Merchandise Planning for Jewelry & Watch Brands

What jewelry merchandise planning is

Jewelry and watch merchandise planning is the process of deciding what a brand will offer across fine jewelry, demi-fine, fashion jewelry, and timepieces: which collections get funded, at what depth and in which sizes, through which channels, and against which gifting occasions. The financial frameworks are standard — an open-to-buy sets the envelope, an assortment plan selects the line, a buy plan converts it into purchase orders.

What is unusual is that this category breaks two assumptions that most retail planning quietly depends on:

• →That cost is fixed once an item is created. In jewelry, gold, silver, platinum and stone costs move on their own markets, on their own timing, while the plan is still open.
• →That space is the binding constraint on depth. Here it is capital. Value density is the inverse of furniture, and what limits a door is tied-up money, insurance and shrink exposure rather than square footage.

Almost everything distinctive about planning this category follows from those two facts.

A scope note first, because the overlap is real. If you plan fashion jewelry as an accessories line — style and colour depth, fixed costs, no serialisation, no memo — then the accessories and bags approach describes your problem more accurately, and much of what follows will not apply. This guide is for lines where a moving cost base, a real size run, ownership status, or service parts are part of the picture.

Margin is re-planned, not monitored

In most categories, margin planning is a set-and-monitor exercise: initial markup is established when the item is costed, and variance against it is explained afterwards in the maintained markup bridge.

That model assumes the cost stops moving. In jewelry it does not. A metal move between costing and production can erode an IMU that was agreed months earlier, and because the move is exogenous, no amount of merchandising discipline prevents it. The plan finds out in the margin bridge, by which time every response has expired.

The mechanism that fixes this is not exotic: carry cost as a planning input that can be re-run, rather than as a fixed attribute set at item creation. When the metal or stone assumption changes, initial and maintained markup re-price across everything that depends on them, and the consequence becomes visible while the season is still open.

What that visibility buys is a choice, and there are four:

• →Adjust retail, where the price architecture and the competitive set allow it.
• →Change specification — metal weight, stone size, plating thickness — where the design can absorb it.
• →Shift mix toward pieces with lower metal content, which is often the fastest lever.
• →Accept the compression knowingly, and fund it from elsewhere in the plan.

All four are real options in month two and none of them exists in month eight. That timing difference is the entire argument for treating cost as dynamic. Planning margin on a moving cost base works through the mechanics.

One boundary worth stating immediately: this is not hedging. Buying protection against a commodity move is a treasury function and stays there. Seeing the merchandising consequence early is a planning capability, and it is a different thing.

Ring size is a genuine size run

An apparel-trained planner arriving in this category usually under-estimates this, because "jewelry" does not sound like a sized business. It is one.

Ring sizes distribute unevenly across a run, concentrated heavily in the middle with a long thin tail that still has to be carried. They shift by door catchment in the way size curves shift by region. They should be built from stockout-corrected sell-through rather than from receipts, because a size that cleared in week three never reported its true demand. And they break identically: a range missing its tail sizes is a broken run — the product is present, just not for that customer, which is exactly the experience of walking into a store where the assortment stops at a medium.

Bracelet lengths and watch strap sizes behave the same way, with the additional wrinkle that strap length is frequently adjustable at the point of sale, which changes the depth calculation but not the underlying distribution.

The practical discipline is the same as good apparel practice: build size curves per collection and per door tier, not from a house-level average. A solitaire collection and a stacking-ring collection do not share a size distribution, and a flagship door and a mall door do not either. Size curve allocation is the right unit for distributing depth once the curve exists.

Capital, not space, is the depth ceiling

This is the second broken assumption, and it changes how the depth conversation should be run.

In apparel, home, or toys, the question "how deep can this door go" is ultimately answered by space — facings, floor area, back-of-house capacity. In jewelry, a case holds far more value in far less volume, and what actually caps depth is the combination of tied-up working capital, insurance cost, and shrink exposure.

Three consequences follow that are easy to get wrong:

• →The under-assorted case may be correct. A door that looks thin against its space is often exactly as deep as the balance sheet permits. Planning depth against space in this category systematically over-buys.
• →Depth decisions are portfolio decisions. Because capital is fungible across doors and collections in a way space is not, adding depth in one place genuinely takes it from another — which is not true when the constraint is a fixed number of facings.
• →Carrying cost is a first-order input. Working capital tied in inventory is not a reporting metric here; it is a planning constraint that belongs alongside the open-to-buy envelope.

Occasions, not seasons

Jewelry demand concentrates into gifting occasions — holiday, Valentine's Day, Mother's Day, engagement season — and each behaves as its own short window with its own mix, its own price-point skew, and its own purchaser.

