Product variant matrix: separate absent, zero and unknown

A product variant matrix should distinguish a possible option combination from a confirmed variant and its quantity. Complete this exercise to hand over six checked cells and two precise questions without converting missing facts into inventory.

Updated: 17.10.2026

What exactly does a variant identify?

A product family groups versions of one model. An option describes a dimension of choice, such as finish or capacity. A variant identifies one particular combination. Inventory quantity is a separate fact about that variant. Keeping those meanings apart prevents an apparently tidy spreadsheet from making claims nobody supplied.

Our fictional family B1 is a bottle with matte and gloss finishes and 350, 500 and 750 ml capacities. The IDs below are internal exercise labels, not allocated barcodes. This is a planning exercise for an owner preparing a product brief, not a demonstrated store configuration, customer experiment or recommendation to alter a live catalog.

WooCommerce describes option combinations and per-variation data. Those documents support the concepts, not our invented product facts. The catalog category and filter plan solves grouping across products. Here the question is narrower: which combinations within one family are actually known to exist?

Which six cells belong in the starting grid?

Write both finishes against all three capacities before deciding anything about availability. Two finishes multiplied by three capacities gives six possible cells. That calculation establishes the size of the option grid, not the number of products you sell. The supplied facts must determine which cells represent existing variants.

Treat this table as immutable input for the exercise. The supplier explicitly says matte 750 ml is not manufactured. It says nothing about whether gloss 350 ml exists. B13 is an existing gloss 500 ml variant, but its quantity has not been supplied. An unanswered existence question and an unanswered quantity question are not interchangeable.

The English example deliberately uses different merchandise from the Azerbaijani exercise. Its logical mapping is exact: B11 corresponds to V1, B12 to V2, B13 to V3 and B14 to V4. Matte 750 ml corresponds to the declared nonexistent cell; gloss 350 ml to unknown existence. Positive quantities differ, but the decisions do not.

How do you separate absent, zero and unknown?

Ask about existence first. If existence is confirmed, ask about quantity second. This two-question sequence matters because zero is a known quantity, whereas unknown is missing information. A quantity field cannot answer whether a supplier manufactures a combination at all. Keep the explanation beside the decision rather than relying on cell colour.

For B11 and B14, the supplied quantities are positive. That alone is not permission to promise an order or delivery date. Allocations, other commitments and fulfilment timing were not provided. The new commitment check addresses a separate decision about covering demand. Do not quietly turn this identity worksheet into that decision.

Write two different questions for the owner. For gloss 350 ml, ask whether the combination exists. For B13, ask for a checked quantity for that existing variant. “Check stock” is too vague for the first question and does not preserve what is already known about the second. Leave both unresolved until the appropriate answer is supplied.

How can an owner complete the grid in fifteen minutes?

Start with one model rather than the entire store. Write its family ID and the exact option names. The expected output is B1 plus finish and capacity, with a fixed list of values for each. Avoid adding lid colour or material merely because those properties could exist in another business. Nothing in this input authorizes those dimensions.

Next, enumerate the six combinations once each. Check that every finish appears against 350, 500 and 750 ml, without duplicated rows. The expected output is a complete possible grid, not a list of six new item records. Only after enumeration should you attach the four confirmed IDs from the supplied table.

Now copy existence and quantity into separate columns. Expected output: four confirmed variants, one explicitly nonexistent combination and one combination of unknown existence. Among the four confirmed variants there are two positive quantities, one zero quantity and one unknown quantity. Do not give the nonexistent combination an invented quantity just to make the column numeric.

Finish by writing a reason for each of the four boundary decisions. Someone reopening the file should be able to distinguish missing evidence from a negative fact without asking you. This is an inspectable paper result, not measured user completion time. Fifteen minutes is the proposed task budget, not a reported performance result or guarantee.

What should the handoff template contain?

Copy the following block into a blank local note. Keep the original table above it so another person can recompute every conclusion. For real business use, later add who confirmed each fact and when it was checked. Do not populate those provenance fields with fictional staff names or implied verification dates.

Family: B1
Options: finish [matte,gloss]; capacity [350,500,750] ml
Columns: finish | capacity | existence | ID | quantity
Question A: does gloss 350 ml exist?
Question B: what is the checked quantity for B13?
Never replace unknown with zero or nonexistent.
Deliverable: six cells, four confirmed IDs, two separate questions.

An AI assistant can help format the note, but it cannot supply the missing supplier answer. If you use one, instruct it to preserve the source values and produce only the grid and questions. Then compare its output with the fixed input yourself. The task is equally possible without AI, using a plain text editor.

A later importer will need an entirely different implementation brief. This note does not define a platform field mapping, inventory policy, barcode or product feed. The product schema guide concerns communicating product facts to search systems. A correct schema does not retrospectively establish the missing facts in this matrix.

Which four mistakes should the reviewer reject?

Reject “six combinations means six SKUs”. The grid has six possible cells but only four confirmed variant records. Reject “remove B12 because quantity is zero”. The supplied facts still identify B12 as an existing variant. Keeping it visible preserves the distinction between the product choice and a current quantity observation.

Reject “gloss 350 ml is not made”. That conclusion was supplied only for matte 750 ml. A missing answer cannot be transferred from a neighbouring cell. Finally reject “use zero for B13 until somebody checks”. That turns an information gap into a numerical fact and can mislead later processing even if the intention is temporary.

A stock discrepancy review compares records and observations under matching identities. This exercise prepares the identity distinctions before any count comparison. The purchase-line receipt review asks whether delivered goods match a specified purchase line, another task that cannot be settled by enumerating options.

In earlier public writing about perfection and effectiveness, I described projects delayed by my pursuit of perfection. The practical application here is finishing one bounded, checkable model before expanding the task. This bottle example is not a new personal client story, evidence of a successful store intervention or a measured business result.

How do you save and reopen a useful result?

On Windows, open Start, search for Notepad and create a new blank document. Name the heading Variant matrix V1. Copy the original six rows, the four decision explanations, the mapping note and both unresolved questions. Press Ctrl+S and save variant-matrix-V1.txt in Documents. Close the editor and reopen that exact file from File Explorer.

Confirm the six combinations are still distinct, B12 still has zero, B13 still has unknown quantity and gloss 350 ml still has unknown existence. Check that no invented fifth ID appeared. Reopening verifies the local artifact was preserved; it does not verify inventory, a storefront selector or a live import. Keep that boundary in the handoff.

If you later choose physical units for an order, their batch dates may require a separate FEFO batch picking plan. A variant grid does not select batches, reserve stock or authorize shipment. Its completed outcome is deliberately smaller: a consistent list of known combinations and precise questions about the missing ones.

Sources