D The Wholesale Apparel Business Model

This chapter teaches you how money actually moves through a wholesale apparel brand: the three price points and the multipliers between them, the seasonal calendar that dictates every deadline, the order types that carry wildly different risk, the payment terms that decide whether you survive, and the cash conversion cycle that kills profitable companies. Everything is worked in real numbers. At the end we translate all of it into tables, fields, constraints and screens your ERP must have.

What you need to know first

Start with the shape of the industry. A brand designs clothes and has them made. It does not usually own factories. A retailer buys those clothes and sells them to the public. When the brand sells to the retailer, that transaction is wholesale. When the brand sells to the public itself (its own website, its own store), that is direct-to-consumer, abbreviated DTC.

Most brands do both. The two channels have almost nothing in common financially, and half the pain in this chapter comes from running them out of one company.

In wholesale, the brand is the vendor and the retailer is the account or customer. The person at the retailer who decides what to buy is the buyer. A single retail chain may be one account with many doors. A door is one physical store. "We're in 40 doors at Nordstrom" means the product is stocked in 40 branches.

Styles, colorways and SKUs

Now the product vocabulary, which chapter B introduced and we lean on constantly here. A style is a design: "the Ridgeline utility jacket." A colorway is that style in one color: Ridgeline in Olive. A SKU (stock keeping unit) is the smallest thing you can actually count and ship: style + color + size.

Ridgeline / Olive / Medium is one SKU. A style with 3 colors and 6 sizes is 18 SKUs. This matters because money is planned at the style level and inventory is counted at the SKU level, and confusing the two is the single most common beginner error in apparel software.

Three prices attach to every style, and you must never mix them up:

• Cost is what the brand pays to have the garment made and delivered to its warehouse. Also called landed cost once you have added freight and import duty.
• Wholesale price is what the retailer pays the brand per unit. Sometimes called the list price or the WSP.
• Retail price is what the shopper pays. Written MSRP (manufacturer's suggested retail price) or RRP (recommended retail price). It is a suggestion. The retailer legally owns the decision.

Margin is profit expressed as a share of the selling price. Markup is profit expressed as a multiple or share of cost. They are different numbers for the same fact and the industry uses both in the same sentence. A jacket costing $26 and selling wholesale for $65 has a markup of 2.5× and a margin of 60%. Learn to convert on sight: margin % = (price − cost) ÷ price.

P&L, invoices and the terms of payment

A P&L (profit and loss statement) lists revenue, subtracts costs, and reports profit for a period. The top block is revenue minus COGS (cost of goods sold) which equals gross profit. Below that sit operating expenses (salaries, rent, software, trade shows), which are not tied to any one garment. Gross profit minus operating expenses is operating profit.

A P&L records revenue when goods ship, not when the money for them arrives. Get that backwards and a healthy statement will tell you nothing about the bank balance. Cash is a separate story, and that gap is the subject of half this chapter.

When you ship goods you send an invoice: a document saying "you owe us $18,400, due on this date." The unpaid invoices you are owed are accounts receivable or AR. The rule for when payment is due is the terms: "net 30" means pay the full amount within 30 days. Money you owe suppliers is accounts payable or AP.

Seasons, linesheets and purchase orders

A season is a batch of styles designed, sold and shipped together: Spring/Summer 2027, Fall/Winter 2026. Seasons are the industry's unit of planning. Almost every number a brand tracks is "per season." A linesheet is the sales document for a season: a page per style with photo, colors, sizes, wholesale price, MSRP, fabric content, minimums and delivery window. It is the thing a buyer writes an order from.

A purchase order or PO is a customer's commitment to buy: these SKUs, these quantities, this price, ship between these dates. When the brand orders from a factory it issues its own PO, sometimes called a cut ticket (the instruction to cut the fabric). Both directions matter, and chapter B's order-to-cash and procure-to-pay flows are exactly these two document chains.

Finally, two abbreviations you will see all chapter. FOB ("free on board") is a price quoted by the factory that includes making the garment and loading it onto the ship, but not ocean freight, insurance or import duty. MOQ is minimum order quantity: the smallest run a factory will accept, per style and often per color. MOQs are why small brands carry so much of the wrong size.

The three price points and the multipliers between them

Everything in wholesale pricing is a chain of multiplications from cost up, checked against a market price coming down. Start from a real cost sheet.

Step one: the cost sheet and landed cost

COST SHEET  -  Style W-4210 "Ridgeline" utility jacket
Origin: Vietnam   FOB terms   Order qty 1,000 units

  Shell, 100% cotton twill  2.10 m @ $4.20/m        8.82
  Lining                    1.30 m @ $1.60/m        2.08
  Trims: 1 zip, 6 snaps, 2 drawcords                2.35
  Labels, hangtag, polybag, carton share            0.85
  CM (cut & make, incl. factory overhead+margin)    6.40
                                                 -------
  FOB PRICE PER UNIT                                20.50

  + Import duty @ 20% of FOB (illustrative rate)     4.10
  + Ocean freight, drayage, customs brokerage        1.35
  + Inbound receiving at the warehouse               0.30
                                                 -------
  LANDED COST PER UNIT                              26.25

Read this top to bottom. The first block is what the factory charges: raw materials (the "bill of materials"), trims, packaging, and CM (cut and make), which is the factory's labor and profit. Add them and you get the FOB price, $20.50. That is the number a factory quotes you. Beginners treat it as their cost. Several more charges land before the goods reach a shelf, and those charges are the difference between a healthy price and a loss.

The second block is everything between the ship's rail and your warehouse shelf. Two words in it need a definition. Drayage is the short truck move from the port to the warehouse. And most small brands do not run their own warehouse. They rent one, called a 3PL (third-party logistics provider): an outside company that receives your cartons, stores them, then picks, packs and ships orders for a per-unit fee. The 30 cents on that line is what the 3PL charges to take the goods in.

Duty and the HTS code

The bigger line is duty: the tax the importing government charges when goods cross the border. Duty is charged per product, and the product is identified by an HTS code, a number from the Harmonized Tariff Schedule, the official US list that gives every importable thing a classification and a rate. Apparel lives in chapter 61 (knitted or crocheted garments) and chapter 62 (the rest, mostly woven).

The baseline rate that applies to countries the US trades with normally is the most-favored-nation or MFN rate, and in apparel it is high compared with most other goods. Pull the official schedule and count: as of July 2026, the stated MFN rates across chapters 61 and 62 run from free to 32%, the median is about 9.4%, and roughly a quarter of the stated rates are 15% or more.

The 2025–2026 tariff churn

On top of that baseline sit whatever emergency, retaliatory or negotiated measures happen to be in force, and that layer changed repeatedly through 2025 and 2026. In February 2026 the US Supreme Court ruled that the "reciprocal" tariffs imposed under the International Emergency Economic Powers Act (IEEPA, a law that lets a president regulate trade during a declared national emergency) were unlawful, and ordered the collected duties refunded.

Days later the administration replaced them with a temporary 10% import surcharge under Section 122 of the Trade Act of 1974, a provision that allows a short-lived across-the-board surcharge to address a payments crisis. That surcharge was proclaimed on 20 February 2026, took effect on 24 February and was written to run only to 24 July 2026 unless Congress extended it.

In parallel the administration opened new Section 301 investigations, which are country-by-country inquiries into unfair trade practices that can end in fresh tariffs. Through all of that churn the average applied tariff rate on US apparel imports was still 21.6% as of May 2026, against 15.2% in January 2025, according to the US Fashion Industry Association's 2026 benchmarking study.

A free-trade agreement is a treaty that lets qualifying goods enter at a reduced or zero duty, and a few softened the blow: as of mid-2026, CAFTA-DR (the Dominican Republic–Central America agreement) and USMCA (the US–Mexico–Canada agreement) were the main remaining routes to preferential treatment. Every figure in this paragraph is "as of July 2026" and every one of them is expected to move.

