Native ads posing as articles

How it works

Advertising has a credibility problem: you know it is trying to sell you something, so you discount it. Journalism, for all its troubles, still gets read as information. Native advertising rents the second to deliver the first. A brand, an agency or sometimes a government pays a news site to publish a piece that looks, in font, layout and tone, like the articles around it. Somewhere on the page sits a label: “Sponsored”, “Partner content”, “Promotion”, “Brought to you by”, in the smallest, greyest type the designer can defend.

That label is the entire legal difference between an ad and a deception, and it is designed not to be seen. In two experiments by researchers at the University of Georgia, only 7 % and 17 % of readers recognised a sponsored article as advertising; a label worked about seven times better when it said “advertising” rather than “brand voice” or “presented by”. The EU’s blacklist of unfair practices has banned unlabelled advertorials since 2005; the US FTC’s native-advertising guidance says the same. So the industry obeys the letter (a label exists) and optimises the font.

Three formats to know. The advertorial: a full article, usually a how-to or a “trend” piece, in which the answer turns out to be one product. The recommendation widget: the “You may also like” grid under real articles, sold per click to whoever pays, which is why it fills up with miracle supplements, crypto “secrets” and lookalike news sites. And the public-sector version: government “information campaigns” bought in friendly media, in Hungary and elsewhere, that present a policy as news, quote no critics, and are labelled, if at all, in the footer. The budget comes from taxes; the reader experiences it as reporting.

Example

@cityvoice_daily, 6h Five ways young families are beating rising energy bills this winter, and why the new NordWarm heat pump tops the list. (Partner content)

@anna_scrolls, 5h Four of the five ways are “buy NordWarm”. The byline is “Content Studio”. And the partner-content tag is lighter than the date.

Notice what the piece borrows: the paper’s logo, its typeface, its slot in the feed between two real stories. Notice what it skips: a journalist’s name, and anyone saying the heat pump is overpriced.

Why it works on you

• Context is credibility. You trust the page, so you trust the paragraph. The ad does not need to lie; it only needs to sit in the right place. Compare undisclosed paid influencers: the same trade in a personal voice.
• Labels are designed to be missed. A grey word in eight-point type, above the headline where nobody looks, satisfies the rulebook and almost nobody else.
• Useful content disarms you. The advertorial really does contain four tips. By the time the product arrives you have already nodded three times.

What to do

On the page: look for the label before you read, not after: top of the article, next to the byline, bottom of the page. “Content Studio”, “Brand Lab” or a company name as the author means the paper’s journalists did not write it. Then ask the native-ad question: who benefits if I believe this? If one product, one company or one policy comes out looking like the only option, read it as an ad. Treat “recommended” widgets as bought space and check where the link actually goes (see checking domains and dates).

When you catch yourself: notice when an “article” has made you want a specific thing. That is fine; ads are allowed to work. Just finish the thought: would I want it if this were labelled as an ad in big red letters?

In a comment section: do not shout “fake news” at a labelled advertorial; it is legal, and the shouting makes you easy to dismiss. “This is paid content (see the tag), so read it accordingly” is enough, and it is true.

Don’t confuse it with

• Clearly marked ads, such as banners or “Advertisement” in readable type, are not native ads. The problem is the disguise, not the sponsorship.
• Journalism that happens to favour a company may be bad journalism, but if nobody paid for the placement it is not this.