How to build a visual assortment board

• ✓A visual assortment board is the locked line laid out by where it goes — cluster, channel or delivery — with depth stated on every card.
• ✓It is built between two locks — from the line lock at line review to the buy — and after the buy it is read, not rebuilt, before each delivery is allocated.
• ✓Choose one axis for the grid and keep the others as filters; a board laid out on two axes at once answers neither question cleanly.
• ✓Breadth is shown by presence — a card is in a column or it is not — and depth by a number on the card, never by card size.
• ✓Every row and column has to sum: options to their planned units — or their tier’s total — and their minimums, clusters to their fixture capacity, the board to the receipt budget.
• ✓In the buy meeting a change is a trade: a unit added to one card names the card it comes from.
• ✓What a card stands for changes by vertical — a style-colorway with a size run in apparel, a style-colorway in a size and width run in footwear, a shade on a planogram in beauty, a model on a floor set in home.

Who builds an assortment board, and when

The assortment board is built between two locks. It starts when the line locks at line review and is itself locked when the buy is placed, and it is the working surface for everything decided in between: which options each cluster or channel carries, how deep, and in which delivery. Before the line locks there is nothing stable to assort; after the buy is placed, the decisions on the board have become purchase orders. What sits upstream of it is covered in how to build a line board and how to build a line plan; what locks the line is covered in how to run a line review.

The board is used at three moments, by overlapping groups of people:

• The assortment build. Planning and buying build the first version from the locked line, the store clustering and the receipt budget. The planner owns the numbers — units, the open-to-buy, the sums. The buyer owns the product view — which options each cluster’s customer will buy, and what the vendor can deliver at what minimum. The allocator brings what the doors can take: fixture capacity, presentation minimums and the size curve by cluster.
• The buy meeting. The board is walked cluster by cluster or delivery by delivery with buying, planning and allocation in the room, and merchandising or design present for the story and the hero. Decisions are made against a picture the whole room can read, and the buy plan is written from the board the meeting signs.
• Store and channel assortment reviews. After the buy and before each delivery is allocated, the locked board is read again against the stores: fixture counts, floor layouts, and any door that has changed format since the clustering was set. Where the brand sells wholesale, the account-level version comes earlier, after the prebook and before production is committed: the accounts’ orders show which options each account carries, and the board shows that pattern back to planning.

The people change less than the question does. At the build the question is whether the board is complete and sums; in the buy meeting it is whether the room agrees with it; at the store review it is whether the doors can actually present what the board gave them.

What the board is built from

Five inputs, all of which exist before the first card is placed. If any one is missing, the board can still be drawn, but it cannot be checked.

• The locked line. Every option with its planned price point, tier, delivery, carryover or new flag, hero designation and planned units — the output of the line plan and the line review. See locking the line for what the lock commits.
• The store clustering. Doors grouped by what they sell and how much — by volume grade, climate, format or customer profile — with a door count per cluster. The cluster, not the individual door, is the board’s working unit, because a board with a column per door is a spreadsheet with pictures in it.
• Fixture capacity. How many options of the category each cluster’s fixtures can present in a delivery. It is the ceiling on breadth by cluster, in the same way that supplier minimums put a ceiling on the option count in the line plan.
• Presentation minimums. The fewest units a door can hold of one option and still present it — one size run with the core sizes doubled in apparel, a pair per size in footwear, a filled facing plus a tester per shade in beauty. It is the floor under depth by door.
• The receipt budget by delivery. Units and open-to-buy for the category in each delivery, carried from the merchandise financial plan through the line plan. It is the number the board has to sum to.

Supplier minimums travel with the line rather than staying on the line plan. A factory minimum per option and a mill minimum per fabric and color are what stop the board from thinning an option below the depth it was quoted at, so carry them onto the board as a note per row or per fabric group.

How to build a visual assortment board in eight steps

Steps one to three set the board up, four to seven fill and check it, and eight puts it in front of the room. A board taken into the buy meeting before step seven spends the meeting finding arithmetic errors instead of making decisions.