Those numbers will have shifted by the time you read this, so take the lesson rather than the figures. Duty is data: a rate looked up per HTS code, per country of origin, per date, with history kept. The rate you used when you costed the style in October is not necessarily the rate you paid when the goods landed in July, and as 2026 demonstrated, you may later be owed a refund on the difference. It must never be a constant in your code, and never a single column you overwrite.

Freight, allocation and transit times

Freight is allocated: you pay per container, then divide across the units in it. Ocean transit from Asia commonly runs about two to three weeks to the US West Coast and four to six weeks to the East Coast, though it varies a lot by lane, carrier and routing, and that is before you add port congestion, customs clearance and the truck to the warehouse.

Air freight arrives in days but costs several times as much per kilo, which is why "we'll air it" is an emergency measure that eats an entire style's margin.

Where the goods actually come from

Country of origin does real work. It sets your duty rate, your lead time, your minimum order quantity and your compliance paperwork, so it belongs in your data model from day one. Asia still makes most of what America wears, but the map moved fast in 2026. Official US trade statistics for the first five months of that year, as reported in the USFIA benchmarking study, look like this:

The headline is that China's share fell below both Vietnam's and Bangladesh's for the first time in decades, and that brands spread their orders across more countries than before. For you as an engineer, that trend is a requirement in disguise: the same style will be made in two or three countries within a couple of seasons, each with a different duty rate, lead time and factory minimum. Country of origin therefore belongs on the production order or the inventory lot, never on the style.

Step two: wholesale price

Wholesale price is set as a multiple of landed cost. Published guidance for wholesale fashion pricing puts the practical floor at roughly 2.2×–2.5× landed cost, because below that there is not enough gross margin left to pay for sample development, showroom commission, trade shows and the deductions we will meet later. We set $65.00, which is 2.48× our $26.25 landed cost, giving a gross margin at list of 59.6%.

Step three: MSRP and keystone

Keystone is the oldest convention in retail, and it is a doubling: the shop sells at twice what it paid. Wholesale $65 becomes retail $130, and the retailer earns a 50% margin. In fashion, keystone sits close to the floor. A store paying rent, staff and card fees cannot survive on much less, so anything under 2× makes you unattractive to buy. Published guidance for wholesale fashion pricing puts typical multipliers in these bands, though they vary by category, country and how strong your brand is in the room:

We are contemporary, so MSRP is $148, which is 2.28× wholesale. Sanity-check the whole chain: $148 ÷ $26.25 = 5.6× landed cost. The same guidance offers a rule of thumb of roughly 3.5×–5× from landed cost to MSRP, so we sit a little above the top of the band. That is a comfortable place to be. Fall below the band and there is no room left for anyone's discounts.

Note the direction of travel in practice. Start at the price the customer will accept for that kind of jacket in that kind of store ($148), divide by the segment multiplier to get $65 wholesale, then check backwards that your landed cost fits under $28 or so. If it does not, you change the garment: lighter shell, five snaps instead of six, a different country. Costing is a negotiation with the design. You move the garment until the arithmetic works.

Initial markup versus maintained markup

These two terms belong to the retailer but you must speak them fluently, because your buyer is judged on the second one.

Buyer's math on style W-4210, 1,000 units

  Ticket (MSRP)                             148.00
  List wholesale cost                        65.00
  INITIAL MARKUP  (148 - 65) / 148         = 56.1%

  Season selling result:
     580 units sold at full price 148.00 = 85,840
     270 units sold at -30%       103.60 = 27,972
     150 units sold at -50%        74.00 = 11,100
                                          --------
     Season revenue                        124,912
     Average unit retail (AUR)              124.91  (84% of ticket)

  Buyer's real cost:
     net of 8% volume discount               59.80
     net of 8/10 EOM cash discount           55.02
     less markdown money from brand           2.99
     NET COST PER UNIT                       52.03

  MAINTAINED MARKUP
     (124,912 - 52,026) / 124,912          = 58.3%
  Markdown %  (148,000 - 124,912)/124,912  = 18.5%
  GMROI  72,886 / (55,016 / 2)             = 2.65
     (avg inventory at cost = what was invoiced,
      before the markdown-money recovery)

Walk through it. The ticket is the price printed on the tag, the MSRP, and buyers use the word constantly. Initial markup (IMU) is the margin on the ticket the day the goods land: 56.1% here.

But almost nothing sells entirely at ticket. In this season 58% of the units sold at full price, 27% at 30% off and 15% at half price. That drags the average unit retail (AUR, the average price actually collected per unit) down to $124.91, or 84% of ticket. Maintained markup (MMU) is the margin the buyer actually kept after those markdowns, and after everything the brand gave back.

Two of the giveback lines need naming, because you will meet them for the rest of your career. A cash discount is a percentage the buyer may subtract for paying quickly: here it is the 8% on the "8/10 EOM" line, explained in full later in this chapter. Markdown money (also called a markdown allowance or margin support) is a payment from you to the retailer to compensate them for the discounts they took on your unsold product. It almost always arrives as a subtraction from an invoice rather than as a bill you receive.

In this example MMU came out at 58.3%, higher than IMU, only because the brand handed over an 8% volume discount, the 8% cash discount and markdown money worth 5% of the invoice. Without that support the same selling pattern would have produced a materially worse number, and the buyer would not repeat the order.

GMROI and hearing the buyer's target

GMROI (gross margin return on inventory investment) is gross margin dollars divided by average inventory at cost: $2.65 of margin per dollar tied up. Buyers compare styles on GMROI because it captures speed as well as margin. A style with lower margin that sells out in six weeks can beat a high-margin style that sits for five months.

As a rough reading: a GMROI above 1.0 means the category at least earns back the money tied up in it, and above 2.0 counts as strong in specialty retail. Our 2.65 is a style a buyer repeats.

Learn to hear this in conversation. When a buyer says "I need a 60 maintained," they are quoting a margin target: after my expected markdowns, I must end the season keeping 60 cents of every retail dollar. Your job is to make that arithmetic work: through initial price, through markdown money, or through a product that sells. Your ERP should be able to produce a per-account, per-season maintained-markup estimate on demand, because that is the number the buyer will quote back at you when they decide whether to reorder.

Why wholesale margin is structurally thinner than DTC

Compare the same jacket sold two ways. The gap comes straight out of the arithmetic. When you sell wholesale you collect the wholesale price rather than the retail price, so the same garment produces roughly half the revenue, and every deduction then comes off that smaller number.

ONE JACKET, TWO CHANNELS  (per unit)

                          DTC        Wholesale     Wholesale
                        (own site)   (specialty)   (a major)
  Gross price            148.00        65.00         65.00
  Discounts + deductions -22.20         0.00        -15.07
  Net revenue            125.80        65.00         49.93
  Landed cost            -26.25       -26.25        -26.25
  ------------------------------------------------------------
  Gross margin            99.55        38.75         23.68
  Gross margin %          79.1%        59.6%         47.4%

  Payment processing      -3.95            -             -
  Consumer fulfilment     -9.00            -             -
  Returns allowance       -4.20            -             -
  Customer acquisition   -38.00            -             -
  Rep / showroom comm.        -        -6.50         -5.98
  Factoring commission        -        -1.30         -1.20
  Freight + 3PL + ticket      -        -1.85         -1.85
  Sample amortisation         -        -1.50         -1.50
  ------------------------------------------------------------
  CONTRIBUTION            44.40        27.60         13.16
  as % of the $148 ticket  30.0%        18.6%          8.9%

Two words in that table are worth pinning down first. Contribution is what one unit leaves behind after every cost that moves with the sale (the goods, the selling costs, the shipping) but before the fixed overheads of running the company, such as salaries and rent. Amortization means spreading a one-off cost over the units it helped produce: you spend $1,500 developing and sampling a style, you expect to sell 1,000 units, so you charge $1.50 to each unit.

