Why Was the Online Versus In Store Distinction Removed in Rev. 5?

In ISIC Rev. 5, published by the UN Statistics Division, retail trade is no longer split primarily by whether a sale happens online or in a shop. A retailer selling the same goods to the same kind of customer is doing the same trade activity, whether the order arrives through a website, a telephone, or a shop counter. The channel is a way of selling, not a different kind of trade.

What Did the Old Channel Split Actually Do?

In Rev. 4 thinking, a retail sale could be sorted by where it happened: in a shop, or outside a shop, including online. That sorting felt natural when online retail was a distinct novelty, with distinct sellers, distinct shops, and distinct customer habits. A retailer with a shop counter and a retailer with a website looked like different kinds of businesses, and a channel split captured that visible difference. Over time, the visible difference faded, while the statistical problems of the split grew sharper.

Many retailers now sell through several channels at once, so a channel based split would force an artificial choice for mixed sellers. A clothing retailer with a high street shop and a website sells clothing retail in both places, using the same stock, the same buying team, and the same customer base, with orders arriving through different routes. Forcing that retailer into an online box or an in store box would misdescribe its work, because its work is clothing retail, carried out through two routes at once. The channel is a route, not a second business hiding inside the first.

Why Does Channel Sorting Damage Comparisons Over Time?

Classifying by channel would split identical businesses by their sales route, which would make trade statistics harder to compare over time. Two clothing retailers doing the same retail work, one mainly online and one mainly in store, would land in different headings, while a single retailer shifting its sales mix from counter to website would appear to change its kind of trade even though its goods and customers stayed the same. Time series built on channel headings would therefore wobble whenever selling routes changed, which they do constantly, even when the underlying retail work is stable.

How Is Intermediation Different From the Channel Question?

Online marketplaces that only arrange sales for others are intermediation, not retail. That distinction is separate from the channel question and is covered in the intermediation cluster. It is worth stating plainly here, because online selling and online arranging are often confused. A retailer that sells its own stock online is doing retail through an online channel. A marketplace that brings buyers and sellers together, without selling the goods itself, is doing arranging work. Rev. 5 keeps those two apart, even though both operate online, because one transfers goods and the other arranges transfers.

Small differences in work can therefore matter, especially where a heading boundary turns on whether something is supplied, arranged, transformed, or repaired. A business that starts as a retailer and adds a marketplace, or starts as a marketplace and begins selling its own stock, changes its work in a way that channel language alone cannot capture. The heading question must return to the work: who sells, who arranges, what is transferred, and who buys. Channel language can describe how an order arrived, but it cannot answer those heading questions on its own.

What Should a Reader Compare Instead of Channel?

Rev. 5 did not start from a blank page. Rev. 5 grows out of Rev. 4, and for online vs in store distinction the UN Statistics Division publication has 22 sections from A to V, 87 divisions, 258 groups, and 463 classes, endorsed by the UN Statistical Commission in March 2023. Around online vs in store distinction, headings were split in places, merged in places, and reworded to reflect new ways of working. That history matters for anyone comparing statistics across years, because a series that changes revision may also change heading, even when the underlying work has not changed. For online vs in store distinction, never quote a code without noting its revision, and check correspondence material before treating an old and a new series as continuous.

Removing the online versus in store split was therefore a decision to sort retail by its work rather than by its routes, and this guide has described that decision without making it a ruling for any seller. It is statistical classification background only, and it does not determine any reader’s code, principal activity, or registration. For a formal heading, the retail work should be matched to the UNSD Rev. 5 wording and confirmed with the national statistical institute or business register that holds the record.