That last point is the one most often missed. The purchaser is frequently not the wearer, and gift-givers behave differently: they cluster on recognisable pieces and known price points, they are more brand-led and less size-confident, and they buy later than self-purchasers do. A plan that treats an occasion as simply a demand spike on the existing mix will get the mix wrong even when it gets the volume right.

The planning implications are straightforward but rarely implemented:

• →Phase receipts to land ahead of each occasion individually, rather than smoothing across a season that contains several.
• →Plan the mix per occasion, not just the volume — gift-heavy occasions skew toward different pieces than self-purchase periods.
• →Keep an evergreen core replenishing underneath, because the pieces that carry the case between occasions are a different business from the occasion buys themselves.

Planning the gifting calendar covers the sequencing in detail, including the sizing question for engagement-driven demand.

Ownership status is part of the position

In fine jewelry and watches, not all stock in a door is owned, and not all owned stock is in a warehouse.

Memo inventory is placed with a retailer and remains the brand's property until it sells. Consigned stock works from the other direction. Both are commonplace, and both break a plan that carries one inventory number.

Collapsing them produces two errors that push the buy in opposite directions:

• →Counting memo stock as sold overstates performance and understates exposure — the brand still owns units it believes it has moved.
• →Ignoring consigned stock sitting in doors understates availability and triggers buys that should not happen.

A plan that runs both errors simultaneously, which is the normal case, is not merely imprecise — it is being pulled in two directions by its own data. The fix is to carry owned, memo, and consigned as distinct positions against the same plan, which is a data-model decision rather than a feature.

The settlement side — who owes what, when title transfers, how memo agreements reconcile — is a finance and ERP process and stays there.

Service parts are real demand

For watch brands especially, the service business consumes a continuous stream of straps, links, clasps, batteries, and movement components. These carry no first-sale revenue, compete for the same production capacity as sellable pieces, and consume the same open-to-buy.

Planned outside the merchandise plan — which is the norm — they surface as a service backlog rather than as a planning variance. That routes a planning problem to the service team, who cannot fix it, and it happens late.

The answer is the same one that works for testers and gift-with-purchase in beauty: plan them as named demand lines inside the same buy, each with a derivation — installed base and failure rate for movement parts, attach rate for straps — rather than as a lump-sum estimate.

The boundary: serialisation and provenance

Fine jewelry and watches are frequently tracked at the level of the individual piece — serial numbers, certificates, stone grading reports, and the provenance chain attached to them. Two units of the same reference are genuinely different units, with different documents and different histories, and for high-value pieces the document is part of the product.

RetailNorthstar does not do serialised, piece-level inventory tracking, and we are not going to describe it as though it does. That belongs in an ERP or a POS with serial support, and a brand running this platform would keep it there and plan alongside it. If piece-level provenance is the problem you are trying to solve, this is not the tool for that job.

The distinction that makes the two systems coexist is between planning a position and proving a piece. A merchandise plan needs to know how many units of a reference sit where, in what ownership status, against what demand. It does not need to know which certificate belongs to which stone — and a system that tried to do both well would do neither.

One connected plan

The spreadsheet-era answer is a file per problem: a collection workbook, a metal cost sheet updated by finance, a size-curve tab, a memo tracker owned by sales, a service parts file owned by operations, and a door list correct on the day it was exported. Each may be individually accurate. What breaks is the connection — they reconcile at month-end, by hand, and the questions that actually govern the business (what does this metal move do to the collection's margin, and which levers are still open? which doors are carrying broken size runs right now? how much of what we think we sold is still ours?) require rebuilding the roll-up from scratch.

A connected model handles it differently. The assortment plan carries collections with size-level depth attached, margin planning re-runs against a cost that is an input rather than a constant, the buy plan sizes orders against a capital ceiling with service parts as named demand, and allocation runs at door and size level with ownership status intact.

For how this maps to the platform specifically, including the honest fit assessment, see the jewelry and watch industry page.

Related resources

• →Jewelry & Watch Brands — RetailNorthstar — Platform fit for jewelry and watch planning teams
• →Planning Margin on a Moving Cost Base — The mechanics of re-planning margin
• →Planning the Gifting Calendar — Phasing occasions individually
• →Accessories & Bags Brands — RetailNorthstar — The better fit for fashion jewelry planned as accessories
• →Target Costing Formula — Working backwards from a required margin
• →Size Curve Allocation Formula — Distributing depth across a size run
• →Working Capital Tied in Inventory Formula — The constraint that binds depth
• →Initial Markup — Glossary — The number a cost move erodes