Every organisation has sent one. The bad press release examples collected in this article are fictional — labelled as such, built from placeholder names — but the failures they display are the most documentary thing in public relations, because a press release is the organisation’s record in miniature: dated, signed in effect, and filed by every desk that receives it. When a release fails, what has failed is rarely the writing. It is a discipline, broken at document scale, in public, with a dateline on it.

So this article works as an autopsy. The good format has been described positively elsewhere in these pages and will not be re-taught here. Instead, ten specimens are dissected one by one, each for the check that was skipped: the fact not yet established, the sentence a stranger could not restate, the claim no checker could verify. The pattern to watch is that none of the ten failures required talent to avoid. Each required a pause — and the pause is the cheapest instrument in the field.

The Ten Specimens

Each specimen below is an invented document, written for this article and plainly labelled, in the placeholder manner this series uses for templates. The dissection after each one names the discipline it breaks and the check that would have caught it before sending.

A note on method before the dissections begin. Each failure is stated in its simplest form, but in practice they travel together: the vague release is usually also the jargon release, the unsupported claim rides inside the badly formatted document, and the sender who cannot be contacted is rarely punctual either. That is because the failures are not ten separate accidents. They are one unfinished process seen at ten points — the discipline was interrupted, and each downstream check that would have caught the earlier failure was skipped along with it. Read the specimens in order and the interruptions compound. The compounding, not any single specimen, is what a desk actually experiences.

Vagueness

Specimen (fictional): ‘Company X, a leading innovator, is delighted to announce an exciting new chapter. Our groundbreaking solution is set to transform the industry and delight customers everywhere. Full details to follow soon.’

Nothing in the specimen is false, because nothing in it is a fact. No event, no date, no place, no person — the five questions a release exists to answer are all deferred to a ‘soon’ that performs the document’s only real work: promising. The discipline broken is the record test: a release that cannot say what happened has not yet got anything to publish, and a teaser is a promise issued against an empty shelf. The desk’s verdict is instant and correct. A document with no news in it is not an announcement but an advertisement for a future announcement, and it is filed accordingly — nowhere.

Jargon

Specimen (fictional): ‘ABC Corp today operationalised its next-generation, synergistic ecosystem leverage, unlocking scalable paradigm alignment across legacy verticals.’

The specimen’s author knew what it meant. Nobody else can, and that is the failure. The translation test applies with special force to a release, because the release’s reader is a professional re-writer: a journalist who cannot restate a sentence will not quote it, will not paraphrase it charitably, and will not rescue it — she will use someone else’s document, or nothing. Jargon elsewhere in an organisation’s writing is a fog. In a release it is a refusal to be quoted accurately, issued to the very people whose accuracy the sender is relying on. The check takes ten seconds: read the sentence to a stranger. If the stranger cannot say it back in plain words, the sentence is not finished.

Formatting

Specimen (fictional): a release that opens with three paragraphs of company history, reaches its announcement in the fourth, carries no dateline, and closes without a boilerplate — the news present, somewhere, in the wrong order.

The format of a release is not decoration; it is a checking device — headline, dateline, lead, body, boilerplate, contact, in the order a desk reads. A document that withholds its fact until the fourth paragraph asks the desk to excavate it, and desks under deadline do not excavate; they move to the next document, which respected their arithmetic. Formatting failures are the most forgivable in this collection and the most purely self-inflicted: the fact was true, checked and ready, and the document’s shape declined to present it.

What the specimen really shows is a confusion about the document’s reader. A release is not a brochure, whose job is to hold attention; it is a working paper for a person mid-triage, deciding in seconds what deserves minutes. Every convention of the form — the headline that states the fact, the dateline that fixes it, the lead that answers the five questions — exists because that reader’s time is the scarcest thing in the transaction. A sender who spends it on ceremony has not been courteous. He has been expensive.

Newsworthiness

Specimen (fictional): ‘XYZ Ltd is proud to announce the fifteenth anniversary of its founding, and looks forward to many more years of excellence and success.’