Showing breadth and depth on the same board

Breadth and depth divide the same units, so a board that shows only one of them hides the trade between them. Breadth is shown by presence: a card is in a column or it is not. Depth is shown by a number on the card, with a visual cue that groups the numbers into bands rather than replacing them.

• Presence for breadth. A card in a cluster column means the cluster carries the option; an empty cell means it does not. Read down a column for the cluster’s breadth and across a row for the option’s reach. Avoid faded or half-transparent cards for “maybe” — a maybe on the board becomes a yes in the buy plan unless somebody remembers to remove it.
• A number for depth. Units per door on the card in each column, and total units at the end of the row. The number is the decision; every other mark is a cue to help the room find it.
• Depth bands as the cue. A small marker — a bar, a fill, a label — that groups units per door into three or four bands defined for the category: at the presentation minimum, standard, deep. Bands let the room see where depth is concentrated without reading every number. Define them in the board’s header so everyone reads them the same way.
• Not card size. Scaling cards by units turns the board into a chart and makes size read as importance: the deepest basic becomes the biggest image, which is the hero’s job rather than the basic’s. Keep every card the same size and let the numbers and bands carry depth.

Once the board is drawn, three patterns are worth reading for, because each is invisible in a table and obvious on a board.

• Thin everywhere. An option present in every cluster at the minimum. It reads as wide reach, and it buys no selling depth anywhere — the size run breaks early in every door at once, and from then on the option records as slow when the sizes customers wanted were simply gone.
• Deep where it is untested. A new option at standard or deep depth in every cluster. The board is betting the whole read before the read exists, and the bet is the same size in the doors least likely to confirm it.
• The under-filled column. A cluster whose fixture holds more options than the board gives it. A fixture with room on it gets filled from whatever stock is available to send, which is the assortment decision made by default rather than in the meeting.

Read the board twice, once by column and once by row, because each direction surfaces a different error: columns show what a store will look like, and rows show what a style will sell through.

An illustrative assortment board: one delivery, nine options, three clusters

Every figure below is illustrative — not a benchmark, not drawn from any brand, and chosen so the arithmetic can be checked.

Take Drop 1 of the women’s knit tops plan worked through in how to build a line plan: nine options — four Good at $39, three Better at $65 and two Best at $98 — carrying 1,680, 900 and 500 units, 3,080 in all. Assume two of the Good options and one of the Better options are carryover and the other six are new, and that the Better option B2 is the drop’s hero.

In this illustration the brand sells through three store clusters and e-commerce. Cluster A is five large doors whose knit-top fixture presents nine options per delivery; cluster B is ten core doors presenting seven; cluster C is eight small doors presenting four. E-commerce draws on one stock pool. The presentation minimum is eight units per door: one run of XS to XL with S, M and L doubled.

The first draft fills A and B to capacity and gives C the three carryover options — G1, G2 and B1 — which leaves C one slot short of the four its fixture holds. Every Better option holds the 300 units the line plan gave it. The empty slot in C is the question the board puts to the buy meeting.

The room decides the hero has to reach every door, so B2 goes into C at the minimum: eight units in each of eight doors, 64 units. The meeting holds the Better tier at its 900 units for the drop, so the 64 have to come from inside it, and they come from e-commerce, where a unit cut does not break a door’s size run: 24 from B2’s own e-commerce line (80 to 56), 20 from B1’s (56 to 36) and 20 from B3’s (120 to 100). B2 rises to 340 units, and B1 and B3 fall to 280 each. The line plan example’s factory minimum for a Better option is 275 units, so B1 and B3 now sit five units above it — a further cut of more than five units to either would put that option below the minimum it was quoted at.

Read the columns first. An A door takes 176 units of the drop, a B door 88 and a C door 36 — twice and nearly five times as much — and the option counts step down nine, seven, four with the fixtures. Cluster C carries the two Good carryovers, the Better carryover and the hero, so its customer meets an opening price, a proven Better option and the season’s story, each at a depth that presents a size run. Cluster B carries everything except Best, and cluster A carries the whole drop. E-commerce carries 1,032 units, about a third of the drop, including the deepest lines of the two Best options, which are in only five doors.

