How Do You Reduce Overproduction Fashion Teams Build Into the Opening Buy?
Picture the Monday trading meeting where twelve new hoodie options are on the screen, each carrying a full-depth opening order, and the first question is which of them the business would still be happy to own in January. That is the practical shape of the effort to reduce overproduction fashion buyers face: broad initial commitments feel safe when the design team is enthusiastic and factory slots are tight, yet they quietly create weak-SKU inventory that is cleared at a markdown, often while the real winners run short.
Overproduction here means units made beyond what customers buy at an acceptable price inside the selling window. It is a commitment decision, taken before demand evidence exists, so the fix is to change when and how much you commit rather than to cut the range or accept lower sales. The method has four parts: segment proven core from uncertain newness, commit in stages with a defined read period, protect winners with reserved supplier capacity, and measure markdown and residual stock after every test. Lead times are what force early commitment in the first place, which our overview of the fashion supply chain sets out.
Every figure in the tiles above comes from the hypothetical worked example further down. None is a benchmark, a supplier price or a typical result, and each should be replaced with your own cost, price and clearance assumptions.
Overproduction, Overstock and Residual Stock: Keeping the Terms Apart
Teams often use these words interchangeably, so post-season reviews argue about labels instead of causes. Separating them shows which KPI exposes each one.
Terms to separate in a post-test review
| Term | Working definition | First KPI to watch |
|---|---|---|
| Overproduction | Units made beyond demand at an acceptable price in the selling window, created at the order before demand is known | Full-price sell-through against units received |
| Overstock | Stock above the cover the business wants to hold at a point in time | Weeks of cover (available stock divided by forecast weekly sales) |
| Residual stock | Units left when the planned selling window ends, after any markdown | Residual units and their value at cost |
| Stockout on a winner | Demand that cannot be served because a proven style was bought too shallow or reordered too late | Lost sales and weeks out of stock |
Definitions vary by retailer. Write yours down before the first read so that every review uses the same ones.
The last row matters as much as the first. Any plan to reduce overproduction fashion buyers can defend has to count stockouts on winners as a cost, otherwise it simply trades markdown for lost sales.
Step 1: Segment Proven Core Versus Uncertain Newness
The first decision is not how many units to buy but how much evidence sits behind each option. A repeat block in a proven fabric and colour has a sales history and a known size curve. A new silhouette in a new fabric at a new price point has none, and treating the two with the same depth is the root of most surplus.
A three-way segmentation of the buy
| Segment | What qualifies | First-buy approach | Reorder approach |
|---|---|---|---|
| Proven core | Repeat block, same fabric, colour and price tier, with run-rate history across more than one season | Commit to full planned depth; protect availability | Replenish on a cover-based reorder rule agreed in advance |
| Adjacent variant | New colour, trim or decoration on a proven block | Commit part of planned depth; hold the rest as a conditional tranche | Reorder against the first read, with capacity held |
| Uncertain newness | New silhouette, fabric, wash or price tier with no sales history | Test quantity only, in a representative channel | Scale only if the pre-agreed gate is met |
Segment criteria are a planning convention, not a standard. Set the qualifying history that suits your own category.
The aim is different depth, not less ambition. The ratings below show how the segments differ on the questions that drive buy size. For choosing which tested options deserve more, see our guide to which styles deserve a bigger buy.
How the three segments differ
1 dot = low, 5 dots = high
| Proven core | Adjacent variant | Uncertain newness | |
|---|---|---|---|
| Demand certainty | ●●●●● | ●●●●● | ●●●●● |
| Sensible depth at first buy | ●●●●● | ●●●●● | ●●●●● |
| Markdown exposure if wrong | ●●●●● | ●●●●● | ●●●●● |
Indicative ratings for planning discussion, not measured data.
Step 2: Use Staged Commitments and a Clear Read Period
A staged commitment splits one intended volume into a test tranche that is firm, and one or more follow-on tranches that are conditional on evidence. The read period is the fixed window in which the test tranche is judged. Both must be written down before launch, or the read becomes a debate about whether the numbers are good enough.
