Introducing Endless AI and Endless MCP. Operate from anywhere.

Most brands believe a wider range sells more. Past a certain point it sells about the same and costs more to carry. Each new SKU adds a minimum order quantity, a safety stock buffer, a forecast that starts with no history, and a share of the warehouse. The revenue it adds often comes from customers who would have bought a neighbouring SKU anyway.

Assortment planning is the set of decisions about which products to carry, how deep to buy each one, and where each one is sold. For a retailer it is a store-by-store exercise. For a brand selling through its own site, wholesale accounts, and retail partners, the questions are different, and they come back every season.

This framework covers the four decisions in order: breadth, depth, channel, and exit.

Decision 1: Breadth — should this product exist in the range?

Every proposed SKU, including a new colour or size, should pass three tests before it gets a buy.

  1. The incremental demand test. Estimate how much of the new SKU’s forecast is new demand and how much moves over from existing SKUs. A fourth scent in a line of three usually takes most of its sales from the other three. Look at what happened the last time you added a variant of the same type: did total category units go up, or did they spread across more SKUs?
  2. The minimums test. Divide the supplier’s minimum order quantity by the SKU’s expected weekly sales. If the first buy covers more than a season, the SKU starts life as overstock.
  3. The channel test. Name the channel the SKU is for. If the answer is “all of them,” check whether your retail accounts will list it. A SKU that only sells DTC still needs DTC-level depth and carries the same minimums.

A SKU that fails two of the three should not get a full buy. It can still launch as a limited run, a pre-order, or a DTC exclusive.

Decision 2: Depth — how much of each product to buy

Depth is where assortment planning meets cash. The same budget can buy 40 SKUs deep or 80 SKUs thin, and both choices have a cost.

A practical way to set depth is to group the range by role and give each role a different target.

RoleShare of SKUs (typical)Depth targetWhat a stockout costs
Core: steady sellers, carried every season20–30%High cover, rarely below safety stockLost sales and lost retail placement
Seasonal and trend40–60%Bought to sell through by season endMissed upside, but markdown risk if overbought
Test and new10–20%Thin first buy, fast reorder optionLittle; the buy exists to test demand
ExitWhatever is leftZero new buysNone

The shares above are a starting range, not a rule. Every SKU needs a role, and the role sets the buy. A common failure is buying test SKUs at core depth because the team is excited about them.

Depth also depends on sizes and variants. If a style comes in five sizes, the buy has to follow the size curve: the share of demand each size takes. Buying flat across sizes guarantees broken size runs in the middle sizes and leftover stock at the ends. Size curves should come from your own sell-through by channel, because a retail account’s curve often differs from your DTC customers’.

Decision 3: Channel — which products go where

Brands selling through several channels rarely need the same assortment in each.

  • DTC can carry the full range, including tests, limited runs, and long-tail variants. Your site is where you learn.
  • Wholesale accounts want a curated range they can reorder reliably. They value a core that is always in stock over breadth.
  • Retail partners on EDI have the strictest constraints: listed items, case packs, lead times, and compliance rules. Adding a SKU to a retailer requires a new item setup, and dropping one can affect your shelf space.

A useful exercise is a channel grid: every SKU down the side, every channel across the top, and a mark for where it is sold. Then check three things.

  1. Every retail-listed SKU has core-level depth. If it is listed at a retailer, it needs to be in stock for them.
  2. No SKU is sold in one channel only because nobody decided otherwise.
  3. Shared inventory is allocated deliberately. When DTC and a retail account draw on the same stock, decide in advance which channel gets priority when stock runs low, and at what level of cover.

Decision 4: Exit — when to stop carrying a product

Cutting SKUs is harder than adding them, because each one has a supporter. Set the exit rules before the season, so the decision is mechanical when the data comes in.

A SKU is a candidate for exit when two or more of these are true:

  1. Its sell-through over the last two seasons is in the bottom quartile of its category.
  2. Its gross margin after markdowns is below the category floor.
  3. Removing it would move most of its demand to a neighbouring SKU (check by looking at what customers who bought it also bought).
  4. It is not listed at a retail account where removal would cost shelf space.
  5. Its minimum order quantity forces a buy larger than a season of demand.