Then read the rows. The two new Good options, G3 and G4, are tested in fifteen doors and online rather than all twenty-three, at 20 units per A door against the Good carryovers’ 24. The 80 units a C placement would have taken, and the 20 held back from the A doors, go online instead — 180 units against the carryovers’ 80 — where one stock pool can read the test without breaking a size run in any door. The hero is in all twenty-three doors but at the minimum in C — a presence buy rather than a selling buy, and the first number on the board to reread once the drop is on the floor. And in the line plan example the two Best options share one fabric, whose 500 units sit 50 above the 450-unit mill minimum: between them, X1 and X2 can lose no more than 50 units before the fabric buy breaks.

Every row sums to its total and every column to its cluster: 880 + 880 + 288 + 1,032 = 3,080 by cluster, and 1,680 + 900 + 500 = 3,080 by tier. The meeting changed where 64 units sit and changed nothing about how many there are, which is the test of a buy meeting that traded rather than added.

Working the board in a buy meeting

The buy meeting is where the board stops being a proposal. Buying, planning and allocation make the decisions; merchandising or design attends for the story and the hero, and finance joins where the open-to-buy is in question. Each function reads a different part of the same board, which is why the board is drawn rather than circulated as a spreadsheet.

• The planner holds the sums. The delivery’s receipt budget, the open-to-buy, the planned units per option and the minimums. Their job in the room is to say what a change costs before it is agreed.
• The buyer holds the product and the vendor. Which options each cluster’s customer will buy, which carryover earned its depth, and what the factory will accept if a quantity falls below the quoted minimum.
• The allocator holds the doors. Fixture capacity, presentation minimums, the size curve by cluster, and how the stock will physically land. A depth the doors cannot present ends up in the back room or the distribution center rather than on the floor.
• Merchandising or design holds the story. Whether the hero reads as one, whether each cluster’s column still tells the season’s story, and which options carry it. See building the color story for reading color across a laid-out range.

A running order that keeps the decisions ahead of the arithmetic:

1. Confirm the inputs: the locked line version, the cluster door counts, the fixture capacities and the budget. Settling these first keeps the meeting from arguing over numbers that are simply out of date.
2. Walk the hero and the story across every cluster, so the decisions that shape the season are made while the room is fresh.
3. Walk each cluster column, largest first, against its capacity and its customer.
4. Resolve the empty slots and the options that are thin everywhere.
5. Read the totals again: rows against planned units and minimums, columns against capacity, the board against the budget.
6. Record and lock: every change with its source of units, its owner and its reason, then a dated version of the board.

The rule that keeps the meeting honest is that every change names where its units come from. In the illustrative example, putting the hero into cluster C was a 64-unit decision, and the meeting made it by moving units from three e-commerce lines inside the Better tier rather than by raising the tier’s total. A meeting that adds units instead of moving them is spending open-to-buy nobody budgeted, and the board will look better than the plan that has to pay for it.

The board allows five moves, and each has a source and a destination. Adding a cluster to an option needs a free slot in that cluster — or a swap — and units from inside the tier. Dropping a cluster frees a slot and units that have to go somewhere named. Deepening an option in a cluster takes units from another cluster or another option. Thinning one releases units, but never below the presentation minimum or the supplier minimum. Swapping one option for another in a cluster changes breadth without changing the count, and is the move to reach for first.

Record the reason as well as the result. The next delivery’s board starts from this one, and a cluster that lost an option for a reason will ask for it back; without the reason on file, the same argument is held again. The same discipline applies to the line review, and the line review agenda and sign-off template carries a decision log that works for the buy meeting too.

How the board flows into the buy plan and allocation

The locked board produces three things the next steps need, and each lands somewhere different. The board is where depth is decided; the buy plan is where it becomes a purchase order, and allocation is where it becomes a shipment.

