Visual assortment planning: draft it, review it, commit it on one board

The assortment that gets built three times

The familiar sequence starts with a merchandiser building the category in a spreadsheet: rows of style-colors with delivery, channel and tier columns, and a running total against open-to-buy. For the review, that spreadsheet becomes something people can look at: a slide deck, printed cards on a wall, or images in a whiteboard tool. The meeting makes its decisions on the picture. Someone captures them in notes and phone photos, and over the next days a planner works through the notes and re-keys every add, cut and move back into the spreadsheet and the planning system.

Every handoff in that sequence is a transcription. Each one costs time, and each one can drop a decision. Four things go wrong, in roughly this order:

• The room decides without the numbers. The picture shows the product, not what it costs the budget. An added colorway looks free on a wall, and its open-to-buy cost arrives days later, after the people who made the decision have left.
• Attribution is lost. “Move the hero into Drop 2 for stores only” becomes a line in someone’s notes. If the delivery is entered and the channel is not, the plan carries the style without its context, and allocation finds out later.
• The record and the agreement diverge. What the room agreed and what got keyed in are two different documents. When they disagree, nobody can say which one is the assortment.
• The meeting happens twice. When the re-keyed numbers come back over budget or off mix, decisions the room thought were closed are reopened, usually by email and without the room.

None of this is a people problem. A visual review and a numeric plan held in separate places will always need someone to carry decisions from one to the other.

When the board is the plan

A data-backed board removes the transcription by making the picture and the record the same object. Three things follow from that:

• Placement is the data entry. Dropping a style-color into the Drop 2 column of the stores board, in the Better tier of knit tops, records all four attributes at once. Moving it changes them. There is no separate form to fill in afterwards, and no attribute left behind in someone’s notes.
• The impact is visible while the decision is still a conversation. Open-to-buy, option counts, margin and mix move as products are placed, so a proposal is argued with its cost on the screen instead of costed after the meeting has closed.
• Commit is the handoff. When the final meeting agrees, it commits the board. What the room sees is the assortment plan, with every attribute set, and the buy is built from it.

One condition matters more than any feature: the board has to be connected to the plan before the first draft, not added at the meeting. A board that only shows up for the review is just a nicer deck, and the transcription comes back.

Three stages, one board: draft, team review, final meeting

The assortment workflow has three stages, and each asks the board a different question. What stays constant is the board itself: the draft is not exported for the review, and the review is not rebuilt for the final meeting. Each stage starts from where the last one left the board.

The first draft is where a data-backed board pays off most, though it gets the least attention. A merchandiser who watches open-to-buy and the option count while placing brings a draft to the team review that already fits. The team review can then spend its time on the season instead of on arithmetic. The running order and decision log for the final meeting follow the same pattern as a line review; the line review agenda template works for it with the axis changed.

Four ways to lay out the same assortment

Teams assort along whichever axis the decision in front of them needs: by category and class, by delivery or drop, by channel or store cluster, by price tier or story. Each layout answers one question well and hides another, so a good review walks the same assortment through more than one layout. On a data-backed board these are views of one plan rather than four files, and a move made on one shows on the others.

Choosing the axis, designing the card and showing breadth against depth are covered step by step in how to build a visual assortment board; the tier and color layouts have their own guides in good-better-best line architecture and building the color story.

What should move as products are placed

Four readouts carry most of the weight. Each should update when a product is placed, moved or removed, at whatever level the board is showing: the class, the drop, the cluster or the tier.

The readouts inform a decision. They do not make it. A readout over target is a question for the room, not a veto: a newness share above plan can be the right call for a drop whose story depends on it, as long as the room makes that call knowingly. The arithmetic behind each figure is on retail-plan.com, in the IMU calculator and the assortment planning template.

One change in the final meeting, with the readouts on screen

A hypothetical season: women’s knit tops, Drop 2, with an open-to-buy of $180,000 at cost, a target of 14 options and a newness target of 60%. The team review leaves the board at 14 options, $176,400 committed and 8 of the 14 options new. In the final meeting, design asks for a new hero colorway: 600 units at $12.00 cost and $48 retail. On the board, the readouts go over budget the moment it is placed. To fund it, the room cuts a carryover colorway: 450 units at $11.00 cost and $39 retail.

The hero adds $7,200 at cost and $28,800 at retail; the cut releases $4,950 and $17,550. The drop ends $1,350 inside its budget at 14 options, with $11,250 more planned sales than the team review left it, and the blended initial markup moves from 72.0% to 72.1%, because the hero’s 75.0% markup replaces a colorway at 71.8%. Newness ends at 64%, four points over target. The room can see that and accepts it, because the hero is the drop’s story.

On a wall, the same change looks like one card swapped for another. Its cost would have been found when someone re-keyed it, and the newness overshoot might not have been noticed at all. Every figure in this example is hypothetical, chosen to make the arithmetic visible, not a benchmark.