Three cost lines in the wholesale columns also need naming:

• Rep / showroom commission is what you pay the people who sold the goods (an in-house sales representative, an outside sales agency, or a showroom that carries your line), normally a percentage of what they write.
• Ticketing is the labor of attaching the retailer's own price tickets to every garment before it ships.
• Factoring commission is the fee charged by the finance company that carries your invoices, and it gets a full explanation later in this chapter.

Reading the bottom line

Selling the jacket yourself leaves $44.40 a unit. Selling it to an independent boutique leaves $27.60. Selling it to a department store, after their discounts and deductions, leaves $13.16, under nine cents of every dollar the shopper spends.

It takes 3.4 wholesale units at a major to equal the contribution of one DTC unit. These are illustrative figures for one garment rather than industry averages. Treat the shape of the comparison as the lesson and expect your own cents to differ.

Brands do wholesale anyway because the DTC column carries one line that does not scale: customer acquisition cost, the advertising spend needed to find each buyer. It rises as you grow, because you exhaust the cheap audience first. The wholesale column has no such line.

One buyer meeting can produce a 400-unit order with zero media spend, and a store puts your product in front of people who will never see your ads. Wholesale buys reach and volume at the price of margin. That is the trade, and it is why the strategic question is always "what mix, and can my cash survive it?"

Hold onto two facts, because every structural problem in the rest of this chapter descends from them. Thin margin per unit is only survivable at volume. And volume in wholesale is bought months before it is paid for.

The seasonal calendar

Apparel runs on seasons instead of on the calendar year, and the seasons run about six months ahead of the weather. When a buyer sits down with your Fall linesheet it is February, it is cold outside, and they are ordering coats they will receive in July and put on the sales floor in August.

How many seasons?

Two majors and up to two or three minors is the standard shape for a small-to-mid brand:

Very large brands add monthly drops on top (small collections released on a fixed date, often announced days ahead), and streetwear brands may run almost entirely on drops. But the two-major backbone is what the wholesale trade calendar, the trade shows and the retailers' buying budgets are organized around, so it is what your software must model first.

Treat the months above as the industry's center of gravity. Individual brands shift their windows by a few weeks to hit a particular retailer's buying period, and the ranges published by the trade vary by a month at the edges.

Month by month, one Fall/Winter season

FALL/WINTER 2027  -  the 19 months of one season

  2026 JUN  Concept, trend, colour palette locked
       JUL  Fabric sourcing; mill minimums checked
       AUG  Sketches to tech pack; first fabric orders
       SEP  Proto samples round 1 from factory
       OCT  Proto round 2; fit sessions; costing round 1
       NOV  Line review: cut styles, lock the assortment
       DEC  Salesman samples (SMS) ordered from factory
                    *** sample money leaves here ***
  2027 JAN  SMS arrive; photography; LINESHEET PUBLISHED
                    *** prices must be final NOW ***
       JAN  Market opens. Appointments + trade shows.
       FEB  Peak order writing. Majors place prebook.
       MAR  ORDER BOOK CLOSES (no new prebook accepted)
       MAR  Bulk POs issued to factory. 30% deposit paid.
                    *** first big cash out ***
       APR  Fabric in-house at factory; production begins
       MAY  In-line inspection; trims land
       JUN  Final inspection; balance 70% paid; vessel sails
                    *** second big cash out ***
       JUL  Arrival, customs entry, duty + freight paid
       JUL  Receive at warehouse; ticket; pick, pack
       JUL  First deliveries ship. INVOICES ISSUED.
       AUG  Balance of deliveries ship
       SEP  Net-30 accounts pay
       OCT  Net-60 accounts pay
                    *** first cash in, 7 months after
                        the first cash out ***
       OCT  Retailer markdowns begin on slow sellers
       NOV  Markdown money and RTV claims arrive
       DEC  Off-price liquidation of your own overstock

Some vocabulary before you trace the money:

• Dated plans like this one are called a T&A or time and action calendar, the industry's name for a task list where every row has an owner and a date, and a slip on one row pushes every row below it.
• A mill is a fabric factory, separate from the garment factory that sews.
• A tech pack is the specification document you hand a factory: measurements, construction, stitching, materials, labels.
• A proto (prototype) sample is the factory's first physical attempt at the garment, made to check the idea works.
• A line review is the meeting where the team cuts weak styles and fixes the assortment, the final list of styles and colors the season will actually offer.
• SMS stands for salesman samples: a small run of sellable-quality garments made purely so the sales team has something to show buyers, paid for by you and rarely resold at full value.
• Greige (say it "gray-zh") is undyed, unfinished fabric. Mills often hold greige and dye to order, which is why fabric lead times are long and why a color change late in the process costs so much.
• In-line inspection means checking quality while the sewing line is still running, so a fault can be caught before the whole run is wrong.
• And final inspection in June is normally done against an AQL (acceptable quality limit), which is a sampling standard: an inspector checks an agreed sample of the cartons and the whole lot is accepted or rejected according to how many faults turn up.
• The March line that reads "ORDER BOOK CLOSES" is the date after which you stop accepting new prebook orders for the season, because everything after it is too late to make.

Tracing the money against the clock

Sample costs go out in December 2026. The first bulk deposit goes out in March 2027. The balance goes out in June. Duty and freight go out in July. And the first customer dollar arrives in September — nine months after the first real spend, and six months after the deposit. That distance is the entire problem, and no amount of profitability closes it.

Notice also the hard gate in January: the linesheet publishes with final prices. From that moment your wholesale price is a public commitment to every buyer, but your bulk fabric price is not yet fixed and your duty rate may change before the goods land. Brands absorb that risk. Some now write a tariff-adjustment clause into their terms and conditions, which is exactly what you would expect after 2025 and 2026. Most eat it.

Market weeks and trade shows

Market week is a concentrated period when buyers travel to a city and see many brands back to back, either in permanent showrooms (a brand's or an agency's sales office) or at a trade show (a convention hall of booths). New York, Los Angeles, Dallas, Atlanta, Las Vegas, Paris, Copenhagen and Florence all run them. In the US the biggest cluster is the group of Las Vegas shows that run in the same week (MAGIC, PROJECT, SOURCING at MAGIC and OFFPRICE), with a smaller New York edition as well.

The rhythm matters more than any particular date: the February round sells Fall, the August round sells Spring, and the September New York round catches late Spring and immediate stock. Exact dates move every year, so look them up on the organizer's calendar when you plan the season and store them on the season record rather than in somebody's head.

Trade shows are a real cost line: booth, build, shipping samples, travel, staff, and often a sales agency's cut on anything written there. Budget them per season and hold them against the orders written at that show. Whether a booth that produces less than its own cost in gross margin is worth repeating is a call you should be able to make from a report.

Order types and why they behave differently

Four kinds of order flow through a wholesale brand. They look similar on paper (customer, SKUs, quantities, price), and they carry completely different risk.

Prebook / futures

The buyer commits months ahead against goods that do not exist. The PO carries a start ship date and a cancel date, together the ship window or cancel window. If you cannot deliver inside that window the retailer may cancel with no penalty to them, and you are left holding branded inventory nobody ordered.

Prebook is the good kind of risk. You produce against demand you already know. The retailer takes the fashion risk. You take the execution risk. A brand whose book is mostly prebook can size production almost exactly and carry very little dead stock. This is why brands push hard for early commitment and why cancel dates are enforced ruthlessly by retailers, because the whole value of a prebook to them is that it arrives on their floor plan on time.

At-once / immediates

The buyer orders from stock you already own and expects it to ship in days. At-once is high margin, requires no forecasting from the customer, and delights the buyer. It is also entirely your risk: you speculated on that inventory. Immediates typically come from three sources:

• overproduction on a prebook style,
• a deliberate stock buy,
• or canceled orders you are trying to place elsewhere.

The operational difference from prebook is enormous. At-once must check ATS (available-to-sell: units on hand minus what is already promised to someone else, defined in chapter B and cached in chapter 8) in real time and reserve stock at order entry.

