Why does ISIC cover activities, not products?

ISIC Rev. 5, UN Statistics Division, classifies what a unit does, not the individual goods or services it sells. That activity focus is intentional: a factory, a shop and a consultancy can sell overlapping items yet belong in different activity groups because their production processes differ. Product classifications exist separately, and keeping the two apart stops a reader from mistaking a product list for an industry grouping.

What changes when the unit of sorting is work, not goods?

Sorting by activity changes the question a statistician asks. Instead of asking what item left the premises, the statistician asks what kind of work was done to produce it. That shift is the whole point of ISIC Rev. 5, UN Statistics Division, and it explains results that look odd until the shift is grasped. A factory, a shop and a consultancy can sell overlapping items, yet sit in different activity groups, because their production processes differ in character, inputs and output type. The overlap is in what is sold. The difference is in how it is made, delivered or advised upon, and ISIC sorts on the difference, not the overlap.

Keeping that distinction clear also protects product statistics. Product classifications exist separately, and they do their own job well: they sort goods and services themselves, regardless of who made them. Merging the two jobs would damage both. An industry table would start to read like a catalogue, and a product list would start to pretend it described who did the work. Separation lets each system answer its own question cleanly, and lets a reader combine them knowingly rather than confusing one for the other.

How does the wider statistical system use an activity grouping?

An activity grouping does not stay inside one table. It connects to national accounts, to business registers, to product classifications and to the classifications used for occupations and trade, and its value grows when those connections are kept clean. A national accounts compiler, for example, needs activity groupings to break an economy-wide total into parts that mean something, and a business register needs them to organise units by what they do. If the activity grouping were secretly a product grouping, every connected use would inherit the confusion: accounts would split totals by item sold rather than by work done, and registers would group unlike work together merely because the outputs overlapped.

How does a statistical office apply the activity focus in practice?

In practice the focus is applied through a chain that must stay consistent from start to finish. A survey questionnaire asks about activity in words. A coding manual turns those words into categories. A published table presents the results under category titles. ISIC Rev. 5, UN Statistics Division, provides the shared reference for all three stages, so the coder, the analyst and the reader work from the same structure. The version label must travel with the data at every stage, from the first coding decision to the final chart, because a lost label is how cross-revision errors enter an otherwise careful analysis, and an activity focus cannot survive if the categories themselves are silently swapped mid-chain.

The working method for a reader follows the same chain in reverse. Start from the activity described in words, then move through the hierarchy from broad to fine, checking each level against the UNSD structure. Working through ISIC activities rather than products level by level, ask whether the definition still fits, whether an excludes list redirects the activity, and whether a finer level adds meaningful detail. For ISIC activities rather than products, an unclear answer at any level means staying at the broader level that is clearly correct and recording the question, rather than picking a fine category that cannot be justified from the evidence available.

What does the analyst’s desk show about this method?

Picture an analyst preparing a table for readers who did not collect the data. The analyst first writes the activity in plain words, then traces it through the UNSD Rev. 5 structure, reading each note, and only then records a category. A colleague later asks why a neighbouring category was not chosen, and the analyst can point to the excludes list that decided the matter. The table is published with its version label visible, so a reader in another country can check the same structure and reach the same understanding. That chain, from words to notes to a labelled result, shows the activity focus in miniature: the decision was made on the work described, the notes were the evidence, and the label lets a stranger verify it.

When doubt remains, prefer the broader category that is clearly right over the finer category that is merely plausible. A correct broad figure respects the reader. A detailed figure that cannot be defended misleads the reader while looking more helpful, and that trade is never worth making.

This guide describes statistical grouping by activity only. It does not fix any reader’s code, principal activity or registration, and for a specific record the confirming authority is the national statistical institute or business register that holds the record.