• Option totals to the buy plan. Each row’s total units become order quantities, split by size from the size curve and by delivery, at confirmed costs and against the open-to-buy. Where a size curve is still being set, the size curve calculator on retail-plan.com works through the split.
• Presence and units per door to allocation. The columns tell allocation which doors receive which options and how many units per door at first delivery, and the size curve by cluster splits those units by size. The board sets the intent; allocation turns it into door-level, size-level shipments when the stock lands.
• The decision record to next season. What each cluster carried, at what depth, and why, is what the next hindsight reads against sell-through by cluster. See hindsighting the line.

Two things break at this handoff. The first is a board change that never reaches the buy plan: a cluster added in the meeting and not in the order quantities, so the doors are promised stock that was never bought. The second is allocation working from a different clustering than the board — a door the store review moved from cluster B to cluster C still sits in B in the allocation file, so it receives B’s depth on C’s fixture. Both are re-keying failures, and both are visible only when somebody lays the board, the buy plan and the allocation file side by side.

The numeric side of this handoff — depth by channel against open-to-buy, size curves, the buy plan itself — is covered on retail-plan.com in line plan vs assortment plan and the assortment planning template. This guide stays on the board.

How the assortment board changes by vertical

The eight steps hold in every vertical. What changes is what one card stands for, what sets the presentation minimum, and what the columns are.

• Apparel. A card is a style-colorway carrying a size run, and the presentation minimum is one run per door with the core sizes doubled, so a door can take an option only at or above the depth one size run needs. The board is read by delivery — one board per drop or floor set, clusters across — because each delivery has to merchandise on its own in every cluster; see planning drops on the line board. The size curve by cluster, applied in allocation, turns the board’s units per door into units per size. Grid-sized categories need far more units per door to present one option — denim by waist and inseam, intimates by band and cup — which cuts how many options a small door can carry; see the denim, intimates and swim guide.
• Footwear. A card is a style-colorway committing a size run of pairs, plus a width run where widths are offered — a medium and a wide fit are two runs, not one. The presentation minimum is a pair per size in the run a door carries; a smaller door can be assorted a trimmed run of the core sizes, which lowers its minimum, and the card has to say so or allocation ships the full run. Core styles carry over for years and hold their slot across seasons while fashion styles rotate by drop; where a performance style is updated to a new version, the outgoing and incoming versions can sit in the same cluster column in the changeover delivery, so the card marks which one the cluster is selling through. See the footwear and accessories guide.
• Beauty. A card is a shade within a franchise, and the shade range plays the size curve’s part: a door carrying a franchise needs enough of the range to serve its customers, and a missing shade leaves no trace in the sell-through because nobody could buy it. The planogram is fixture capacity made literal — a set number of facings per gondola — and the retailer’s gondola reset dates set the deliveries, so a shade that misses a reset waits for the next one. Testers take units per door that never sell and are replaced on their period-after-opening, and the shelf life of sealed stock caps how deep a door can be stocked. See the beauty and personal care guide.
• Home and furniture. A card is a model in a finish, cover or configuration, and the board is laid out by floor set and store format: which models sit on each format’s floor, and which are offered by special order without floor space. Where stock is held in a distribution center rather than in the store, the store column carries the floor-sample decision and the depth decision moves to the distribution-center line. Container minimums and ocean lead times fix the buy months ahead of the floor set, so the board locks earlier than an apparel board and changes by swap after it.
• Sporting goods and outdoor. The columns can be regions or climate clusters rather than volume grades, because seasonality by sport follows the weather — a ski or cycling option belongs in the doors where that season is running. Hard goods are cards by model and model year, soft goods by colorway and size run, so one board carries two kinds of card and cannot be summed without stating which. Dealer prebooks placed months ahead set the wholesale columns, and MAP pricing limits markdown as an exit, so a cluster that was over-assorted clears through closeout channels or carries into the next model year. See the home, outdoor and sporting goods guide.

Across all five, the test of the board is the same: every card states its unit, every column states its capacity, and the board sums back to what the line plan approved.

Where assortment boards go wrong

Each of these shows on the board before it shows in the season, which is the practical argument for drawing one.