Prebook must not touch ATS at all at entry, because there is no stock yet. It draws down a production plan instead of an inventory pool. If your software treats them the same you will either oversell your warehouse or block prebooks you can absolutely fulfil.

Replenishment / basics programs

A never-out-of-stock program: a core white tee, a signature denim, a house cap. The retailer reorders continuously, often automatically, and expects the SKU to be available every week of the year. Margins are usually a little lower, because basics compete on price, but the volume is steady, the forecasting is statistical rather than a guess about trend, and the styles amortize their development cost over years.

Basics change the inventory model. Instead of "make 1,000, sell 1,000, done," you run a rolling stock position with a reorder point and a lead time. Chapter 1's ledger handles the movements. What the business needs on top is a coverage view: at the current weekly rate, how many weeks of supply do I have, and when must I place the next cut ticket to avoid running out given my supplier's lead time?

That lead time varies widely by factory, country and fabric, so measure your own rather than assuming a number. The reorder point is the stock level at which you must place that next order to arrive before you run dry.

Special makes and private label

Special make-up (SMU) is your style produced in an exclusive color or fabric for one account. Private label is you manufacturing under the retailer's own brand name. Both are made to order for a single customer, usually at a lower price in exchange for volume and certainty.

The risk splits two ways, and both ends are extreme. On one side there is no fashion risk at all, because the goods are sold before they are made. On the other, concentration risk is total: if that customer cancels, walks away, or claims a quality fault, there is no secondary market for 4,000 units with somebody else's brand in the neck. Special makes also demand real capability from your systems: customer-specific SKUs, customer-specific packaging and labeling, customer-specific pricing, and often customer-specific compliance documents.

Payment terms as a business weapon

Terms decide when you get paid, and in a business where profit is 9% of the ticket, timing is worth more than price. Here are the ones you will meet.

Net 30 / net 60 / net 90. Pay the full invoice within that many days of the invoice date. Small specialty stores are commonly net 30. Department stores and chains push for net 60 and beyond. Each extra 30 days is not free to you: on a $900,000 season, assuming money costs you 12% a year, 30 extra days costs about $8,900 in financing.

8/10 EOM. Long the standard department store term in apparel, and a trap for beginners. It means: take an 8% discount if you pay within 10 days of the end of the month in which the invoice was dated. A convention applies: invoices dated on roughly the 25th or 26th and later are treated as belonging to the following month, giving the buyer an extra cycle. In practice large retailers take the discount as a matter of policy whether or not they pay on time, and disputing it costs you the account.

WHAT A CASH DISCOUNT REALLY COSTS  (season = $900,000)

  Granting 30 extra days (net 30 -> net 60)
     900,000 x 12% x 30/365                 =  $8,877

  Granting an 8% cash discount
     900,000 x 8%                           = $72,000

  Annualised cost of the discount:
     8/(100-8) x 365/50 days pulled forward =   63.5%

  For comparison:
     2/10 net 30  -> 2/98 x 365/20          =   37.2%
     3/10 net 60  -> 3/97 x 365/50          =   22.6%

Read the middle number. Offering 8/10 EOM to pull cash forward by about 50 days is equivalent to borrowing at 63.5% a year. It is one of the most expensive forms of finance in existence, and it is a standard term in department store apparel. The comparison at the bottom shows why: a 2/10 net 30 is also expensive at 37% annualised, and a 3/10 net 60 is a comparatively sane 23%. If you must give a discount, give it against a longer stretch of days.

Dating, anticipation and deposits

Dating. An agreement that the clock starts later than the invoice: "net 30, December 1 dating" on goods shipped in September means nothing is due until 30 days after 1 December. Retailers ask for dating to get seasonal goods in early without paying early. It is a genuine concession you can trade: give dating, take a larger order or an earlier commitment.

Anticipation. An extra discount a retailer deducts for paying before the due date, computed at an agreed annual interest rate and scaled down to the days saved (an annual rate of 12% over 15 early days is 12% × 15 ÷ 365, so about 0.49% of the invoice). It is a legacy department store practice, is often taken unilaterally, and should be negotiated explicitly or excluded in your terms and conditions. Whatever you agree, your system must be able to recognize a deduction line labeled "anticipation" and decide whether it was contractually allowed.

Deposits. For new accounts, unrated accounts (retailers with no published credit rating from a credit agency), international accounts and special makes, ask for money up front, commonly 30% or 50% at order confirmation with the balance before or on shipment.

This is the most under-used lever a small brand has, because a deposit converts a receivable into cash six months early at zero financing cost. Buyers who genuinely want the goods will pay it. Buyers who will not are exactly the ones most likely to cancel.

Letters of credit

Letters of credit. A bank guarantees payment against documents. The buyer's bank promises to pay you when you present, say, the bill of lading (the carrier's document proving the goods were shipped, often written B/L), the packing list and the inspection certificate.

Letters of credit, or LCs, are standard on large export orders and with unfamiliar international accounts. They are slow and they cost fees on both sides, and they are document-fussy: a mismatched spelling can cause the bank to reject the presentation. But they turn a stranger's promise into a bank's promise.

The cash conversion cycle: how profitable brands die

The cash conversion cycle (CCC) is how many days your money sits in inventory and receivables before it comes back. The formula is standard across industries:

CCC = DIO + DSO - DPO

  DIO  Days Inventory Outstanding
       how long goods sit before they ship
  DSO  Days Sales Outstanding
       how long invoices sit before they are paid
  DPO  Days Payable Outstanding
       how long you sit on your own supplier bills

Typical wholesale apparel brand:

  DIO   goods paid for in June, shipped July-Aug     ~ 45- 90
  DSO   net 30 to net 60, plus late payers           ~ 45- 75
  DPO   30% deposit + 70% before shipment            ~   0- 15
                                                     ----------
  CCC                                                ~ 90-150 days

Three abbreviations, one idea:

• DIO (days inventory outstanding) is how long your money sits as goods.
• DSO (days sales outstanding) is how long it sits as an unpaid invoice.
• DPO (days payable outstanding) is how long you get to sit on somebody else's money before paying them.

Add the first two, subtract the third, and you have the number of days your own cash is locked up. The ranges above are a realistic shape for a small wholesale brand rather than a published benchmark; measure yours from your own invoices and payments as soon as you have a season of history.

The killer is DPO. In most industries you buy on terms: you get the goods, you pay in 60 days, and your supplier funds your inventory. Apparel factories generally do not work that way. The common structure is 30% deposit on order confirmation and 70% before shipment, with new or small brands frequently asked for 50% down.

You pay for goods before you own them, then wait 45–75 days after shipping to be paid. Your DPO is effectively zero, and in cash terms worse than zero, because much of the money leaves before the goods exist.

That is why the apparel cash cycle is so much harsher than the textbook picture in which supplier credit quietly funds your stock. In apparel, you fund it.

A worked example: a profitable brand runs out of money

Meet Meridian, a contemporary womenswear brand. Team of six. Two seasons a year. Both seasons this year are profitable. Watch the bank account. The company is invented. The arithmetic is the part to take seriously.

MERIDIAN  -  SEASON P&L, both seasons profitable

                             FW26           SS27
  Net wholesale booked      900,000      1,150,000
  Post-ship deductions -9%  -81,000       -103,500
                          ---------      ---------
  Net realised revenue      819,000      1,046,500
  Landed COGS              -362,000       -462,000
                          ---------      ---------
  Gross profit              457,000        584,500
    gross margin %            55.8%          55.9%

  Rep / showroom commission -63,000        -80,500
  Factoring commission      -18,000        -23,000
  Outbound freight + 3PL    -27,300        -34,860
  Samples, booth, travel    -84,000        -86,000
                          ---------      ---------
  Season contribution       264,700        360,140

  Combined contribution                    624,840
  Annual overhead (6 people, rent, tools) -456,000
                                         ---------
  OPERATING PROFIT                         168,840

That is a real business. Two profitable seasons, 55.8% and 55.9% gross margin, $168,840 of operating profit on about $1.87M of realized revenue, roughly a 9% operating margin, which is respectable for a small wholesale brand. Now the same year in cash.