Designing the stages
- 1Set the test quantityLarge enough to read, small enough to clear if wrong. Where feasible, 500 pieces per style and colour is a proposed test quantity; it varies by fabric, dye, trims and factory.
- 2Fix the read periodState the selling weeks, the channel and the metrics. Do not move the window after seeing early numbers.
- 3Pre-agree the gateWrite thresholds for sell-through, returns and margin, and the action at each outcome, before delivery.
- 4Back-solve the latest decision dateCount back from season end by the follow-on lead time, leaving enough selling weeks to justify the reorder.
A framework to adapt. The thresholds belong to the buyer and differ by category and price point.
The back-solve in step four is simple arithmetic: latest decision week equals season length, less follow-on lead time, less the minimum selling weeks that justify a reorder. With an illustrative 20-week season, a six-week follow-on lead time and six selling weeks as the minimum, the latest decision falls at the end of week 8, which leaves slack beyond a three-week read period. If the sum falls before your read period ends, the style is too late to stage. Our lead times guide explains what drives the follow-on figure.
An illustrative decision gate for one test style
| Outcome at the end of the read period | Signal against buyer-set thresholds | Action |
|---|---|---|
| Above the upper threshold | Sell-through, margin and returns all clear their gates | Release the follow-on tranche into reserved capacity |
| Below the lower threshold | Sell-through under the gate, or returns above the limit | Stop; plan markdown and residual stock clearance for the test units only |
Thresholds are intentionally left blank: set them from your own margin structure and category history.
Step 3: Protect Winners With Supplier Capacity Reservations
Staging has a cost: a winner identified in week three is only useful if the factory can make more of it before the window closes. A capacity reservation is an agreement with the supplier to hold production slots, and often fabric and trims, for a follow-on order that has not yet been confirmed.
Reservation terms differ between suppliers and programmes, and each one moves risk from one party to the other. Ask early and put the answers in writing.
Reservation terms to agree before the test lands
| Term | Question to settle | Why it matters |
|---|---|---|
| Reserved volume | How many units, across how many styles, and can the allocation move between styles? | Winners are unknown at booking, so flexibility between styles is worth more than total volume |
| Expiry date | By what date must the order be confirmed or the slot is released? | Must sit after the read period and before the latest decision date |
| Fabric and trim position | Is yarn or fabric held, booked or only planned? Who owns it if unused? | Capacity without materials does not shorten the lead time |
| Cost of holding | Is there a deposit, a fee or a price assumption attached? | A reservation that costs more than the surplus it prevents is not worth holding |
A checklist of questions, not a template contract. Terms are negotiated per supplier and programme.
Designing the same discipline across more than one supplier is covered in our guide to flexible sourcing for an established retailer.
Step 4: Track Markdown and Residual Stock After Each Test
A test only helps if the outcome is recorded and fed into the next decision. A one-line log per option, completed at the end of the read and again at season end, shows what the tested newness cost.
A minimal post-test log
| Field | How to calculate it | When to record it |
|---|---|---|
| Full-price sell-through | Units sold at full price divided by units received (state the denominator) | End of read; end of season |
| Weekly rate of sale | Units sold divided by weeks in stock, which normalises for stockouts | End of read |
| Returns rate | Units returned divided by units sold, with reasons if available | End of read and after |
| Markdown value | Units cleared multiplied by the gap between full price and clearance price | End of season |
| Residual units and value | Units left at the end of the window, valued at cost | End of season |
Definitions of sell-through and markdown rate vary by retailer. Keep one definition across all tests so that results are comparable.
Reviewed across several tests, the log shows which kinds of newness repay a larger first buy and which should always be tested small.
The KPIs Explained: Markdown, Stock Turn, Gross Profit and Sell-Through
Four measures carry most of the argument, and they are easy to blur. Each is defined below with its formula; the worked example that follows uses the same definitions.
