80108.0
S250505.0
M3801103.5
L300605.0
XL190209.5
Style total1,2002504.8

Size M holds 380 and sells 110 a week: 3.5 weeks, below the four-week lead time, so a reorder placed today lands after M is gone. XS and XL hold 80 and 190 selling 10 and 20: 8.0 and 9.5 weeks, roughly double the style figure, and the units that will be on the clearance rail. S and L sit at 5.0. The style reads 4.8 weeks — enough to reorder with time in hand — and that figure describes none of its sizes: it is a shortage in M and an overstock in the tail averaged into a number that looks fine. Once M is out the style stops presenting as a full run, the rate of sale on the remaining sizes falls, and their cover lengthens further, which is the broken-run trap: the weeks of supply on the leftover sizes rises because the size that drove the traffic is missing.

The remedy is in the computation and in the buy. Compute cover by size for any item sold in a run, read the reorder against the fastest core size and the markdown exposure against the slowest tail sizes, and treat the style figure as a total rather than a position. In the buy, the size curve that set 380 of M against 190 of XL is what the size-level cover is now testing, and the minimum buy depth guide works through how deep a run has to be before the tail sizes can be bought at all without being bought into markdown. The same logic applies to shade ranges in beauty, half sizes and widths in footwear, and any dimension across which one SKU number is reported as one.

Weeks of supply by channel and by location

The grain decision in step one has a second dimension: where the stock is. A business figure pools every unit against every sale; a channel figure splits stores from ecommerce from wholesale; a location figure reads one door or one distribution centre. Each is a true division and they answer different questions, because stock in one place cannot sell in another until someone moves it. A category can read 6.0 weeks across the business while door A holds 2.0 and door B 12.0, and the business figure says nothing is wrong. Where stock can be moved — a DC that replenishes doors, an ecommerce pool that ships from stores — the pooled figure is the planning read and the location figures are the allocation read; where it cannot — a wholesale account’s own stock, a franchise partner’s — the location figure is the only one that exists.

Channels differ in what belongs in on hand and in how the rate behaves. A store’s on hand is what is on the floor and in the stockroom, and its rate is what that door sells; its cover is read against the DC-to-door replenishment lead time, usually days, so store cover runs short by design and the DC carries the floor. The DC’s on hand is what is available to allocate, its rate is total outbound demand, and its cover is read against the vendor lead time — the long one. Ecommerce on hand is sellable stock in the fulfilment network, less units reserved against open orders and less the returns queue not yet back in sellable; its rate is online demand, which moves with paid traffic in a way store demand does not, so a shorter trailing window suits it. Wholesale on hand is the stock the brand still owns against the orders it has taken, and its denominator is the shipping plan rather than consumer sales — a different measure, covered in the wholesale-and-DTC guide.

The allocation read follows from the spread. Cover by door for one style, sorted, is the list of where stock is short and where it is long, and the transfer or the next allocation drop is the lever that moves units from the long end to the short end without buying anything; the allocation guide works through how that drop is sized. The one rule carried over from the basis section applies here with more force: a department cover across doors is total stock over total sales, never the average of the door figures, because a small door with a long cover and a large door with a short one average to a number that describes neither, while the pooled figure at least describes the stock. The allocation term covers the vocabulary.

Targets by lifecycle, derived from lead time, review period and safety stock

There is no universal number of weeks to hold, and this guide offers none. What there is, for a replenished item, is a mechanism that produces the floor from three inputs the business already knows. Cover floor = lead time + review period + safety stock, all expressed in weeks of demand. Lead time is how long an order takes to become sellable stock; the review period is how often the position is looked at and an order can be placed, so a weekly review adds up to one week before the gap is even seen; safety stock is the buffer against demand running above the rate and the delivery running late, sized from the variability of both. It is the reorder point in weeks: when on hand plus on order falls to the floor, the next order is placed, and that order has to carry the item from the day it lands to the day the one after it lands. On hand itself will sit below the floor the week before a delivery, by as much as the lead time, because the order covering that gap is already in flight; what must never fall below the floor is the position. On hand peaks at roughly the floor plus the order quantity in weeks the week a delivery lands, and the average on hand across the cycle — the trough plus half the order quantity — is the figure a stock-to-sales ratio or an inventory turn would see.

