Where does raw material end and processed product begin?

A buyer or supplier meeting raw against processed for the first time needs one sentence first: the boundary inside UNSPSC where an input stops being raw material and starts being a product in its own right, in UNSPSC as managed by the United Nations Development Programme since 1 January 2025. The boundary is set by the codeset wording, not by how finished something looks to a buyer.

Level of detail for raw against processed

What it shows	Where it misleads
Broad view of raw against processed	that the boundary inside UNSPSC where an input stops being raw material and starts being a product in its own right
Close view of raw against processed	that the boundary is set by the codeset wording, not by how finished something looks to a buyer
Practical use of raw against processed	the claim that the boundary inside UNSPSC where an input stops being raw material and starts being a product in its own right

Finding the shelf for raw against processed

Think of raw against processed as occupying one shelf position defined by the claim that the boundary inside UNSPSC where an input stops being raw material and starts being a product in its own right. The boundary is set by the codeset wording, not by how finished something looks to a buyer. That position decides which families a buyer searches when looking for raw against processed and which reports will count it. Borderline inputs are where two segments can both seem plausible, which is exactly when the class wording matters. A shelf position for raw against processed that is roughly right is still wrong for reporting, because reports add shelf positions for raw against processed together without asking how rough they were.

From broad shelf to raw against processed

To narrow toward raw against processed, start broad and ask at each pair whether the boundary inside UNSPSC where an input stops being raw material and starts being a product in its own right still holds. The boundary is set by the codeset wording, not by how finished something looks to a buyer. Each yes moves the reading one level finer, and each no sends the reader back to a neighbouring pair for raw against processed. Borderline inputs are where two segments can both seem plausible, which is exactly when the class wording matters. That loop for raw against processed, repeated patiently, is the whole method, and there is no shortcut version of it for raw against processed that stays reliable.

Where raw against processed goes wrong in practice

With raw against processed, the error to watch for is borderline inputs are where two segments can both seem plausible, which is exactly when the class wording matters, set against the fact that the boundary inside UNSPSC where an input stops being raw material and starts being a product in its own right. The boundary is set by the codeset wording, not by how finished something looks to a buyer. Errors of this kind with raw against processed survive because the resulting code still looks valid, with the right number of digits and a plausible title for raw against processed. Only a check against the current codeset wording for raw against processed exposes them, which is why checking is a habit rather than a one-off task.

Before moving on from raw against processed, run one last check: does the claim that the boundary inside UNSPSC where an input stops being raw material and starts being a product in its own right still hold in the current codeset wording for raw against processed? The boundary is set by the codeset wording, not by how finished something looks to a buyer. Borderline inputs are where two segments can both seem plausible, which is exactly when the class wording matters.

The test of understanding raw against processed is simple: can the reader explain why the boundary inside UNSPSC where an input stops being raw material and starts being a product in its own right without reaching for a code first? Borderline inputs are where two segments can both seem plausible, which is exactly when the class wording matters. If the explanation of raw against processed holds, the code will follow naturally, and if it does not, no code chosen for raw against processed yet will rescue it.