KPI definitions used on this page
| KPI | Formula | Reads as | Where definitions vary |
|---|---|---|---|
| Sell-through | Units sold divided by units received or available over the same period | How much of what arrived has been bought | Denominator (received, or available including opening stock); full-price versus all-price units |
| Markdown | Markdown rate = value or units sold below full price, divided by total sales value or units. Depth = (full price minus sale price) divided by full price | How much margin is given away to clear stock | Whether the rate is by value or by units, and whether it includes promotions |
| Stock turn | Cost of goods sold divided by average stock at cost over the same period | How many times the stock investment is recovered | Retail-value or cost basis; length of period; whether annualised |
| Gross profit | Net sales excluding VAT, minus cost of goods sold. Gross margin = gross profit divided by net sales excluding VAT | What the buy earned after product cost | Which costs are in cost of goods (landed cost, duty, freight, markdown funding) |
Gross margin is not markup. Markup is (selling price minus cost) divided by cost, so a 60 per cent gross margin on a product is a markup of 150 per cent.
A simple illustration of stock turn, with hypothetical figures: cost of goods sold of £240,000 and average stock at cost of £60,000 over the same period give four turns, since 240,000 divided by 60,000 equals 4. A broad buy that sits in the warehouse lowers that figure even when every unit eventually sells.
Worked Example: A 30,000-Unit Speculative Buy Against a 3,000-Unit Initial Test
The comparison below is hypothetical and illustrative. It shows how exposure, read period and lost sales trade off, not what is typical or likely. Every input is a labelled assumption, and the arithmetic has been checked with a short script.
Assumptions (all illustrative)
| Input | Assumed value | Notes |
|---|---|---|
| Options | 6 hoodie options, same price tier | Plan A: 5,000 each. Plan B test: 500 each |
| Landed cost per unit | £14 | Plan B test units assumed £1 higher (£15) because small runs usually cost more per unit |
| Full price (excluding VAT) | £35 | Full-price gross profit £21 per unit, a gross margin of 60.0 per cent |
| Clearance price (excluding VAT) | £10 | Loss of £4 per surplus unit against cost; markdown depth 71.4 per cent |
| Season | 20 weeks of full-price selling | Surplus cleared once, at the end |
| Plan B read and follow-on | 3-week read; 6-week lead time with capacity reserved | Follow-on lands at the start of week 10 and is sized to the remaining demand |
| Demand scenario 1: concentrated | Weekly full-price demand per option: 400, 250, 100, 50, 30 and 20 units | 17,000 units across the season |
| Demand scenario 2: broad | 250 units a week for every option | 30,000 units across the season |
All inputs are assumed for illustration. Real costs, prices, demand and lead times differ by product and programme.
Plan A commits £420,000 at cost on day one (30,000 units at £14). Plan B commits £45,000 (3,000 units at £15), then buys follow-on units only where the test supports it.
Season outcome under each demand scenario (illustrative)
| Measure | A: scenario 1 (concentrated) | B: scenario 1 (concentrated) | A: scenario 2 (broad) | B: scenario 2 (broad) |
|---|---|---|---|---|
| Units bought | 30,000 | 11,850 | 30,000 | 19,500 |
| Cost committed | £420,000 | £168,900 | £420,000 | £276,000 |
| Units sold at full price | 14,000 | 11,750 | 30,000 | 19,500 |
| Surplus cleared at £10 | 16,000 | 100 | 0 | 0 |
| Full-price sell-through (full-price units divided by units bought) | 46.7% | 99.2% | 100.0% | 100.0% |
| Full-price demand not served | 3,000 | 5,250 | 0 | 10,500 |
| Net sales (excluding VAT) | £650,000 | £412,250 | £1,050,000 | £682,500 |
| Gross profit | £230,000 | £243,350 | £630,000 | £406,500 |
| Gross margin | 35.4% | 59.0% | 60.0% | 59.6% |
Illustrative model. Plan B follow-on quantities are sized exactly to remaining demand, which is optimistic; a real buyer would carry some error in either direction.
In scenario 1 the speculative buy leaves 16,000 surplus units, because most of the range sells below its depth while the two leading options run out of stock, and the staged plan ends with a higher gross profit despite serving fewer full-price customers. In scenario 2 the position reverses: every option is a winner, the surplus is zero, and the staged plan gives up sales during the weeks before its follow-on lands.
Season gross profit by plan and scenario (illustrative)
Gross profit in pounds
Hypothetical outputs of the model above. They show how the ranking depends on the demand pattern, not what any real range will earn.