• Assorting a line that has not locked. Depth is committed to options the line review then cuts. The full case is in line board vs assortment board; the fix here is to build from the dated, locked version.
• A board that does not sum. Rows drifting from planned units, columns over their fixture capacity, a total over the budget. Each difference is small on its own; together they are an open-to-buy overspend that surfaces at the buy.
• The same depth everywhere. Every option in every door at the same units per door looks fair and gives every cluster the same assortment regardless of what its customers buy. The tell on the board is a grid of identical numbers.
• Depth drawn as card size. The deepest basic becomes the biggest image on the board, and the room reads it as the story. Keep cards one size and let numbers and bands carry depth.
• Three files. The board edited in the meeting, the buy plan in a spreadsheet, allocation in a third file. Every re-key is a chance for the three to disagree, and the disagreement is found by the stores when the stock lands.
• A hero in name only. Designated, present in every cluster, and bought at the minimum everywhere — so it breaks its size run first in the doors where it was meant to anchor the story.

On a wall, in a spreadsheet, or in software

An assortment board can be pinned to a wall, built in a slide deck or laid out in a spreadsheet with images in the cells; every check in this guide is arithmetic a spreadsheet can do. The limit is not the medium but the number of copies. A wall board cannot sum itself, a slide deck cannot be filtered by delivery, and a spreadsheet beside a separate buy plan and allocation file has to be reconciled by hand after every meeting. If you are starting from files, the line board template covers the visual grid and the assortment planning template on retail-plan.com covers the numbers. For what separates a board connected to the plan from a picture of it, see visual line planning software.

• Line board vs assortment board: what each one decides →
• Assortment board (glossary) →
• How to build a line plan, step by step →
• How to build a line board, step by step →
• How to run a line review →
• Locking the line: what sign-off commits →
• Balancing carryover and newness on the line →
• Planning drops on the line board →
• Hero style (glossary) →
• Hindsighting the line →
• Line plan vs assortment plan (retail-plan.com) →
• Assortment planning template (retail-plan.com) →

Frequently asked questions

What is a visual assortment board?

A visual assortment board is an image-first layout of a locked line that shows which options go to which store clusters, channels or deliveries, and how deep each one is bought there. Each card carries the option’s image, price, door count, units per door and total planned units, so breadth and depth can be judged by eye. It is built after the line locks at line review and before the buy is placed, and it is the surface a buy meeting works from.

How is an assortment board different from a line board?

The line board decides what the range is — which styles, colorways and price tiers make up the season. The assortment board decides where that locked range goes and how deep: which clusters, channels or doors carry each option, and at how many units. The line board comes first, and its locked output is the assortment board’s input.

When in the season is an assortment board built?

Between two locks. It starts when the line locks at line review, because before that there is nothing stable to assort, and it is itself locked when the buy is placed, because after that its decisions are purchase orders. In between it is built by planning and buying and walked in the buy meeting. After the buy it is read again rather than rebuilt, in store and channel assortment reviews before each delivery is allocated.

What goes on an assortment board card?

What the buy meeting decides from: the image, the option as a style and colorway, the price and tier, a flag for carryover, new, hero or channel exclusive, the door count, the units per door in each cluster, and the option’s total planned units. Cost, margin and sales history belong in the drill-down rather than on the face of the card.

How do you show breadth and depth on the same board?

Show breadth by presence — a card is in a cluster column or it is not — and depth by a number on the card: units per door in each column and total units at the end of the row. A depth-band marker can group the numbers so the room sees where depth is concentrated, but card size should not carry depth, because the biggest image reads as the most important option rather than the deepest.

Is an assortment board the same as a planogram?

No. A planogram shows where products physically sit on a fixture. An assortment board decides which options each cluster or channel carries and how deep, before the stock is bought. In beauty and other fixture-led categories the planogram sets the assortment board’s capacity — the facings a gondola holds are the ceiling on how many shades a door can carry — but the two documents answer different questions.

Can you build an assortment board in a spreadsheet?

Yes. Every check an assortment board needs — rows against planned units and minimums, columns against fixture capacity, the total against the receipt budget — is spreadsheet arithmetic, and images can sit in cells. The limit is copies rather than the medium: a board in one file, the buy plan in a second and allocation in a third have to be reconciled by hand after every change.