When a wall, a deck or a spreadsheet is enough

Plenty of teams run a good assortment review on printed cards and a spreadsheet. If the assortment is small, there is one channel and a few deliveries, and the planner who owns the spreadsheet can enter the meeting’s decisions the same afternoon without error, then the transcription is cheap and the tooling would be overhead.

The honest trigger is the transcription itself. When re-entry after the meeting takes days, when channel or cluster boards have to agree with each other, or when decisions are reopened once the numbers come back, the gap between the picture and the plan has started to cost more than closing it. The broader comparison of board tools is in line board vs spreadsheet, Miro and Airtable, and the software category in visual line planning software.

Five things to ask in a demo

• Move a product to another drop and show me the plan. Which fields changed, and did the total for each drop move?
• Show me the impact before anything is committed. Open-to-buy, option and SKU counts, margin and mix, at the level the board is showing.
• Switch the layout without rebuilding. Go from the category grid to the channel board: is it the same plan, or a second board someone has to keep in step?
• Show me the first draft and the final meeting on the same board. Does the merchandiser’s draft carry through the reviews, or is it exported and rebuilt?
• Show me the committed assortment in the buy. Does the plan the buy is built from carry every attribute the board set, without anyone re-keying it?

Visual assortment in Canvas

Line Board is published by RetailNorthstar, which makes Canvas, so this section describes our own product. In Canvas, the merchandiser drafts on the same boards the final meeting commits: category and class grids, delivery and drop boards, channel and store-cluster boards, and price-tier and story boards. Products are dragged or placed onto them. Each placement, move or removal updates open-to-buy against plan, option and SKU counts, units, sales and margin, and mix balance. When the final meeting commits, the board is the assortment plan, with every attribute its placement set, and the buy is built from it.

Teams running Canvas in production tell us this is the part of the workflow they value: a visual, creative way to build and agree the assortment, backed by data, with the impact visible while it is still being discussed, and nothing to re-key once the group has agreed.

• ✓Visual assortment planning builds, reviews and commits the assortment on one image-first board backed by the plan’s data.
• ✓The workflow has three stages on the same board: the merchandiser’s first draft, the team review, and the final meeting with stakeholders that commits it.
• ✓Placement is the data entry: where a product sits records its class, delivery, channel or cluster and tier, so nothing is transcribed after the meeting.
• ✓Four layouts (category, delivery, channel or cluster, tier or story) answer different questions; a good review walks the same assortment through more than one.
• ✓Open-to-buy, option and SKU counts, units, sales and margin, and mix should move as products are placed, so a trade is argued with its cost on screen.

• How to build a visual assortment board →
• Line board vs assortment board →
• Locking the line →
• Planning drops on the line board →
• Line review agenda template →
• Visual line planning software →
• Canvas: visual assortment in RetailNorthstar →

Frequently asked questions

What is visual assortment planning?

Visual assortment planning is assortment planning done on an image-first board backed by the plan’s data. The merchandiser places products onto boards laid out by category, delivery, channel or store cluster, and price tier; the placement itself records what the assortment carries, where and when; and open-to-buy, option counts, units, margin and mix update as the board changes. The same board is drafted alone, reviewed by the team and committed in the final assortment meeting, so the agreed board is the assortment plan.

How is visual assortment planning different from a visual assortment board?

The assortment board is the artifact: the locked line laid out by cluster, channel, delivery or tier, with breadth and depth on the cards. Visual assortment planning is the workflow around it: who drafts it, how the team reviews it, how the final meeting commits it, and whether the board is a picture of a spreadsheet or the plan itself. How to build the board step by step is covered in the guide to building a visual assortment board.

Who works on the board before the final assortment meeting?

The merchandiser who owns each category drafts the first assortment alone, inside the category’s open-to-buy. The merchandising team then reviews the drafts together, with planning, to make the categories add up to one season. Only then does the wider group of design, planning, buying, wholesale or sales and leadership meet to finalise and commit it. Each stage works on the same board, so nothing is exported between them.

Does a visual assortment replace the assortment plan spreadsheet?

Only if the board is backed by the plan’s data. If placing a product on the board records its delivery, channel, cluster and tier in the plan, and the totals update as it moves, the board is the assortment plan and there is nothing to re-key. If the board is images on a wall, a slide deck or a whiteboard, the spreadsheet stays the record and someone has to transcribe the meeting into it afterwards.

When is a wall, a deck or a spreadsheet good enough?

When the assortment is small enough that one person can transcribe the meeting the same afternoon without error, there is one channel and few deliveries, and the room does not need to see open-to-buy move to make a decision. The case for a data-backed board starts when re-entry after the meeting takes days, when several boards for channels or clusters have to agree, or when decisions keep being reopened once the numbers come back.