Exposure at the read date tells a different story from final profit. At the end of week 3, Plan A in scenario 1 holds 27,450 unsold units, worth £384,300 at cost, that cannot be cancelled. Plan B holds 1,400 unsold test units, worth £21,000 at cost, before it decides anything. Exposure is not loss, since in scenario 2 Plan A sells every unit eventually, but it is the risk behind the decision while evidence is still arriving.
What Would Have to Be True for Each Plan to Win
Neither plan is better in general, and any decision to reduce overproduction fashion buyers make should rest on which conditions apply. Each wins under conditions that a buyer can test against the evidence available before the order, and the break-even ratio gives a quick check. A missed winner unit costs £21 of gross profit in this model; a surplus unit cleared costs £4. Staging therefore pays when it avoids at least 5.25 surplus units for every full-price unit it fails to serve.
Conditions that favour each plan (illustrative)
| Condition | Favours the 30,000-unit buy | Favours the 3,000-unit test |
|---|---|---|
| Demand pattern | Broad: most options sell near planned depth | Concentrated: a few winners carry the range |
| Evidence before the order | Strong, from proven blocks and past colour performance | Weak, because the options are new |
| Follow-on lead time and slot | Long, or no reserved capacity, so a winner cannot be repeated in the window | Short enough, with reserved capacity and booked materials |
| Clearance loss per surplus unit | Small, so surplus is cheap to clear | Large, so each surplus unit is expensive |
A decision aid built on the illustrative model, not a rule. Swap in your own figures to see which column your next buy belongs to.
In scenario 1 the model avoids 15,900 surplus units and gives up 2,250 full-price units, a ratio of about 7.07 to 1, above the 5.25 break-even. In scenario 2 it avoids none and gives up 10,500, so the ratio is zero. The break-even moves with the clearance price: at a deeper clearance that costs £8 a unit rather than £4, the ratio falls to 2.625, and staging pays more easily. Treat the figure as a prompt to do your own sum, not a target.
A test tranche still needs a representative channel and a supplier that can repeat the sealed sample. Where a style is already proven, an early full-depth commitment may be the better decision.
If a winner emerges from the test, the question becomes how much to reorder and when, which our guide to the reorder point formula covers.
Common Mistakes When You Try to Reduce Overproduction Fashion Programmes
Most failures come from the process around the test rather than the test itself.
- Judging a test on units sold without normalising for weeks in stock, so a stockout looks like weak demand
- Running the test in a channel that does not represent the wider customer, then scaling on its result
- Staging every style, including proven core, which creates stockouts on products that never needed a test
- Counting markdown but ignoring residual stock carried into next season
Each is cheap to prevent by writing the rules before launch. For a hoodie-specific version, see how to reduce hoodie overstock without losing bestseller sales.
The Decision: What to Do Before the Next Opening Buy
To reduce overproduction fashion buyers can start by sorting the next season into the three segments, and commit full depth only where the evidence supports it. Put the uncertain newness into a test tranche with a written gate, back-solve the latest follow-on decision date, and ask suppliers for a reservation whose expiry sits between the two. Monitor full-price sell-through during the read and markdown plus residual stock after season end, and compare them with the previous season.
Evidence that would change the decision includes a follow-on lead time longer than the remaining window, a test channel you cannot trust, or a style whose history is stronger than you thought. If the broad buy you are about to sign were wrong on half of its options, which units would you hold in January, and what would you have wanted to know in week three to avoid them?
How Ruhrose Approaches Testing and Staged Buys
Ruhrose develops hoodies, jersey, denim and woven casualwear for UK buyers, with production in Bangladesh, and sampling comes before bulk. Where a buyer wants to reduce overproduction fashion ranges carry by testing newness first, we discuss a smaller initial run, which is a risk-reduction model and not risk-free, and a proposed test quantity of 500 pieces per style and colour where it is technically and commercially feasible. It varies by fabric, dye, trims and factory, and it is not a guaranteed minimum. We would rather agree the read period and the follow-on plan at the sampling stage than after the goods land. The approach is described in our test-and-scale guide.