Illustrative cover floors by lifecycle derived from lead time, review period and safety stock in weeks of demand, with what governs the ceiling in each case. Not benchmarks.
LifecycleLead timeReview periodSafety stockCover floorWhat governs the ceiling
Replenished basic, domestic3115Order quantity in weeks
Replenished basic, imported124218Order quantity in weeks
Seasonal fashion, chase open6118Weeks left + exit allowance
Seasonal fashion, chase closed———noneWeeks left + exit allowance
Clearance———noneWeeks to the exit date

The rows are illustrative planning arithmetic, not observed figures. A domestically replenished basic on a three-week lead time, reviewed weekly with a week of safety stock, has a five-week floor; the same basic sourced on a twelve-week lead time with a monthly review and two weeks of safety stock has an eighteen-week floor, and nothing about the garment changed. Seasonal fashion has a floor only while a chase is possible — a six-week chase lead time plus a weekly review plus a week of buffer is eight — and once the chase window closes the floor disappears, because there is nothing a low cover can trigger. What remains is a ceiling: the trading weeks left plus the exit stock the plan intends to carry out, above which the stock is a markdown waiting for a date. Clearance has neither floor nor target, only a date: cover at the markdown rate of sale has to reach zero by the exit, or the markdown deepens.

Two refinements. Safety stock is the input most often set by habit rather than derived; the seasonal safety-stock guide covers how it is sized from demand and lead-time variability for an item that will not be reordered, and the lead-time guide how the variability is measured. And the floor is set per item, not per department: a department floor is a stock-weighted blend of items on different lead times, and reading every item against it under-protects the long-lead ones and over-stocks the short-lead ones. The benchmarks page carries illustrative planning ranges by channel for the department view; the floor for a reorder is derived, not looked up.

How weeks of supply drives the reorder, the chase, the markdown and the open-to-buy

Weeks of supply earns its place by being the figure that converts a stock position into a decision with a date on it, and four decisions read it. The reorder, on a replenished item: projected cover — on hand plus on order — falling toward the floor is the trigger, and the on-hand figure says whether an order already placed will land before the stock runs out or a second one is needed now. The replenishment guide turns the floor into a reorder point in units — the weekly rate times the floor in weeks — and a pack-rounded order quantity, which is cover arithmetic run backwards.

The chase, on a seasonal item inside its chase window: forward cover below the floor the chase lead time sets is the signal, and the window is the deadline — the same reading a week after the window closes is not a chase but a stock-out the plan has to accept, re-allocate around or trade through on substitutes. The markdown, on a seasonal item: forward cover above the trading weeks left plus the exit allowance is the overstock week, and the excess in weeks times the forward rate is the stock the window cannot sell at the ticket. The third worked example shows that excess as 35 at retail from week four — the markdown the plan already contains. Intake comes before price in both: a receipt that can be pulled forward, pushed back, cut or cancelled inside its lead time moves the cover line without spending margin, and only when the intake row has nothing left to give does the decision move to a promotion or a markdown.

The open-to-buy, pre-season and in-season: a cover target is how a sales plan becomes a stock plan. Planned closing stock for a period is the planned sales of the weeks ahead times the cover target, the open-to-buy identity solves receipts from the stock targets, and receipts less what is already on order is the open-to-buy. Change the cover target and every period’s receipts move; read actual cover above the target in-season and the open-to-buy for the periods ahead has shrunk by the excess, before anyone cancels anything. The how-to-set-OTB guide works the identity, the open-to-buy calculator runs it, and the open-to-buy term covers the vocabulary; the sales plan is the denominator the whole chain rests on.

Reading it in-season

The in-season read is weekly, on the cover line of the WSSI or the cover column of a stock report, and it has an order. First, confirm the denominator: forward against the current forecast, with the trailing figure beside it, and the gap between the two read as a statement about the forecast before anything else is read. A forward figure that has not moved while actuals have is a forecast nobody has touched. Second, find the two weeks that matter — the first week the position, on hand plus on order, crosses below the floor on a replenished item, and the first week it crosses above the weeks left plus the exit allowance on a seasonal one — and ignore every week between, because cover between the floor and the ceiling is the plan working. Third, go to the row that caused the crossing: intake landing into a week that no longer needs it, or a rate of sale that has fallen with nothing moving on the stock line. The first is an intake decision and the second a demand one, and the same cover reading has a different owner depending on which row moved.