Exit should mean no new buys. Existing stock then needs its own plan: sell through at full price where possible, move it to the channel where it sells best, or mark it down on a schedule. See the markdown guide linked below.

What makes assortment plans go wrong

Most assortment mistakes come from planning on stale or partial data.

  • Planning DTC and wholesale separately. Two teams each plan their own range against the same stock, and the first channel to sell wins the inventory.
  • Using last year’s sales as demand. Last year’s sales include stockouts, which hide demand, and markdowns, which inflate it.
  • Ignoring open purchase orders. A range review that looks only at on-hand stock misses units already committed to arrive.
  • Reviewing once a season. A test SKU that sells out in three weeks needs a reorder decision in week three.

Endless keeps inventory, open purchase orders, and orders from every channel on one record, with demand forecasting built on that data. The sell-through, cover, and on-order numbers an assortment review needs are current rather than assembled from exports.

A seasonal assortment review in eight steps

  1. Pull sell-through, margin after markdowns, and cover for every active SKU, by channel.
  2. Assign every SKU a role: core, seasonal, test, or exit.
  3. Apply the exit rules and mark candidates.
  4. Run the three breadth tests on every proposed new SKU.
  5. Set depth by role, and by size curve where it applies.
  6. Build the channel grid and check retail-listed SKUs for core depth.
  7. Check the total buy against open-to-buy for each month.
  8. Set the in-season review dates and the reorder triggers for test SKUs.

A worked example: one category review

A brand sells a candle category of 14 SKUs across DTC, two wholesale accounts, and one retailer on EDI. The review uses the last two seasons of data.

SKU groupSKUsShare of unitsSell-throughMargin after markdownsRetail-listed
Three core scents, large size346%88%61%Yes
Three core scents, small size318%82%58%Yes
Four seasonal scents424%71%49%No
Two limited collaborations29%64%44%No
Two older scents23%38%22%No

Applying the framework:

  1. Roles. The six core SKUs are core. The four seasonal scents are seasonal. The collaborations were tests. The two older scents are exit candidates.
  2. Exit. The two older scents fail sell-through, fail the margin floor, are not listed at retail, and share most buyers with the core scents. Both exit. Existing stock moves to a DTC bundle and then a scheduled markdown.
  3. Breadth. The team wants to add two new seasonal scents. The incremental demand test, based on the last time a seasonal scent was added, shows about half of a new scent’s sales come from the other seasonal scents. They add one new scent, not two.
  4. Depth. Core SKUs get cover set by the retailer’s lead time. Seasonal SKUs are bought to sell through by season end, with the new scent bought at test depth and a reorder trigger at week four.
  5. Channel. The retailer asks to list a seasonal scent. The channel test passes only if the brand can hold core-level depth on it for the full listing period, which it cannot at current minimums. The scent stays DTC and wholesale this season.

The result is a 13-SKU range instead of the 16 the original plan called for, with more of the budget on SKUs that sell through at full price. The review took one afternoon because the data was ready. In most brands it takes a week, and most of that week is assembling the data.

Questions buyers ask

What is assortment planning?

Assortment planning decides which products a business carries, how much of each it buys, and where each is sold. It balances breadth of range against the cash, minimums, and forecast risk each additional SKU brings.

What is the difference between assortment planning and merchandise planning?

Merchandise financial planning sets the money: sales, margin, and inventory targets by category and month. Assortment planning decides which specific products fill those targets and in what depth.

How many SKUs should a brand carry?

There is no fixed number. Add a SKU when it brings new demand, its minimum order fits its expected sales, and a channel will carry it. Remove one when it sells in the bottom quartile, misses the margin floor, and its demand would move to a neighbouring product.

Commerce is chaos.

Tame your tech stack with one system that brings it all together—and actually works.

Get a Demo

Share this post