Keystone is not your margin

A customer is standing at the counter with a bridal set she’s tried on twice this month, and now she’s asking for ten percent off to close today. You do the math you always do: the piece is keystoned, so there’s room. You say yes before she finishes the sentence. It feels like an easy call, because it is — right up until the end of the quarter, when the number in the account doesn’t look like the number you were picturing when you said yes.

Keystone was only ever the sticker price

Keystone — doubling your wholesale cost to set the tag — is the oldest pricing rule in the trade, and it’s a reasonable place to start. Say a stone-and-setting combination costs you $3,000 wholesale. Keystone puts the tag at $6,000. On paper, that’s a $3,000 margin: a hundred percent markup on cost.

On paper is the important phrase. Keystone prices the tag. It says nothing about what happens to that $3,000 between the day you write the ticket and the day the money is actually yours to keep — and for most sales, quite a lot happens in between.

Where the margin actually goes

Follow that same $6,000 ring through an ordinary month, and watch what a hundred-percent markup turns into:

• The discount taken to close. Ten percent off $6,000 is $600 — and that $600 doesn’t come out of the tag, it comes straight out of the margin. $3,000 becomes $2,400. A fifth of the profit, given up to save a sale that, at the price you’d already set, was probably closing anyway.
• The trade-in credited generously. Same visit — she brings in an old pendant against the purchase. Your bench values it at $400 in stone and metal. She remembers a friend getting closer to $650 somewhere else, and you split it at $550 to keep the visit moving. That’s $150 more gone, and it won’t show up anywhere as a discount — it’ll show up as a pendant in your case that cost more than it’s worth.
• The repair done at cost as a favour. A few weeks later she’s back for a prong check before a trip. Five minutes on the bench, and you wave the $65 ticket because she just spent real money with you. It’s the right thing to do. It’s also $65 that never gets counted against anything.
• The piece that sat two years. A different item now — a $4,000 wholesale estate piece that doesn’t move on any convenient schedule. It sits in the case two years before the right customer finds it: two years of capital that could have turned over three or four times in something else. Call it ten percent a year in what that money could otherwise have earned — $800, gone before the piece even discounts to finally sell.
• The memo return that shipped on your account. A parcel comes in on memo, nothing matches what the customer wanted, and it goes back. Return shipping and insurance on a $5,000 parcel runs somewhere around $50 — a rounding error on any single return, and not a rounding error the fortieth time it happens in a year.
• The special order re-sized twice. A custom band gets ordered half a size small, then — after a week of wear and a change of mind — half a size back up. Two trips to the bench at $45 in labour each is $90 that was never in the original quote, because the quote assumed one sizing, once.

Add up just the three that touched the bridal set — the $600 discount, the $150 generous trade-in, the $65 waved repair — and a sale that priced out at a $3,000 margin landed closer to $2,185. That’s still a good sale. It is not the sale you thought you’d made when you wrote the ticket at keystone.

Why labour-heavy goods run at double-key

This is exactly why bench-heavy goods — hand fabrication, heavy custom work, anything more likely to come back for a resize, a repair, or a warranty visit than a case piece is — commonly get priced at double-key, four times wholesale, rather than two. It isn’t greed. A $2,000 wholesale custom piece priced at keystone gives you $2,000 to absorb a design revision, a resizing, a bench hour nobody bills separately, and the ordinary favours you end up doing for a customer who watched you build something for her. Priced at double-key, that same piece gives you $6,000 to absorb the same list — and on custom work, the list runs longer than it does on anything you simply unlock from the case.

Why a flat multiple loses money on custom, specifically

Case goods are predictable enough that keystone, averaged across enough sales, works. Custom work isn’t average — that’s the entire pitch you make to the customer. Every one-off carries its own design revisions, its own findings, its own bench time, its own chance of a resize or two before it’s right. Quote a special order off the same flat multiple you use for a stone you unlock from the case, and you’re pricing the exception as if it were the rule. The stores that lose money on custom generally aren’t bad at custom work — they’re pricing it with a number built for something else.

Know your real number before you discount

None of this means keystone is wrong, or that double-key is the answer to everything you sell. It means the multiple on the tag was never the number that mattered. The number that mattered is what’s actually left after the discount, the generous trade-in, the favour, the carrying cost, and the rework that’s specific to that category of goods. A store that knows its real number per category — bridal, estate, custom, repair-inclusive — can say yes to a discount because there’s genuinely room, or say no because there isn’t, before the conversation at the counter. Not after the deposit’s already in the drawer.

This is the number the books Gemercel runs are built to show — the real margin per category, after the costs particular to custom and bench work, not just the multiple on the tag. If you’re pricing off keystone and finding the real number out at the end of the quarter, put your store on the list and see it before you quote the next one.