Fourth, drop a level. A category cover reading that looks fine can hold a style that is out and a style that will never sell, and a style reading that looks fine can hold a broken size run; the department figure is the agenda for the meeting, the size-level figure is where the action is. Fifth, check the numerator before trusting any of it: stock in transit counted as on hand, stock reserved for a wholesale order still in the available figure, stock in a returns queue not yet back in sellable — each inflates cover without a unit being available to sell. The available-to-sell term covers what belongs in the on-hand figure for a cover read and what does not.

The reforecast is the step that closes the loop. Every in-season cover reading that disagrees with the plan is one of two things: a one-off the plan should absorb, or a trend the plan should adopt. The reforecast guide works through how a revised sales curve rebuilds every remaining week’s cover, and the reforecast term covers the cadence. A cover line read against a stale plan is a report about a season that is no longer happening, and the weeks of supply on it are weeks of a demand that does not exist.

Reviewing a cover report: five questions before the number is believed

A cover report arrives as a column of weeks, and the column is the last thing to read. Five questions, in order, turn it into something a reorder or a markdown can rest on; each is a definition check before it is a stock check, and a report that answers them in its header is the one worth reading.

First, which denominator, and over how many weeks? A column that does not say whether it is trailing or forward, or how long its window is, cannot be compared with another column or with last week’s. Second, which basis? Units, retail or cost, and the same on both sides — a report that carries retail stock from the merchandise system and unit sales from the till has divided two different things. Third, what is in on hand? In transit, reserved, in the returns queue, at a third-party location: each one in the numerator is cover that cannot be sold this week, and a definition change in the on-hand feed moves every figure on the report without a unit moving.

Fourth, what is each line being read against — a floor, a ceiling or a date — and is it the right one for the item’s lifecycle? A replenished basic at eight weeks against a five-week floor is trading as designed; a seasonal item at eight weeks with four left is an overstock. The same number is a different verdict on a different lifecycle, and a report that flags every figure above a single threshold flags the wrong lines. Fifth, at what level? A category figure is the agenda; the style figure is where the decision sits; the size figure is where the stock-out is. A report that passes all five is a short one, with the denominator, the basis, the on-hand definition and the floor or ceiling in its header and the lines that crossed one in its body — which is the point. Cover earns its place by turning a stock position into a date, and the five questions are what keep the date honest.

The open-to-buy template, where the cover target becomes receipts

The open-to-buy template carries the chain this guide ends on: a phased sales plan, a cover or stock-to-sales target per period, the closing-stock targets it produces, and the receipts and open-to-buy the identity solves from them. It is the place to see what one change to a cover target does to every later period’s buy.

Get the OTB Template

Free for brand, retailer, wholesaler and manufacturer teams. We email the link to your work address; requests from companies we can’t verify are reviewed first.

By submitting, you agree to .

Nine ways a weeks of supply figure misleads

Each divides correctly and describes the wrong thing. None shows up as a spreadsheet error; each shows up as a reorder placed late, a chase that lands into a sell-down, or a markdown taken on a style that was about to run out.

A zero-sales week in the divisor

An out-of-stock week sells nothing, and averaged into a trailing window it lowers the rate and inflates the cover: 2,100 units over weeks of 300, 0, 300 and 300 reads 9.33 weeks against the 7.00 the item actually carries. Drop the stocked-out week from the average, or the report tells you the stock is safe precisely because it ran out.

An average of averages

Weeks of supply is a ratio, so a department figure is total stock over total weekly sales, never the mean of the SKU figures. SKU A at 1,000 units over 100 a week reads 10.0; SKU B at 200 over 400 reads 0.5. Their mean is 5.25; the pooled figure is 1,200 ÷ 500 = 2.4. The mean is weighted by count of SKUs and the pooled figure by stock, and only the second describes how long the department lasts.

