Why press releases do not count
This is the conversation we have most often. A client sends fifteen links, most of them announcements that ran on news websites, and cannot see why none of them help. The answer is independence. A press release is you talking about yourself, and no number of websites republishing it changes who wrote it.
What we say on that call
We do not argue. We send the client the actual guidelines so they can read them without us, and we walk through four things.
Independence. A company publishing information about itself is a different thing from a journalist investigating that company and deciding to report on it. Both can be accurate. Only one of them is evidence that anyone outside the company took notice.
Primary and secondary. A press release is primary material: the subject’s own account. Wikipedia is built on secondary sources, where someone independent has examined the subject and written about it.
Verification is not notability. This is the one that unlocks the call. A press release can be perfectly good evidence that a company announced a product on a certain date. It is almost never evidence that the company deserves an encyclopedia article. We are not saying a press release can never be cited for anything. We are saying it cannot carry the case.
Republication is not reporting. A release reproduced on a news website or a distribution platform does not become journalism by arriving there. What makes a staff-written report useful is the editorial work behind it: an editor who decided it was worth covering, a writer who did the work, fact-checking, and real depth on the subject.
The check we run on a link
It takes about thirty seconds. Is the piece labelled sponsored, paid or partnered. Is it independent of the subject. Does it cover the subject in depth or mention it in passing. How is it written: neutral, or promotional. Is the publication itself reliable.
Two rules we apply to ourselves while doing it. We do not assume that everything published by a major outlet was independently reported, so we read the piece and check whether there is original reporting in it or a rewritten announcement. And we do not invent policy quotations or overstate what Wikipedia requires. When we tell a client a rule, we check the current guideline and give them the link.
Sixty URLs, one story
The clearest way to think about syndication is to separate how many websites displayed a story from how many independent editorial decisions produced it. Wikipedia’s guideline for organizations and companies addresses this directly: press releases, material derived from them, and syndicated stories do not become multiple independent sources by appearing on multiple domains.
A company sends a release through a newswire. Sixty to eighty websites republish it unchanged. That produces sixty to eighty links and not one independent investigation, because the underlying material came from the company. The same arithmetic applies in the other direction, and this part is fair to legitimate journalism: a properly reported story that a hundred newspapers carry is still one story. It can be an excellent source. It is one source, not a hundred.
Paid placements, by type
The question is never only whether a client paid for publicity. It is whether what came out is independent, reliable and substantial coverage of the subject. Paid placements can be perfectly good advertising and useful marketing. They are a different product from reported journalism, and the publication’s reputation does not make the two interchangeable.
Contributor and executive council programs. Programs such as Forbes Councils, the Entrepreneur Leadership Network and the Fast Company Executive Board are paid memberships that let members publish articles under the title’s banner. Our clients have paid anywhere from about a thousand dollars to ten thousand for these, depending on the programme. They fail as evidence of notability because the article is written by the subject, ghostwritten for them, or written by someone close to them, and it carries their opinion. Wikipedia’s perennial sources list treats contributor content at major business titles as generally unreliable. A contributor label does not automatically make a piece useless for every purpose, but a self-authored or promotional article is weak evidence that anyone independent noticed you.
Sponsored articles and advertorials. Advertising presented in article form. Many publishers label them as sponsored, paid content or brand voice. They fail because payment decided both that the piece exists and what it says. Publication on a respected domain does not turn an advertisement into journalism.
Wire distribution. These services describe themselves accurately: they distribute and amplify announcements across a network. The announcement originates with the subject, and multiplying its URLs multiplies nothing else.
Pay-to-feature packages. A profile, an interview, a cover, social promotion, often sold as a bundle. The thing to check is the actual editorial arrangement. A paid profile can give someone a great deal of space and still have no editorial independence behind it, and length alone establishes nothing.
Interviews and podcasts
Clients are sure an interview in a real publication is coverage. Here is the line. An interview proves that a publication or a show gave you a platform to speak. What Wikipedia looks for is independent coverage that meaningfully examines your work or your significance. The existence of the interview is not that.
A paid or sponsored interview is weak, because the editorial independence is not there. An interview on the client’s own channel is not independent at all. A contributor question-and-answer piece in a recognised publication is usually weak, especially where the subject controlled the answers. An independent journalist interviewing a subject as part of a substantial reported story can genuinely help, depending on the depth of the article and how independent it reads: if it glorifies the subject, it will not survive review. An independent podcast discussing someone’s work in depth can support a case without carrying it. And several interviews repeating the same biography add nothing, because there is no independent analysis anywhere in them.
What does count
News. Staff-written reporting with substantial focus on the subject, not a passing mention and not a rewritten release: wire services such as the Associated Press and Reuters, and newsroom reporting at established titles. For a title that runs both a newsroom and a contributor network, which side of the house the piece came from matters more than the name on it.
Trade press. In-depth features in established industry publications, where there is independent editorial work and real analysis of what the subject contributed. Wikipedia applies particular caution to trade press when the subject is a company, so these support a case more often than they carry it.
Books. Books from established independent publishers that give the subject meaningful space. A passing mention, a self-published biography, or a book the subject commissioned is weak evidence. For authors, this is read against the guideline for creative professionals.