MERIDIAN  -  CASH, JANUARY TO DECEMBER

Mo       CASH IN   FACTORY   DUTY+FR   MKT+SMP      OPEX    BALANCE
-------------------------------------------------------------------
Open           -         -         -         -         -    180,000
Jan            -         -    58,000    34,000    38,000     50,000
Feb      200,000         -         -    46,000    38,000    166,000
Mar      240,000    87,000         -         -    38,000    281,000
Apr      176,500         -         -         -    38,000    419,500
May            -         -         -         -    38,000    381,500
Jun            -   203,000         -         -    38,000    140,500
Jul            -         -    72,000    36,000    38,000     -5,500
Aug            -         -         -    48,000    38,000    -91,500
Sep      328,000         -         -         -    38,000    198,500
Oct      410,000   111,000         -         -    38,000    459,500
Nov            -   259,000         -         -    38,000    162,500
Dec            -         -    92,000         -    38,000     32,500
-------------------------------------------------------------------
TOTAL  1,354,500   660,000   222,000   164,000   456,000     32,500

Where the trough falls

Read the BALANCE column. January opens with $180,000. Spring receivables land in February, March and April and the balance peaks at $419,500. This is the moment a founder thinks the company is doing well and hires someone.

Then May through August, nothing comes in. Spring's money is collected. Fall's goods have not shipped yet, so there is nothing to invoice. Meanwhile the Fall factory balance of $203,000 falls due in June, $72,000 of duty and freight in July, and $84,000 of sampling and trade show cost across July and August.

The account goes negative in July and bottoms at −$91,500 in August. Payroll does not clear.

And notice what happens after the rescue. Fall collections arrive in September and October and the balance recovers to $459,500. But Spring 2027 is a bigger season, so its deposit ($111,000), factory balance ($259,000) and duty ($92,000) are due in October, November and December. The year ends with $32,500 in the bank, having started with $180,000, in a year that earned $168,840 of profit.

The levers, and which ones actually work

We re-ran Meridian's twelve months changing one thing at a time.

Three lessons fall out. First, order discipline (deliberately writing less business than you could) improves the year end by $100,000 but does nothing for the August trough, because the trough is caused by the previous season's timing. Turning growth down is a medium-term fix for a short-term crisis.

Second, renegotiating factory terms looks like the obvious answer and is worth $296,000 at year end, because the Spring balance falls outside the year entirely. But in this case it simply relocated the Fall payment into another dead month and left the trough exactly where it was. Timing beats size.

Third, only levers that pull receivable cash forward actually close the gap.

Deposits and factoring

Deposits are free money, and you should ask for them from every new, unrated, international or special-make account. In Meridian's case, deposits on about a third of the Fall book would have closed the trough entirely.

Factoring deserves a proper explanation because it has long been the financial backbone of the apparel trade, and the apparel version of it carries services you will not find in most other industries. A factor is a finance company that buys your receivables: it takes over the invoices your customers have not yet paid.

Two separate services are bundled under the word, and beginners conflate them:

• Credit protection and collection. Before you ship, you submit the order for credit approval. The factor checks the retailer's creditworthiness. Approval depends on your customer's credit, not yours, which is why emerging brands can use it. The factor also becomes the party the retailer pays, and chases the money. You pay a commission calculated on the face value of each invoice. Whether the factor absorbs a bad debt depends on one word in the contract: non-recourse means the factor eats the loss if an approved customer fails to pay for credit reasons, while recourse means the factor can hand the unpaid invoice back to you and take its money back. Recourse is cheaper and protects you much less.
• The advance. Separately, the factor will lend you a percentage of the invoice immediately rather than paying you at maturity, charging interest on the advanced money on top of the commission. Commission rates, advance percentages and interest are all negotiated and vary with your volume and your customer mix, so get all three in writing and model them separately, because a headline commission means nothing without the other two numbers.

What the advance actually buys

In our model Meridian was already paying a factoring commission but collecting at maturity, which is credit protection only. Turning on the advance eliminates the August trough completely. The lever table ignores the interest the factor would charge on the advanced money, which would shave the year-end balance. Include it when you model your own. That is the trade: a few points of margin in exchange for not going bust in the summer.

The general rule is worth memorizing: growth consumes cash in direct proportion to how far ahead of collection you must spend. In wholesale apparel that distance is six to nine months. Model it explicitly, per season, or you will discover it at the bank.

Channel structure and channel conflict

A brand sells through channels, and channels fight. A channel is a route to market: department stores, boutiques, off-price, marketplaces, your own website. Each one has its own economics, its own rules and its own opinion of what the others are doing.

Majors / department stores and national chains

Nordstrom, Macy's, Saks, Bloomingdale's, and the big specialty chains. Enormous volume, prestige, and door count. They also bring a rulebook. A vendor manual and its routing guide spell out exactly how you must pack, label, document and deliver, and every rule has a price attached. Four terms you need before that sentence means anything:

• EDI is electronic data interchange, a family of standard machine-readable message formats that big retailers use instead of email: the order, the ship notice and the invoice all travel as structured files. For most majors it is the only accepted channel. Chapter 4 covers the mechanics.
• ASN is advance ship notice. It is the electronic "here is exactly what is on this truck, carton by carton" message you must transmit before the freight arrives, so the receiving dock can scan it in rather than open it.
• MABD is must-arrive-by date, the day the goods must be at the retailer's door, which is not the day you ship. Miss it and the delivery may be refused or fined.
• Floor-ready means the goods arrive already ticketed, hung and labeled to the retailer's spec, so store staff can move them straight onto the sales floor without touching them.

Chargebacks and the routing guide

Break any of these and you get a chargeback: money the retailer subtracts from your invoice as a penalty. Published summaries of retailer routing guides put typical penalties, as of 2026, at:

• $200–$500 and up per purchase order for missing the must-arrive-by date,
• $150–$300 per shipment for using a carrier the retailer did not approve,
• $100–$500 per shipment for ASN errors,
• and $50–$150 per carton or pallet for label problems.

The largest chains go further. Walmart's OTIF ("on time, in full") program is commonly summarized as requiring about 98% on-time delivery on shipments Walmart collects, 90% on shipments you prepay and 95% delivered in full, with a fine of 3% of the cost of the goods on non-compliant cases. Target runs an equivalent program, also summarized at 3% of cost with a minimum charge of about $150 per occurrence, measured quarterly rather than on Walmart's rolling basis.

Treat every one of those figures as indicative and read your own signed supplier agreement, because the retailer changes them and the agreement is what binds you.

Majors also expect three kinds of financial support:

• markdown money (defined earlier: your contribution toward the discounts they take on your unsold goods),
• co-op advertising (short for cooperative advertising: you pay a share of the retailer's cost of advertising your product, usually as a percentage of the season's invoicing),
• and RTV rights (return to vendor): the right to ship unsold goods back to you and deduct the value from what they owe.

Specialty boutiques

Independent stores buying 6–36 units of a style. Higher effective margin because they rarely extract discounts or deductions, they usually pay net 30, and they impose fewer compliance requirements. Lower volume per account, higher account count, more credit risk per dollar, and far more administrative overhead, because 200 boutique accounts generate 200 relationships.

Published guidance puts typical minimum order quantities per style for boutique buying at around 5–12 units for emerging brands, 12–24 for established ones and 24–50 as you scale, though every brand sets its own.

Off-price

TJ Maxx, Marshalls, Ross, Nordstrom Rack, Burlington and the closeout trade: dealers who buy leftover stock in bulk, cheaply, to resell. They buy opportunistically at deep discounts and resell branded goods well below department store pricing.

For a brand, off-price is the pressure valve: it converts dead inventory into cash quickly. It is also dangerous. Ship there too often, at too shallow a discount, or too early in the season, and your full-price accounts see your product at 60% off two blocks away and stop buying. Use off-price for genuine surplus, and keep it out of your sales budget.