Cost in the numerator and retail in the denominator

Stock at cost over sales at retail understates cover by the cost complement; stock at retail over cost of sales overstates it by the same factor. The illustrative outerwear category reads 6.00 weeks on either basis held consistently, 2.70 with cost stock over retail sales and 13.33 the other way round. Both wrong figures are plausible, which is why the basis is written on the report.

Counting in-transit stock twice

A shipment that has left the vendor is in transit; once the DC receives it, it is on hand. If the on-hand ledger books it at receipt while the purchase order stays open until invoicing, the same cartons sit in on hand and on order for days, and the projected figure carries them twice. Define the event that moves a unit from on order to on hand, and apply it in both systems.

Trailing cover read on a seasonal build

On the illustrative seasonal dress, trailing cover reads 5.33 weeks at the week-three close while forward cover reads 3.20, and reads 2.77 at week six while forward reads 4.50 against four weeks left. Going into the peak the trailing rate is too low and flatters cover; coming out it is too high and manufactures a shortage. Read the forward figure wherever a plan exists.

A forward figure on a stale forecast

Forward cover is only as honest as the forecast under it. When actuals run behind plan and the plan is not reforecast, forward cover reads to plan while the stock builds; the trailing figure, for all its faults, will have started to rise. A gap between the two is the signal that the forecast needs revisiting, not that one of them is broken.

Treating the floor as a target

A five-week floor means on hand plus on order should not fall below five weeks at the point an order can be placed; it does not mean on hand should sit at five weeks all the time. A replenished item ordered every three weeks oscillates in on-hand cover between roughly the floor and the floor plus three — eight weeks at the moment of receipt on the illustrative basic. Reading eight as overstock and cutting the order is how the floor gets breached a cycle later.

One number for a style with a broken size run

The illustrative style reads 4.8 weeks of supply in total while size M carries 3.5 and size XL 9.5. Against a four-week lead time, the style figure says reorder with time to spare and the size figures say M will be out before anything lands. The style number is the stock-weighted average of a shortage and an overstock, and describes neither.

Weeks and days, months and periods, mixed on one report

Seven weeks of supply is 49 days; a stock-to-sales ratio of 2.0 on a four-week period is eight weeks of cover and on a five-week period ten. A report that carries weeks for one category and a period ratio for another is comparing two things that look alike, and a 4-5-4 calendar makes every third period a different length.

What moves the numerator and the denominator in each vertical

The division is identical everywhere. What changes is what belongs in on hand, how the weekly rate behaves across the calendar, and whether the item is read against a floor, a ceiling or a date. Apparel is the reference case.

Apparel

The numerator is on-hand units by style, colour and size, and the size dimension is where it misleads: a style can read healthy while its core sizes are gone. The denominator is a weekly rate that moves with the season curve, so forward cover is the working figure and the trailing figure is the check on the forecast. Markdowns change the retail value of stock without touching units, which is the reason cover in units and cover at retail part after the first reduction.

Footwear

The numerator is pairs across a size run and often a width run, so size-level cover matters more than in apparel: a run broken in the half sizes stops presenting long before the style’s total pairs run out. Carryover styles that continue across model years are read against a replenishment floor; a model-year change puts the outgoing style on a clearance ceiling the week the new one lands. Prebooked pairs that ship straight on to wholesale accounts never sit in on hand and belong outside the figure.

Accessories and bags

With no size run the unit of cover is the style-colour, and the split that matters is between evergreen core and fashion colourways: the core is read against a floor and the fashion colour against the weeks left in its window. Leather-goods minimums push a colourway buy above demand, so the opening cover is set by the minimum rather than by the plan and the denominator has to be honest about what the colour will sell. Attach rate ties the weekly rate to the apparel or footwear it sells beside.

Home and furniture

Long lead times and container minimums make the receipt lumpy, so cover steps up by a container’s worth and sells down, and the floor is set by an ocean lead time measured in months rather than weeks. Projected cover — on hand plus containers on the water — is the figure a reorder is placed on, because the on-hand figure alone will always say reorder too late. Floor-set changes and dealer prebooks fix the numerator before any sell-through exists, and special-order lines carry no stock and no cover.

