How Are Advertising and Market Research Classified?

Advertising and market research in ISIC Rev. 5, UN Statistics Division, cover creating advertising campaigns and placing advertisements, and studying markets, consumer behaviour and public opinion for clients. Advertising work includes designing advertisements, planning media and buying advertising space or time on behalf of clients. Market research includes surveys, focus groups, data analysis and opinion polling supplied as a service, where the output is knowledge about a market rather than a product for sale.

Does the firm make the message, or does it own the channel?

Start with the work itself. A firm that dreams up an advertisement, writes the copy, designs the visual, plans where it should run, and then buys that space or time for a client is doing agency work. A firm that studies shoppers, runs a survey, sits a focus group down, crunches the answers, and hands a client a report on what people think is doing research work. The distinction lives in what the firm delivers. One delivers a campaign ready to run. The other delivers findings about a market.

• Creating advertising material and campaigns for clients
• Planning media and placing advertisements on a client’s behalf
• Market research surveys and consumer studies supplied as a service
• Public opinion polling supplied as a professional service

A unit that only owns pages or airtime, and sells that space without shaping a campaign, is not doing the same job as the agency that shaped it. The UNSD entry for ISIC Rev. 5 from the UN Statistics Division sorts by the service performed, and that is the test to apply when a firm’s own description straddles both sides.

Why do media owners sit somewhere else?

A newspaper that prints advertisements is, first, a publisher. A broadcaster that runs advertising breaks is, first, a broadcaster. Their advertising revenue is real, but their main work is producing the paper or the programme that carries the advertisement. Counting them as advertising agencies would blur two different markets: the market for making and placing advertisements, and the market for audiences that media owners sell. The same logic keeps a census or an official statistical survey apart from commercial market research. One is a public statistical function. The other is a service sold to a client who wants to understand buyers or voters. The client, the funding, and the purpose all differ, even when the tools look similar: questionnaires, samples, and tabulated answers.

What happens when one firm does both jobs?

Many real firms do. The food brand example from this guide shows how. Picture an agency that designs a campaign for a food brand, buys television and online slots, and then measures how viewers responded. That is advertising work from the first sketch to the bought slot. Now picture a separate research firm that surveys shoppers, asks why they picked one brand over another, analyses the answers, and reports back to the same food brand. Its output is not an advertisement. It is knowledge the brand can act on, perhaps by changing a recipe, a price, or a pack. If a single firm offers both, statisticians look for its main activity, using the UNSD wording and the firm’s own account of what it mainly does, rather than letting the larger client or the louder service decide. That is also why a note of the exact UNSD wording consulted matters: two reviewers should be able to follow the same path and land in the same place.

What does this separation let analysts see?

Advertising spend moves early when business confidence shifts. Firms cut campaigns quickly and restart them quickly, so agency income is a sensitive signal. Research spend tells a different story: what firms are trying to learn before they commit. Keeping creation and placement apart from media ownership also stops one advertisement from being tallied twice, once as agency output and once as media output. And across the whole of ISIC Rev. 5, UN Statistics Division, the scale is large: 22 sections from A to V, 87 divisions, 258 groups, and 463 classes. A small shift in where an activity lands can therefore move a published aggregate in a way readers would notice.

How do the neighbours sharpen the boundary?

Publishing sits in Section J. Broadcasting sits in Section J too. Selling space purely as a media owner is grouped with that media activity, not with agencies. Public administration statistics belong in Section P. Reading those neighbours beside this scope often settles a doubt faster than re-reading this scope on its own, because the neighbour’s wording may describe the same work more exactly. Ask, in plain terms: does the unit mainly produce a campaign, produce knowledge about a market, produce a publication or programme, or produce official statistics? Whichever answer fits the UNSD entry in full, including its exclusions, is the answer to carry forward.

For anyone applying this to a real business, this guide is only a map of how the classification is organised. It is statistical description, nothing more. It cannot settle that business’s code, its main activity, or its registration, and those questions belong with the national statistical institute or business register that holds the record, which can weigh the full facts against the UNSD entry.