Marketplaces and B2B platforms

Wholesale marketplaces (Faire and similar) and B2B order platforms (JOOR, NuORDER and similar) put your linesheet in front of thousands of buyers. "B2B" is business-to-business: software for selling to shops rather than to shoppers. The trade is commission and control.

Marketplaces generally charge a percentage of each order, with a higher rate (and often a one-off fee) on a retailer the platform introduced to you, and a reduced or zero rate on retailers you brought yourself. Published rates change often, so read the platform's current brand pricing page before you sign. A number printed in a book will already be stale.

Then model whatever you agree as a channel-level cost rather than an ad-hoc discount typed onto an order, because the arithmetic is unforgiving. Work an example: a 15% commission against a 55% gross margin takes more than a quarter of your gross profit.

Distributors and agents in other territories

Two different arrangements that beginners confuse. An agent sells on your behalf and takes a commission (commonly quoted around 10–15% in apparel, though it is negotiated), while the retailer remains your customer and your credit risk. A distributor buys from you at a distributor price, then resells to retailers in their territory at their own wholesale price.

The distributor takes the credit risk, the inventory risk and the local compliance burden. You get one customer, one currency, one shipment, and a much lower price. Distributor pricing is usually a further discount off your wholesale, because the distributor needs their own markup to fund local warehousing, duties and selling.

Your own DTC

Your site is a channel competing with your customers. Every promotion you run is a promotion your retail partners cannot match on the same goods. This is the sharpest conflict in the model, and it has two dimensions:

• Price. If you discount before your retailers do, you have devalued their inventory and they will ask you to pay for it.
• Timing. If your site launches Fall in July while your wholesale accounts receive in August, you have taken the first sale of every full-price week.

Both dimensions are schedule problems, so solve them with data. Store a per-channel launch date and a per-channel promotional calendar on the season record, and make the system refuse a DTC discount on a style whose wholesale accounts are still inside their full-price window. A rule the software enforces survives the week somebody wants a quick sale.

MAP policy

MAP is minimum advertised price: a policy stating the lowest price at which a reseller may advertise your product. Note the word advertise.

Under US law the distinction matters. Telling a retailer what price they must sell at is called resale price maintenance. Until 2007 US federal courts treated that as automatically illegal. The Supreme Court's decision in Leegin Creative Leather Products v. PSKS changed the test to a case-by-case weighing of actual competitive effects, so it is no longer automatically unlawful at the federal level. But it remains legally risky, and several US states are stricter under their own laws.

A MAP policy sidesteps the question by being unilateral: you announce the rule, you do not negotiate it, you do not ask anyone to agree to it, and the only enforcement is your own decision to stop selling to a violator. Practitioner guidance is consistent on the requirements: draft it as a policy rather than a contract, apply it to everyone the same way, and keep timestamped evidence of violations.

Note also that resale price maintenance is treated as a "hardcore" restriction of competition in the UK and EU, meaning regulators there presume it is unlawful and have fined brands for it, so a US-style MAP program does not travel. Take actual legal advice before you launch one. This chapter is about software and offers no legal opinion.

Sell-through economics from the buyer's side

You cannot sell to a buyer you do not understand. The buyer is trying to hit a margin plan with a fixed budget, and the clothes are the means to it.

Open-to-buy

Open-to-buy, usually shortened to OTB, is how much money a buyer may still commit for a period: the budget they have left to spend on new stock. The standard formulation is:

Planned receipts = planned sales
                 + planned markdowns
                 + planned end-of-period inventory
                 - beginning-of-period inventory

OTB = planned receipts - merchandise already on order

Example
  Planned receipts for the period      650,000
  Already committed on open POs        250,000
                                      --------
  OPEN-TO-BUY                          400,000

Read the top formula as a budget balance. The buyer knows what they plan to sell, what they expect to lose to markdowns, what inventory they want left at the end, and what they are starting with. Sales and markdowns both drain the shelf, so both increase the amount that has to come in, and the stock already on hand reduces it. The result is "planned receipts". Subtract what is already ordered and you get the money that is genuinely available.

When a buyer says "I'm out of open-to-buy," the arithmetic has no room left in it, whatever they think of your product. The useful response is to ask for a later delivery that falls into the next OTB period, or to ask what they would cancel to make room.

Sell-through

Sell-through rate is units sold divided by the units you had available at the start of the period. For a seasonal buy, that starting figure is the units received.

Retail-planning guidance gives a shape rather than a single number, because the right target depends on category, price point and season length:

• for a 12–14 week season, roughly 40–50% full-price sell-through by week six counts as on plan,
• the end-of-season goal is around 80% or better,
• and falling below about 40% at week six is the trigger for a markdown review.

Buyers watch this weekly. If your style is lagging in week three, it will be marked down in week six and you will hear about it. If it is running hot, you will get a reorder call, and whether you can say yes depends entirely on whether you have at-once stock, which is a decision you made nine months earlier.

Weeks of supply is on-hand units divided by average weekly unit sales. It is the same fact as sell-through, expressed in weeks instead of a percentage. For a replenishment basic, the target is normally set to cover the supplier's lead time plus a safety buffer (spare stock held to absorb a late delivery or a sudden spike in demand), so that you can reorder before you run dry.

For a seasonal style, weeks of supply should be heading toward zero as the season ends, because leftover seasonal stock is a markdown waiting to happen.

End of season: markdown money, RTV, and liquidation

When the season ends and inventory remains, three mechanisms transfer the pain:

• Markdown money (also markdown allowance or margin support) is a payment from you to the retailer to offset the discounts they take on your unsold product. It arrives as a deduction from an invoice. You never receive a bill for it. It is negotiated, it is often assumed, and for department store business it can run to several percent of the season's net invoicing.
• RTV is return to vendor. The retailer physically ships unsold goods back and deducts the value. Whether you have granted RTV rights is a contract question and should be explicit in your terms. Unmanaged RTV is brutal: you get back out-of-season goods, often with retailer price tickets on them, and you pay the freight.
• Off-price liquidation. Either the retailer jobs the goods out, meaning they sell them in bulk to a closeout buyer, or you buy them back and job them yourself. Off-price buyers pay a fraction of wholesale. Whatever you recover is cash. Treat none of it as profit.

Run the bad-season version of our earlier buyer arithmetic and you see why buyers are conservative. If W-4210 had sold only 30% at full price, 25% at −30%, 25% at −50% and 20% jobbed at −75%, the buyer's revenue on 1,000 units drops from $124,912 to $96,200, gross margin falls from 58.3% to 45.9%, and GMROI falls from 2.65 to 1.61. That is the difference between "reorder it" and "we're not carrying this brand next season."

All of which is why sell-through data is worth chasing. Retailers can send you point-of-sale data, the record of what actually rang up at the till, by SKU, by store, by week. It comes either over EDI (message type 852, "product activity data") or through a vendor portal.

Get it, ingest it, and build a weekly sell-through report per account. It tells you what to reorder, what to cut, and what markdown money you are about to be asked for, weeks before the buyer calls. Chapter 4 covers the mechanics. This is the reason to bother.

Unit economics: one style, end to end

Here is W-4210 all the way through: the full waterfall from list price down to contribution, for 1,000 units sold to a department store.