Outdoor

Counter-seasonal categories — snow and summer — peak in different months, so one category’s trailing rate is at its least useful exactly when the other’s is at its best, and a department figure blends a build with a sell-down. Gear on model years is read against a clearance ceiling at changeover or carried into the next model year on a floor; MAP pricing narrows the markdown route, so the cover the ceiling cannot clear carries over. Warranty and repair stock sells nothing and belongs outside the numerator.

Sporting goods

Seasonality runs by sport — ski, golf, cycling, racquet — on calendars that do not align, so cover is computed per sport against that sport’s own forward curve, never on a department average. Team and roster orders ship against the order and belong outside on hand, and dealer prebooks commit quantity before the season opens, so the opening cover is a buy decision already taken and the in-season read is about the at-once stock around it. Prior model years at changeover sit on a clearance ceiling.

Beauty and wellness

Shade ranges carry low-velocity depth at the ends of the range, so a shade-level cover read matters the way size-level does in apparel: the hero shades run out while the range total reads comfortable. Period-after-opening and expiry dating put a hard ceiling on cover regardless of the plan — stock that will date before it sells is not supply. Testers and gratis leave the ledger without a sale, and gondola resets move the numerator on a date set by the retailer rather than by demand.

Toys and games

The gifting peak concentrates the denominator into a few weeks, so trailing cover in October understates the rate ahead by a large factor and forward cover against the holiday plan is the only usable figure. Licensed product sells inside a licence window and a sell-off period, which sets a ceiling by contract rather than by season. Retailer commitments fix the buy before sell-through exists, and a reorder is read against a lead time that includes safety-standard testing.

Baby and juvenile

Registry demand spreads purchases over weeks between an item being added and bought, which smooths the weekly rate and makes a longer trailing window — eight weeks rather than four — the better denominator on replenished hard goods. High cost and large cube mean a few units of excess cover tie up more capital and more floor than the same count in apparel. A recall or a safety-standard change takes stock out of the numerator at once, and model-year changeovers put the outgoing model on a clearance ceiling.

Jewelry and watches

Low velocity and high cost per piece make cover long by construction — pieces rather than units, and a weekly rate that may be below one for a single reference — so cover is read monthly or by model year and in cost dollars, where the precious-metal cost base changes the value of the same pieces without any sale. Memo and consignment goods sit on the floor but not on the balance sheet, and belong outside an owned-stock figure. A model-year change sets the ceiling on the outgoing reference.

