You have stock on the shelf. Something sold last week, and the first question is how to get another one.

In most retail categories that question has an answer. In a shop selling carved stone it does not — there is no second one of that piece, and there will never be one. What is left is a harder job: reading what your stock is telling you, using signals designed for shops where sold goods are replaceable.

The formula quietly assumes a replacement

Inventory turnover is cost of goods sold divided by average inventory value. It answers one question: how hard is the money in your stock working.

The usefulness of that answer rests on an unstated premise — that sold stock can be bought again. When a line sells out you refill it, and turnover becomes a measure of how fast you can grow a line that works. It is not a performance score. It is a rate of expansion.

Nothing goes back on the shelf in a one-of-a-kind range. When the piece sells, the slot is empty. Turnover still tells you how quickly your capital came back, but it can no longer tell you whether the business scales, because the mechanism scaling would use does not exist.

Two shops can post identical turnover figures and only one of them can grow.

There are two purchase paths, and they behave differently

This is where stock plans most often go wrong, so it is worth stating plainly — including how we work.

A single piece is bought as it is. It is photographed as the actual object, described as the actual object, and there is one of it. When it is gone, that design is not available in that stone again.

A batch order is not a stack of those pieces. Batch orders are carved to order, and a batch of a design is a batch of stones: the same design, the same material, the same nominal size, each piece cut from its own block. Our minimum is 10 pieces per order, with designs and materials mixed freely within the order.

So "reorder" is not one thing. Reordering a single piece is impossible. Reordering by the batch gives you more of the design and none of the exact object — why the second batch differs is about how far that goes.

The practical consequence for stock is blunt: you cannot fill a gap, you can only fill a shelf. The empty slot where the sold piece was is a display decision, not a purchasing one.

Sell-through reads backwards

Sell-through is the share of stock sold in a period. In a replaceable range a high number is unambiguously good: demand exceeded supply, order more.

In a one-of-a-kind range the same number has two opposite explanations, and they call for opposite responses.

  • The range is right. Customers want what you picked, and the answer is to buy more of that type — that material, that size band, that subject.
  • The price is low. The piece was underpriced and left money on the table.

The second reading is easy to skip, because manufacturing gives you a second chance: sell out at the wrong price and you raise it on the next run. Here, the underpricing is not recoverable on that piece. It is only recoverable on the pieces still on the shelf, and only if you act before they sell.

A test that separates the two: look at the family, not the piece. If one design sold fast while every related piece is slow, the outlier is probably a price signal. If everything in that material and size band moves, the band is the signal. And if a piece only sells after heavy in-store explanation, price is not the problem — the object needs a better position in the room, which is a different fix entirely. Merchandising carved stone covers what that changes.

The number that does survive: capital-days

There is a measure that keeps working when nothing can be replaced.

Capital-days = the cash in a piece × the days it sits before it sells.

Use landed cost, not the price tag. The money is not back in your account until the piece sells.

Two pieces, with illustrative figures, both inside a real range that runs $109 to $559 retail on stone from 1.32 lb to 9.08 lb:

PieceCash inDays heldCapital-days
Small$1209010,800
Large$4803014,400

The small piece turns three times faster. The large one still ties up more money. Rank by what sells and the small piece wins; rank by money and it does not — and the large piece is both the one to watch and the one worth a deliberate price decision rather than a hopeful wait.

That is the whole argument in one table: for unique stock, the ranking that matters is money-days, not velocity.

Depth is a range question, not a quantity one

"How many pieces should I hold" is the wrong shape of question when no design exists twice. The equivalent decisions are:

  • How many slots does the space actually have? A case that holds eight pieces at proper spacing is an eight-piece range, whatever your budget says.
  • What sits at the top of the range? A high piece anchors the price of everything beside it. A range that stops at the comfortable middle has no ceiling to point at.
  • What does an empty slot cost? Not the missing sale — the missing anchor. A half-empty case reads as picked over, and that is a merchandising cost, not an inventory cost.

One consequence is worth stating because it sounds like bad news: a range where everything sells out may mean the range never tested its ceiling. The pieces that sit are often the ones holding the price band up.

What to ask before you commit

Five questions, in order, for any stock plan:

  1. Which lines are single pieces and which are carved to order? Different purchases, different lead times, different repeatability. A supplier who answers "both" without distinguishing has told you something.
  2. If a single piece sells, is there any route to that design again? The honest answer is the most useful thing you will learn about a catalogue.
  3. For a batch, what is the minimum, and is mixing free? Ten pieces of one design is a very different commitment from ten pieces across five designs.
  4. What lead time is written on the quotation? Ours is quoted per order rather than published as a fixed number, because it depends on material, size, detail, finish and quantity — but it should always be a written date you can plan backwards from. Why lead times vary explains what actually consumes that time.
  5. Can the photographs be reused after the piece sells? On a one-of-a-kind piece the photograph is the object. On a batch it is one stone among many. Writing product descriptions is where that distinction has to be stated rather than discovered.

None of this makes one-of-a-kind stock worse than replaceable stock. It makes it different in one specific way: there is no restocking loop to learn from, so what you learn has to be put back into pricing, placement and range instead of into quantity. That is a slower loop, but it is not a blind one.

The pieces currently in stock are listed with their real weights and dimensions, and each one is the piece in its photographs — crystal skulls in stock is the current range.