STYLE W-4210  -  1,000 UNITS TO A MAJOR
                                     amount      running
  Gross wholesale 1,000 x $65        65,000       65,000
  Volume discount 8%                 -5,200       59,800  <- net inv
  Cash discount 8/10 EOM             -4,784       55,016
  Markdown money 5% of net invoice   -2,990       52,026
  Compliance chargebacks 1.5%          -897       51,129
  RTV / returns 2%                   -1,196       49,933
  ---------------------------------------------------------
  NET REALISED REVENUE               49,933    = $49.93/u
                                               = 76.8% of list

  Landed COGS 1,000 x $26.25        -26,250
  ---------------------------------------------------------
  GROSS PROFIT                       23,683       47.4%

  Rep / showroom commission 10%      -5,980       17,703
  Factoring commission 2%            -1,196       16,507
  Outbound freight not recovered       -750       15,757
  3PL pick / pack / ticket $1.10/u   -1,100       14,657
  Sample + development amortised     -1,500       13,157
  ---------------------------------------------------------
  CONTRIBUTION                       13,157    = $13.16/u
                                               = 20.2% of list
                                               =  8.9% of MSRP

Every line between "gross wholesale" and "net realized revenue" is a deduction that arrives after the sale is booked, and most of them arrive as unexplained subtractions from a check.

The style was priced at a 59.6% gross margin. It delivered a 47.4% gross margin. After selling costs it contributed $13.16 a unit, or 8.9% of what the shopper paid. The same style sold to specialty stores, with no volume discount, no cash discount, no markdown money and no chargebacks, contributes $27.60 a unit, 18.6% of the ticket: more than twice as much on the identical garment.

The percentages assumed for each deduction here are illustrative. The point is the structure, and the fact that you will only ever know your own numbers if you record every deduction with a reason code.

What this means for your ERP

Everything above is now a specification. Here is the translation into tables, fields, constraints and rules.

Three prices, three fields, never derived at read time

A style needs its price points stored separately, versioned by season and effective date, and never recomputed on the fly from a multiplier. Multipliers are how you propose a price. The agreed price is a fact you must be able to reproduce years later for an audit or a dispute.

create table style_price (
  id              uuid primary key default gen_random_uuid(),
  tenant_id       uuid not null,
  style_id        uuid not null references style(id),
  season_id       uuid not null references season(id),
  currency        char(3) not null default 'USD',
  fob_cost        numeric(12,4) not null,
  duty_rate       numeric(6,4)  not null,   -- as of costing date
  freight_alloc   numeric(12,4) not null,
  landed_cost     numeric(12,4) not null,   -- stored, not derived
  wholesale_price numeric(12,2) not null,
  msrp            numeric(12,2) not null,
  map_price       numeric(12,2),
  effective_from  date not null,
  effective_to    date,
  created_at      timestamptz not null default now(),
  constraint chk_margin_floor
    check (wholesale_price >= landed_cost * 1.8),
  constraint chk_msrp_above_wsp
    check (msrp >= wholesale_price)
);
create unique index style_price_active
  on style_price (tenant_id, style_id, season_id, effective_from);

Take the block line by line:

• uuid is a long random identifier, used instead of a counting number so that two systems can generate ids without colliding.
• numeric(12,4) is an exact decimal with four places after the point, which is what you use for money: the alternative, floating-point, introduces tiny rounding errors that accumulate into unexplainable pennies.
• effective_from and effective_to make each row valid for a date range, so old prices stay readable instead of being overwritten. That is what "effective-dated" means throughout this chapter.
• A check constraint is a rule the database itself refuses to break.
• The create unique index line stops two rows claiming the same style, season and start date.

The design choices matter too. landed_cost is stored rather than computed because the duty rate that applied when you costed the style is a historical fact and duty rates change, sometimes several times in a year, as 2025 and 2026 demonstrated. duty_rate is captured alongside it so you can explain the number later.

The check constraint enforcing a wholesale floor above landed cost is a cheap way to stop somebody publishing a linesheet at a loss. Set the multiple to whatever your business actually requires and make it configurable per segment.

tenant_id marks which brand each row belongs to (a tenant is one customer of your software) and connects to chapter 5's row-level security, the database feature that keeps one brand's rows invisible to another.

Note that this table lives at style level rather than SKU level: sizes rarely price differently, and when they do (extended sizing surcharges) you want an explicit override table instead of exploding the price rows.

Season as a first-class dimension

You need a season table with a code (FW26, SS27), a type, and the calendar gates: linesheet_publish_date, order_open_date, order_close_date, production_book_by, ship_window_start, ship_window_end. Season must be a foreign key on styles, prices, customer orders, purchase orders and inventory lots. A foreign key is a column that points at a row in another table and that the database refuses to leave dangling.

Every list screen needs a season filter, and the default must be "the season currently being sold," which the system works out from the calendar gates. Today's date alone will not give you it. Reports that aggregate across seasons must say so explicitly.

Add the retail calendar too. Retailers plan on the National Retail Federation's 4-5-4 calendar, which divides each quarter into 4-week, 5-week and 4-week "months" so that comparable periods contain the same number of weekends, and inserts a 53rd week every five or six years: 2006, 2012, 2017 and 2023 were the recent 53-week years. If you report weekly sell-through against an ordinary calendar and your buyer reports it against 4-5-4, your numbers will never reconcile. Store a retail_calendar table mapping each date to fiscal year, fiscal month and fiscal week.

Order types are different objects with different rules

One customer_order table with an order_type enum (a column restricted to a fixed list of allowed values, here prebook, at_once, replenishment and smu), and behavior that branches on it:

• prebook: requires ship_window_start and cancel_date, both not null. Must not consume ATS at entry. Must allocate against a production plan. Must be blocked from entry after season.order_close_date without an override with a reason code.
• at_once: must reserve from ATS at entry inside the same transaction that checks availability. This is exactly the concurrency and idempotency problem of chapter 3 (idempotency meaning that processing the same message twice must not create two orders), and getting it wrong means overselling the warehouse. Must be blocked if ATS is insufficient, with a partial-fill option.
• replenishment: attaches to a program record with a reorder point, target weeks of supply and a supplier lead time. It generates suggested cut tickets when coverage drops below the point.
• smu: carries a customer_id on the style itself and must be excluded from general ATS entirely, because those units belong to one account.

Hard rules to enforce in the database, not the UI: a prebook line's ship window must fall inside its season's ship window; cancel_date >= ship_window_start; the sum of allocated quantity across orders for a production lot cannot exceed the lot quantity. Chapter 1's append-only ledger is where allocations live: you record an allocation as a movement and never overwrite it in place, so the history of who was promised what survives.

Terms, deductions, and the money that never arrives

This is the part most homegrown systems get wrong, and it is where the business bleeds. You need:

• A payment_terms table describing terms structurally, not as free text: net_days, discount_pct, discount_days, eom boolean, eom_cutoff_day (for the end-of-month convention described earlier), anticipation_allowed boolean, dating_date. From these you compute a real due_date and a real discount_due_date per invoice.
• A deduction table with a typed reason_code (a short fixed label saying why money was withheld, chosen from a list rather than typed as free text) such as cash_discount, volume_discount, markdown_allowance, coop_advertising, compliance_chargeback, rtv, shortage, anticipation, freight_allowance and unknown, linked to the invoice and, where possible, to the specific PO or carton. Every deduction needs a disputed flag, a dispute_opened_at, and a resolution.
• A rule engine, however simple, that flags deductions the customer was not entitled to: a cash discount taken after the discount due date; anticipation on an account where anticipation_allowed is false; markdown money above the agreed seasonal cap; a chargeback for a late delivery on a PO your own records show shipped inside the window.

The report every founder will demand within a month of going live is the net realized revenue waterfall. A waterfall report starts at the big headline number and shows each subtraction in order until you reach what is actually left: gross wholesale → discounts → deductions → net realized, sliceable by account, style, season and channel, with a per-account "deduction rate" ranking. It is the single most valuable screen in a wholesale ERP and it does not exist in most off-the-shelf systems.

Cash: the forecast is not optional

Because the P&L will not warn you, the ERP must. You need a cash forecast that projects forward from data you already hold:

• Inflows. For each open invoice, due_date adjusted by that customer's historical payment behavior (store a rolling avg_days_to_pay per customer), reduced by that customer's historical deduction rate. Plus deposits due on confirmed orders. Plus factor advances if you use them.
• Outflows. For each open purchase order, the deposit and balance milestones with their trigger dates; estimated duty and freight at the projected arrival date; operating expenses as a recurring schedule; season market and sample costs as planned commitments.