See the connected workflow in RetailNorthstar →

Frequently asked questions

What is weeks of supply?
Weeks of supply (WOS) is the number of weeks the current inventory would last at the expected rate of sale: on-hand units or dollars divided by average weekly sales. 2,400 units selling 300 a week is 8.0 weeks of supply. The rate can be trailing — the average of recent actual weeks — or forward, the average of the planned weeks ahead, and the two are different figures on seasonal demand.
What is the weeks of supply formula?
Weeks of supply = on-hand inventory ÷ average weekly sales, with both sides in the same basis — units, retail dollars or cost dollars. Trailing weeks of supply divides by the average of the last four or more weeks of actual sales; forward weeks of supply divides by the average of the next four weeks of the plan, or of the weeks left in the window where fewer remain. Projected weeks of supply adds on-order stock to the numerator: (on hand + on order) ÷ average weekly sales. Days of supply is weeks of supply × 7.
How do you calculate weeks of supply step by step?
Fix the grain and the basis; take closing on-hand stock available to sell; choose trailing or forward and state it; average the weekly sales in that window, excluding any out-of-stock week from a trailing average; divide stock by the average; add on order for the projected figure; then read the result against the floor for a replenished item or the ceiling for a seasonal one. In this guide’s illustrative basic, 2,100 units at the week-one close over a trailing average of 297.5 is 7.06 weeks, and over the forward plan of 300 it is 7.00.
What is the difference between forward and trailing weeks of supply?
They share a numerator and differ in the denominator. Trailing divides by what sold in recent weeks; forward divides by what the plan expects to sell in the weeks ahead. On flat demand they agree — the illustrative basic never reads more than 0.10 weeks apart. On a seasonal build they diverge: the illustrative dress reads 5.33 weeks trailing and 3.20 forward at the week-three close, and 2.77 trailing against 4.50 forward at week six. Planners read forward cover because the question is how long the stock lasts from here, and the trailing figure is kept beside it as the check on the forecast.
Is weeks of supply the same as weeks of cover?
Yes — weeks of cover is the same figure under a second name, and forward cover is the forward version of it. The one distinction worth keeping is in the denominator rather than the name: cover computed against the forecast ahead is forward cover, and cover computed against recent actuals is trailing weeks of supply. A WSSI carries forward cover; a stock report built from the sales ledger alone can only carry the trailing figure, so check which one a column holds before comparing them.
How is weeks of supply different from inventory turnover and the stock-to-sales ratio?
All three put stock against sales and differ in the window. Weeks of supply is a point-in-time figure in weeks: stock now over a weekly rate. Inventory turnover is a period figure: cost of goods sold over average inventory at cost, and 52 ÷ turns gives the weeks of supply the average inventory represents — 6.5 turns is 8.0 weeks. The stock-to-sales ratio is a period snapshot: opening stock over that period’s sales, so a ratio of 2.0 on a four-week period is eight weeks of cover and on a five-week period ten. Sell-through adds the fourth read: a week’s sell-through of 20% is 4.0 weeks of supply at that week’s rate, since (1 − 0.20) ÷ 0.20 = 4.0.
Should weeks of supply use units or dollars?
Either, as long as numerator and denominator use the same one. Units are the truth for a single style and the only basis that survives a markdown, because a markdown changes retail value without changing a unit. Retail dollars suit a category or department, where unit prices differ, and tie to the merchandise plan and the open-to-buy. Cost dollars answer the working-capital question. The illustrative outerwear category reads 6.00 weeks at retail and 6.00 at cost while the cost complement is constant, and 3.71 at retail against 4.29 at cost once markdowns have reduced the retail value of the stock still on hand.
What is projected weeks of supply?
Projected weeks of supply is (on hand + on order due inside the horizon) ÷ average weekly sales — the cover the position would have if every open delivery landed. The illustrative basic reads 7.00 weeks on hand at the week-one close and 13.00 projected with two deliveries of 900 still to come. Report it beside the on-hand figure rather than instead of it: the on-hand figure says what can be sold today, and the projected figure is only as reliable as the delivery dates behind it.
How many weeks of supply should you hold?
There is no universal figure, and this guide offers none. For a replenished item the floor is derived from the mechanism: lead time plus review period plus safety stock, all in weeks of demand, so an illustrative item on a three-week lead time, reviewed weekly, holding one week of safety stock has a five-week floor on the position — on hand plus on order — and on hand peaks at roughly the floor plus the order quantity in weeks the week a delivery lands. For a seasonal item there is no floor once the chase window closes, only a ceiling: the trading weeks left plus the exit stock the plan intends to carry out. Clearance stock is read against the weeks to the exit date.
How does weeks of supply drive a reorder, a chase or a markdown?
Projected cover — on hand plus on order — falling toward the floor on a replenished item triggers the reorder, and the on-hand figure says whether an order already placed will land before the stock runs out. Forward cover below the floor on a seasonal item inside its chase window is the chase signal; the same reading after the chase window closes is a stock-out the plan has to accept or re-allocate around. Forward cover above the trading weeks left plus the exit allowance is the overstock week, and the excess — in the illustrative outerwear category, 0.78 weeks or 35 at retail at the week-four close — is the markdown the plan has already written. The open-to-buy carries the same arithmetic forward: planned closing stock is planned sales ahead times a cover target.
Why does weeks of supply at the style level hide a broken size run?
Because the style figure is a stock-weighted average of its sizes. The illustrative style holds 1,200 units selling 250 a week, 4.8 weeks of supply, while size M holds 380 selling 110, 3.5 weeks, and size XL holds 190 selling 20, 9.5 weeks. Against a four-week lead time the style reads safe and M runs out before a reorder lands, while XS and XL carry eight to nine weeks that will clear at markdown. Weeks of supply is computed at the size level for any item sold in a size run, and the style figure is read as a total, not as a stock position.

See how RetailNorthstar keeps the sales plan, open-to-buy and stock targets on one shared data model — so the cover target that sets a category’s closing stock comes from the plan that sets the buy.

No credit card. No commitments.