Paid press coverage vs earned: how to tell them apart

A paid placement identifies itself if you know what to look for: a URL on a segregated path such as /press-release/, a disclaimer naming the distributor, a wire name where the byline should be, campaign parameters left in the boilerplate links, and nofollow on the links back to you. Earned coverage has none of those and carries a journalist's name. On 28 August 2026 we opened more than 30 published placement URLs from four real campaigns and read the attributes off the live pages; every one of them sat on a quarantined path.

Two things get sold under the word "coverage", and they are not variations of each other. One is a page you paid to have published. The other is a page somebody chose to write. Both can be valuable. Only one of them can be bought, and the difference is legible on the page in under a minute once you know what to look for.

On 28 August 2026 we opened more than 30 published placement URLs from four real campaigns and read the attributes off the live pages rather than off search results. What follows is what that showed.

Five markers of a paid placement

1. The URL sits on a segregated path. Every single one of the 30+ pages we checked lived on a path that quarantines syndicated content: /press-release/, /pressreleases/, /story/news/, /markets-news/Newsfile/, /community/articles/. Not one was in the outlet's editorial stream.

2. There is a disclaimer, and it is not subtle. AP News prints, above the release: "PRESS RELEASE: Paid Content from GlobeNewswire - Mobile. The AP news staff was not involved in its creation." Yahoo Finance labels it a paid press release and directs enquiries to the distributor. The Globe and Mail states that the section contains third-party material including paid content that it has not reviewed. Markets Insider adds a "PRESS RELEASE" chip and the wire's name above the headline. None of that is squeamishness: the FTC's Native Advertising: A Guide for Businesses tells advertisers that where paid content borrows the format of the editorial around it, a clear and prominent disclosure is what keeps the format itself from being deceptive.

3. There is no byline. On all 30+ pages, no human was credited anywhere. The author slot holds the network's name — "TMX Newsfile", "GlobeNewswire", "MarketersMEDIA" — or nothing at all. AP showed only a timestamp.

4. The body is your copy, verbatim. Dateline included: LONDON, May 26, 2026 (GLOBE NEWSWIRE) --, Singapore, Singapore--(Newsfile Corp. - June 9, 2026) -. It usually ends with a contact block and a line pointing at the source version.

5. It appears everywhere on the same day. Fifteen outlets publishing an identical headline within 24 hours is a distribution event, not fifteen editorial decisions.

An earned article inverts every one of these. It has a byline, an editorial URL, words the journalist chose, a publication date that answers to nothing but the news, and — the real tell — it can be unflattering.

What the links actually do

This is where the received wisdom is wrong, and the measurement is worth more than any promise. Across the same 30+ pages:

Read the first and last rows together. The same outlet — AP News — gave a dofollow link on one release and no link at all on another. The variable was not the outlet. It was which network carried the release, and on this evidence the cheaper network produced the better link.

The general shape holds across the whole sweep: the marquee names give the worst links. If search credit is the reason you are buying, the household logos are the wrong thing to pay extra for, and any vendor selling a wall of them as an SEO product is selling you the least SEO-productive part of their inventory. Our own catalogue prints the reading we took for each outlet, and "Unconfirmed" against the ones we have not opened.

A test that separates the two, in your own report

The distributor's campaign tag in the release boilerplate survives into every syndicated copy. In the campaigns we examined, the boilerplate link carried campaign parameters like utm_source=press_release&utm_medium=media_boost, and we found them on 11 live pages.

That gives you a clean test on any coverage report you are handed:

• A page carrying the campaign tag is a copy of the release you paid to distribute.
• A page covering the same news without the tag is independent coverage — somebody read it and wrote their own.

Run it on your own report before you forward it to a board. It takes a browser and the "view source" command, and it is the single most useful minute you can spend on a PR invoice.

What this looks like when it goes wrong

We ran the test on four published case studies from mediaboost.press, a competitor. Three findings, all checkable.

The trade coverage in one case study pre-dates the distribution it is credited to. The CargoAi study lists AJOT, Air Cargo Week, Global Trade Magazine, Forwarder Magazine and STAT Times under "Published across 500+ outlets, including…". Those pieces ran between 4 and 8 June 2026. The wire release is datelined 9 June. None of them carries the campaign tag that every genuine wire copy in the same campaign carries. They came from the client's own trade outreach, and they are the more valuable half of that campaign — a trade title writing about you beats a wire copy. They are simply not this distributor's results.

"500+ outlets" is one wire release fanning out. The verifiable footprint per release, opening everything findable, was roughly 12 to 20 pages — what an outlet count is actually counting, taken apart across four real distributions. Two of the four clients had also been run through two different networks days apart — one on 25 and 29 June — which produces two footprints from one announcement and one plausible route to a bigger number.

One cited link is dead. The MarketWatch URL in the Genviral study returns "Page Not Found" in a real browser. We checked it twice. That is not carelessness so much as physics: press archives rot, and a report published once is out of date within months.

None of that means the product is not real. It is: 30+ live pages, on AP News, Yahoo Finance, Business Insider's markets site, Barchart, the Globe and Mail, Benzinga, Manila Times and NewsBreak, carrying the distributor's own campaign tag — and for one client, a genuine native permalink on globenewswire.com, which is the hardest thing on this list to fake. The delivery is sound. The counting is where the claims get ahead of the evidence.

So what is each one for?

Paid syndication buys a citable record. A page on a domain you do not control, published on a date you chose, at a price you knew in advance, saying what you wrote. That is genuinely useful: it is what an investor finds when they search you, what a partner's compliance team files, and what an AI assistant reads when someone asks about your company. It is also the only one of the two that can be scheduled.

Earned coverage buys a journalist's judgement. That is the entire value, and it is exactly what cannot be purchased. It is why a paragraph in a trade title beats fifteen syndicated copies, and why no distributor — including us — can promise it.

One inconvenient closing fact: buying syndication does not cause earned coverage. In one campaign we examined, six trade titles covered the news one to five days before the wire went out. In another, the only two genuinely independent pieces came from outlets that cover that beat anyway. Neither pattern looks like a wire release generating attention, and if your plan is to buy one and wait for journalists, the evidence here does not support it. Buy the wire for the record it creates, and pitch journalists separately, as a different job with a different success rate.

One more, for calibration: Meltwater measured press-release citation share in AI search falling from 0.4% to 0.2% between April and May 2026 — a halving in a single month. Syndicated pages are a surface AI assistants read. On that measurement they are becoming a smaller one. Treat "AI visibility" as a reason it is worth having a citable page, not as a growth channel with a number attached.