Academic. Peer-reviewed research, academic books, scholarly reviews, and independent scholarly attention to the work. We check citation record and h-index alongside the specific tests in Wikipedia’s academic notability guideline, which is a different route to an article from the general one.
Archives. Older newspaper coverage that never made it onto the open web. This is the category most often missed, and the one we search hardest, because for anyone with a career behind them it is where the real coverage usually is.
How many do you need
There is no fixed number, and anyone who gives you one is guessing. We have seen a draft accepted on the strength of one genuinely independent piece of coverage, and we have seen drafts carrying more than ten independent links declined. The guideline asks for multiple independent, reliable sources with significant coverage, and depth settles it rather than count.
What that means in practice depends on who you are. A company is the hardest case: we look for several unrelated, substantial reports focused on the company itself, because the organizations guideline is particularly strict about what counts as significant coverage. A person needs independent profiles and real reporting on their career, achievements or impact, rather than interviews and passing mentions. An academic is judged on scholarly impact: influential research, citations that mean something in context, major contributions, and recognition from independent scholars. Publishing a great many papers helps, and on its own it is not enough.
When your PR agency says the coverage is enough
This one causes more friction than any other conversation we have, so here is how we handle it. We do not criticise the agency or dismiss its work. What the agency has been doing may be working perfectly well for what it was meant to do. Media visibility and Wikipedia notability are two separate tests, and clearing one does not clear the other.
So we take it out of the realm of opinion. We read every article the agency delivered and mark each one: usable or not usable, and if not, why not. That is a yes or no per link with a reason attached, and the client can see the reasoning themselves. Where the client agrees, we bring the agency into the conversation to confirm which placements were sponsored and which material was syndicated, because they know and we are guessing otherwise. Then we point the next round of effort at coverage that would count.
We never promise an approval to end the argument, and we never tell a client their previous spend was wasted. What we do say is this: a volunteer reviewer will apply the same test we just applied, and if we wave a weak link through to keep the peace, the reviewer will not.
How coverage that counts is actually earned
A journalist writes about you when there is a story they want to tell. That is the whole mechanism. You cannot buy the decision, you can only give someone a reason to make it, and whether they make it depends on their interest, their publishing slots and their judgement on the day. We do not control that and we will not pretend otherwise. What we can do is help work out what the story actually is and who would want it, which is a different piece of work from the drafting.
The one thing to remember
A pile of press releases does not mean you have nothing, and it does not mean you have enough. Some links in that pile may well be usable. The only way to know is for someone to read every one of them and mark it, usable or not, with the reason written down. That is what the source review is, and it comes before anyone asks you for money.
Questions people ask
Why do press releases not count as Wikipedia sources?
Because they are not independent. A press release is the subject talking about itself, so it can verify that a company announced something, but it cannot show that the world took notice. Wikipedia asks for sources written by people with no stake in the subject. A press release fails that test whichever website it ends up on.
A press release ran on fifty news sites. Does that count as fifty sources?
No. The question is not how many websites displayed a story but how many independent editorial decisions produced it. A wire release republished unchanged on sixty sites creates sixty URLs and no independent investigations, because the material still originates with the company. Even a genuinely reported story that many papers republish counts as one source, not dozens.
Do Forbes Councils or similar contributor articles count for Wikipedia?
Usually not. Contributor and executive council programs are paid memberships that let members publish articles under the publication’s name. The piece is often written by the subject or on their behalf, so it is not independent reporting, and Wikipedia’s perennial sources list treats contributor content at major business titles as generally unreliable. A respected domain does not make a self-authored article independent.
Does a sponsored article or paid feature count?
No. If payment decided that the article exists and shaped what it says, it is advertising in the shape of an article, however good the publication. Length does not help either: a long paid profile is still a paid profile. Many publishers label these as sponsored, paid content or brand voice, which is the first thing to check.
Do interviews and podcasts count towards notability?
An interview shows that a publication gave the subject a platform to speak, not that anyone independently examined their work. Interviews are mostly the subject’s own words, so they are primary material. A journalist interviewing someone as part of a substantial reported story can help. Several interviews repeating the same biography add nothing, because there is no independent analysis in them.
What sources do count for a Wikipedia article?
Staff-written reporting with real editorial oversight that covers the subject substantially, not in passing: wire services and newsroom reporting at established titles, in-depth features in recognised trade publications, books from independent publishers that give the subject meaningful space, and for academics, peer-reviewed work, scholarly reviews and citation record. Older newspaper archives count too, and are the most commonly missed.
How many good sources does a Wikipedia article need?
There is no fixed number. We have seen a draft accepted on the strength of one genuinely independent piece of coverage, and we have seen drafts with more than ten independent links declined. The guideline asks for multiple independent, reliable sources with significant coverage, and depth decides it rather than count. Companies are held to the strictest reading, people need independent reporting on their career and impact rather than interviews and mentions, and academics are judged on scholarly impact rather than the number of papers published.
My PR agency says our existing coverage is enough for Wikipedia. Who is right?
Media visibility and Wikipedia notability are two different tests, and an agency can deliver real publicity that still does not satisfy the second one. The way to settle it is to review the actual articles one by one and mark which meet the standard and which do not, with a reason for each. That is a yes or no per link, not an opinion about the agency’s work.
Guides
Policies cited
Ask whether a page is possible
Send a name and anything written about the subject by someone else. You get a written source review before any payment.
Book a review call