What is a group in ISIC Rev. 5?

A group in ISIC Rev. 5, UN Statistics Division, is the third level, numbered with three digits, and Rev. 5 contains 258 groups. Groups sit between divisions and classes, splitting a division into clusters of related activities that share a production character. They are the level where many specialist tables begin to feel specific, without yet reaching the finest detail the classification offers.

What sits between a broad field and its finest detail?

A group is the third level, numbered with three digits, and Rev. 5 contains 258 groups. Groups sit between divisions and classes, splitting a division into clusters of related activities that share a production character. That middle position gives a group a particular usefulness: it is specific enough to feel like a real field of work, yet broad enough to contain several classes whose differences a non specialist reader does not yet need. A division may contain work of several quite different characters, gathered under one broad heading. A group gathers work of one character, and the classes inside it differ in detail rather than in kind.

That character test is what distinguishes a group from a mere subdivision by size or by customer. Two activities belong in the same group because the work itself shares a production character, not because they happen to be similar in scale or in the market they serve. A reader who groups by surface similarity will therefore place work in the wrong group, even when the placement looks plausible at a glance. A reader who asks what kind of work is actually done, and how it is done, will find the group boundary drawn where the production character changes, which is where the classification intends it to be drawn.

How does a group keep neighbouring classes together without blurring them?

A group holds its classes together by shared character, and holds them apart by their definitions. Inside a group, each class has its own explanatory note, its own includes list and its own excludes list, and those notes draw the lines that keep the classes distinct. A reader who reads only the group title will miss those lines and may treat the group as if it were a single undifferentiated kind of work. A reader who reads the class notes inside the group will see both the family resemblance and the differences, and will understand why the group exists as a cluster rather than as a single class.

Borderline cases test that understanding. An organisation may combine several activities, an activity may change character over time, and a new method of production may blur an old boundary, and in each case the question may be whether the work still belongs in the same group, or has moved to a neighbouring group or a neighbouring class. The tools for answering are the same as elsewhere: describe the activity in words, read the notes for both candidate categories, and note the uncertainty rather than forcing a quick answer that later proves hard to defend. A group boundary, like every boundary in the classification, is decided by definitions, not by familiarity with a title.

What happens when a group boundary shifts in a revision?

Revisions split, merge and move categories, and groups are no exception. A group may be split into two groups when a field diversifies, merged with a neighbour when the distinction between them no longer serves statistical purposes, or moved to a different division when the field it belongs to is reorganised. Each change alters what a group total contains, even when the group label survives in recognisable form. A writer comparing group totals across revisions should therefore check the UNSD structure for each revision, rather than assuming that a familiar three digit code, or a familiar group title, still covers the same work.

That check matters for the reference period as well. Revisions and national adoption dates can fall inside a series, so a series that crosses a revision boundary needs its break flagged, and a cross country comparison needs adoption dates considered, because one country may have switched basis earlier than another. Naming the version, ISIC Rev. 5, UN Statistics Division, stating the level, and flagging any break, keeps a group comparison honest. Without those details, a change in a group total may be read as a change in the field, when it is really a change in the boundary that defines the field.

This guide explains an intermediate level of a statistical classification, and it does not sort any reader into that level or any other. As statistical classification only, this guide on a group settles no reader’s code, principal activity, or registration; a specific record is confirmed by the national statistical institute or business register holding it.