What changed in Section G in ISIC Rev. 5?

Section G in ISIC Rev. 5, published by the UN Statistics Division, keeps wholesale and retail as its core, with updated descriptions for intermediation, for non store and online channels, and for the boundary with services. For section g, the revision as a whole contains 22 sections from A to V, 87 divisions, 258 groups and 463 classes, and it was endorsed by the UN Statistical Commission in March 2023. Across Rev. 5, 18 new intermediation classes were added, and trade intermediation is part of that wider recognition of arranging work. Readers should use the Rev. 5 publication and national correspondence tables for any detailed mapping, since this guide only sketches the direction of change.

What stayed at the centre, and what moved around it?

Section G in ISIC Rev. 5, published by the UN Statistics Division, keeps wholesale and retail as its core. That continuity matters, because readers comparing trade statistics across revisions need a stable centre before they can make sense of changes at the edges. Around that stable centre, Rev. 5 updated descriptions for intermediation, for non store and online channels, and for the boundary with services. The revision as a whole contains 22 sections from A to V, 87 divisions, 258 groups, and 463 classes, and it was endorsed by the UN Statistical Commission in March 2023. Across Rev. 5, 18 new intermediation classes were added, and trade intermediation is part of that wider recognition of arranging work. Those numbers describe the revision as a whole and the intermediation change across the economy, not a claim that Section G alone gained 18 classes.

The direction of change can be summed up in one sentence: descriptions became clearer about arranging, about channels, and about edges, while the wholesale and retail centre stayed recognisable. Clearer intermediation descriptions for agents and brokers mean arranging without ownership is easier to place. Updated channel descriptions including online and non store retail mean newer ways of selling are described in fuller wording. Refined specialised versus non specialised store wording means the range test is stated more sharply. Consistency with the wider Rev. 5 structure of 463 classes means Section G sits coherently inside the whole tree. Sharper boundary wording with manufacturing and services means neighbouring work is separated more cleanly. That list is an orientation aid, not a substitute for the full class description, and the detail in the Rev. 5 publication should prevail where a case is unclear.

Does moving online, on its own, change a retailer’s class?

Online selling has grown since the previous revision, but growth in a channel does not mean every online seller changes class. The classification still groups mainly by buyer, goods, and store type, so a channel change alone may leave the class unchanged, and national practice should be checked. A retailer that moved from a shop to online sales should therefore check its current code against the Rev. 5 description for its goods and buyer type, rather than assuming a channel change forces a class change. That check, done with the national institute, is the safe route, because it tests the goods and the buyer instead of treating the website as a new kind of business in itself.

Slow the example down. A shop that sold clothing to households from a counter, and now sells the same clothing to the same kind of households through a website, has changed its route without changing its goods or buyer type. That continuity weighs heavily, and the channel change alone may not move the description. Online selling matters alongside goods and buyer, not instead of them.

Why do correspondence tables matter more than a summary of change?

Readers should use the Rev. 5 publication and national correspondence tables for any detailed mapping, since this guide only sketches the direction of change. Correspondence tables matter because they map old descriptions to new descriptions case by case, with the qualifications and exceptions that a summary must leave out. A summary can say intermediation gained clearer recognition and channels gained updated wording. Only the publication and the correspondence material can show how a particular older description relates to a particular Rev. 5 description, and whether a national version adds a further step in between. For time series work, that case by case mapping is essential, because a change in wording that looks small in a summary can still shift how a borderline activity is counted over time.

The same caution applies to the boundary with services and manufacturing. Sharper wording helps a reader see the edge more clearly, but it does not remove the need to read the neighbouring descriptions in full.

This guide bases its code statements about section g on the UN Statistics Division classification detail and the ISIC Rev. 5 publication in the frontmatter.

This guide sketches change in a statistical system only. It does not assign a code, decide a principal activity, or settle a registration, and confirmation of any of those belongs with the national statistical institute or business register that holds the record.