Render it as a 13-week and a 12-month rolling view with a minimum-balance alert. If the projected balance ever crosses zero, the system should say so loudly and name the month. That one report would have told Meridian in March that August was going to fail.

Supporting fields you need for this to work: on purchase_order, a deposit_pct, deposit_paid_at, balance_terms, ex_factory_date, eta_port, eta_warehouse; on customer, payment_terms_id, credit_limit, factor_approved_amount, avg_days_to_pay, deduction_rate_pct, deposit_required_pct.

Credit, factoring and the approval gate

If the brand factors, credit approval becomes a hard gate in the workflow: an order above a customer's approved amount must not ship until the factor raises that customer's approved credit limit, because shipping it converts a protected receivable into an unprotected one.

Model this as an order status (pending_credit → credit_approved → releasable) with the approved amount and approval date stored, and a nightly reconciliation against the factor's file. The picking screen must refuse to release an order sitting in pending_credit. This is also a natural integration point in chapter 4's terms: factors typically exchange approval and assignment files on a schedule.

Customs, duty and origin are their own tables

Put hts_code on the style and country_of_origin on the purchase order or the inventory lot, never on the style, because the same style gets made in two countries. Keep duty rates in their own effective-dated table keyed by HTS code, origin and date. Then record what actually happened at the border: the customs entry number, the value you declared, the duty you paid and the date, each linked to the receipt and to the commercial invoice that supports it.

That last paragraph sounds like bureaucracy until a court changes the rules. In 2026, after the Supreme Court struck down the IEEPA tariffs, importers spent the year filing to recover duties a court had held were unlawfully collected, and the ones who could file quickly were the ones whose systems could produce, per shipment, exactly what was paid and why. Build a duty_payment row per entry and a duty_refund_claim row that points at it, and you have turned a scramble into a query.

Store the ordinary savings too: if you use first sale valuation (a customs rule that can let you declare the price the middleman paid the factory rather than the higher price you paid the middleman, so duty is charged on the smaller number), you must hold both prices and the documents that prove the chain.

Product compliance data you must store

A garment sold in the US carries legally required information on its label: fiber content, country of origin, care instructions and the identity of the company responsible for it. That identity may appear as a company name or as an RN number. A registered identification number is an identifier the US Federal Trade Commission issues to a business so it can be used on the label in place of the full legal name.

Retailers then add their own requirements on top: their ticket format, their hanger spec, their carton markings. All of it belongs on the style record, versioned by season, because a label reprint is cheap and a truck refused at a distribution center is not.

Supply chain traceability is now a live requirement. US enforcement of the Uyghur Forced Labor Prevention Act kept tightening through 2026. That law presumes any goods made wholly or partly in China's Xinjiang region, or by a listed company, were made with forced labor, and it bars them from entry unless the importer proves otherwise.

US Customs and Border Protection detained $53.6 million of textile, apparel and footwear goods under that law in the first four months of 2026, against $23.2 million in the comparable earlier period, according to CBP data cited in the USFIA study. A June 2026 executive order then directed the agency to strengthen enforcement and importer disclosure further.

In schema terms that means a supplier chain you can walk: a row per node (sewing factory, mill, spinner, and the raw fiber where you can get it) attached to the production lot, each with documents and dates. A folder of PDFs on somebody's drive will not survive a detention.

Channels, MAP, and conflict prevention

Add a channel dimension to customers and orders (major, specialty, off_price, marketplace, distributor, dtc) and a price_list per channel per season so that distributor pricing, marketplace pricing and standard wholesale are separate, auditable rows rather than ad-hoc discounts typed into an order. Enforce that a style flagged off_price_eligible = false cannot be added to an order for an off-price customer, and that off-price orders require an approval role.

For MAP, store map_price with effective dates (it is on style_price above), plus a map_violation table recording observed price, retailer, URL, screenshot reference, observed timestamp, and the action taken. Because a defensible MAP policy must be unilateral and consistently enforced, the value of this table is evidentiary. Never build a workflow where a salesperson "agrees" a MAP exception with a retailer. Record it as a policy change with an effective date instead.

Compliance: chargebacks are a data problem

Retailer routing guides impose per-PO requirements:

• an approved carrier,
• a must-arrive-by date,
• carton labeling in a standard barcode format called GS1-128,
• an advance ship notice transmitted before the freight arrives,
• and a delivery appointment booked a day or two ahead.

Carton labels also carry an SSCC, or serial shipping container code, an 18-digit number that uniquely identifies one specific carton so the retailer can scan it against your ASN instead of opening it. Every one of those is a field you must store per customer and validate per shipment before the truck leaves.

Model a customer_routing_profile holding the requirements, and a pre-ship validation that blocks release when a rule is unmet. A blocked shipment costs an hour. A chargeback costs hundreds of dollars and cannot be reversed without a fight. Chapter 4 covers the EDI mechanics of the 850 (the order), the 856 (the ASN) and the 810 (the invoice). This chapter tells you why the 856 timing rule is worth engineering carefully.

Retail-side metrics you must be able to produce

Buyers will ask, and salespeople will ask on their behalf. Build these as reports:

• sell-through by SKU by account by week (fed by EDI 852 or portal exports);
• weeks of supply;
• season-to-date maintained markup estimate per account;
• GMROI estimate per style;
• and a reorder-opportunity list of styles selling above target rate where you hold at-once stock.

These connect directly to chapter 8's ATS caching and reporting, because they are all read-heavy aggregations over a high-write ledger.

Chapter connections

Field notes & further reading

• NRF 4-5-4 Retail Calendar. The National Retail Federation's official retail calendar, with downloadable three-year files. Explains why retailers count months in 4-5-4 weeks and when the 53rd week gets inserted. Use it to build your retail_calendar table so your weekly reports reconcile with your buyers'.
• Toolio — Fundamental Retail Math Formulas. Clean worked definitions of initial markup, maintained markup, sell-through, weeks of supply, GMROI, markdown percent and open-to-buy, each with a numeric example and a benchmark. This is the arithmetic your buyer is doing in their head, and the source of the sell-through and GMROI thresholds quoted above.
• The Wholesale Fashion Calendar — LINESHEET. Month-by-month order windows, linesheet drop dates and ship windows for SS, FW, Resort and Pre-Fall. The source for the calendar table in this chapter and a good sanity check when you are designing season records.
• Wholesale Pricing Guide — LINESHEET. Multiplier ranges by segment, minimum-order-quantity norms, and the cost-to-retail ratio rule of thumb quoted above. Useful for setting the margin-floor constraints in your price table.
• RetailerHub — What is a Routing Guide. A plain summary of what retailer routing guides mandate and typical chargeback amounts by violation type, plus the OTIF targets and penalty percentages for the big chains cited in this chapter. Read it before you design your pre-ship validation rules, then read your own signed vendor agreement, which is the document that actually governs.
• USFIA Fashion Industry Benchmarking Study. The annual survey of US fashion companies, free to download. Source of the tariff and sourcing figures in this chapter: the 21.6% average applied tariff on US apparel imports as of May 2026 against 15.2% in January 2025, the January–May 2026 country shares, and the forced-labor detention values. The 2026 edition also states that the detention figure rose 56.8%. The two dollar amounts it gives imply a much larger jump, so this chapter quotes only the amounts. Read the current edition each year, because these move fast.
• Harmonized Tariff Schedule of the United States (USITC). The official, searchable duty schedule. Chapters 61 (knitted apparel) and 62 (woven apparel) are yours. Look up your actual HTS codes rather than assuming a rate. This is the authoritative source your duty_rate field should trace back to.
• GS1 US — What is a GTIN. The standards body for barcodes. Explains GTIN-8/12/13/14, how UPCs relate, and how company prefixes are licensed and priced. Every major retailer requires GS1-issued numbers, and the same standards body publishes the GS1-128 carton label and the SSCC used on it.