The anniversary is real, the pride presumably sincere, and the document is still an error — because the error was committed before any sentence was written, in the choice of document. An internal milestone is a fact for the staff, the shelf and the celebration; dressed as a release, it asks an editor to spend readers’ attention on the sender’s self-regard. The editor’s tests are not mysterious: is it news to the desk’s readers, is it timely in their calendar, can it be checked? The specimen fails the first and never reaches the others. Fifteen years is an achievement. It is not, without more, an event in anyone else’s world.

The Missing Public

Specimen (fictional): a single release for a specialist industrial product, addressed at once to homeowners, businesses, investors and everyone — and written, in the end, in terms none of them can act on.

A release addressed to everyone is checked by no one in particular. The discipline broken is the publics discipline: every document has a public that must act differently because of it — buy, attend, object, comply, decide — and the writing follows from naming that public and its stake. The specimen’s mismatch is not a style fault; it is a fact error about who the document is for, and it propagates. The desk cannot tell which of its readers the story concerns, so it concludes, reasonably, that it concerns none of them.

Unsupported Claims

Specimen (fictional): ‘As the best provider in the industry, with unmatched quality and unrivalled expertise, DEF Ltd is pleased to announce…’

The superlative is the claim form this series retires wherever it appears, and in a release it does particular damage. ‘Best in the industry’ is unverifiable by construction: no checker is offered a test by which it could be found false. And a checker who finds one untestable claim in a document discounts the document, not the sentence — the audit standard applied at paragraph scale. The verifiable facts travelling with it — the launch, the date, the specification — arrive pre-doubted. The specimen has spent the record’s credibility on an adjective. The check is one question: what would prove this claim wrong? If nothing could, it is not a claim. It is a costume.

Errors of Two Kinds

Specimen (fictional): a release fixing an event on ‘Friday 31 June’ in a wrongly named town — and, two paragraphs later, spelling the subject of its own announcement incorrectly.

The standard treatment of this specimen lumps all errors together. They are two kinds, and the autopsy separates them. Errors of cosmetics — the typo, the clumsy sentence — cost polish and a little dignity. Errors of record — a wrong date, a wrong name, a wrong figure — cost truth, and they travel: the desk that trusts the document republishes the error under its own name. Speed sharpens the distinction rather than blurring it: proofreading, honestly understood, is verification of the record, and only one of the two kinds is ever corrected in public — the record kind, at the sender’s expense. A checklist that weights a typo and a wrong date equally has misunderstood both.

No Way to Ask

Specimen (fictional): a complete, well-formed release that ends at its boilerplate — no contact name, no email, no telephone number anywhere on the document.

It looks like a small omission. It is the removal of the document’s correction channel. A release is an invitation to check: the desk that wants the second fact, the clarification or the interview uses the contact block, and the desk that cannot ask does the arithmetic instead — run it thin, or not at all. The contact block is also where corrections begin: when the record needs amending, the first conversation normally happens through the channel printed on the document itself, with the same discipline a statement follows. A release that cannot be queried cannot be corrected. A sender who cannot be reached has published a monologue in a dialogue’s clothing.

Timing and Frequency

Specimen (fictional): a routine product update sent at twenty to seven on the Friday evening before a holiday weekend — the sender’s eleventh release that month.

Two failures share this specimen. The first is timing: a release is dated by its burial as surely as by its dateline, and the archive preserves the choice — a desk reads a Friday-evening routine release as the sender’s own valuation of it. (Honest reasons for hard timing exist — a fact that must be on record before an event — and they declare themselves. Routine news has none of them.) The second failure is frequency: an organisation that publishes everything teaches its publics to read nothing. Listening disciplines apply in reverse — the sender who never rations its own output becomes, to its desks, background noise with a logo.

One Text for Every Channel

Specimen (fictional): one release text pasted unchanged into a social post, an email and a messaging thread — the same document, failing three different ways.

The facts of a release do not change by channel; everything else does. Tailoring is a carriage discipline: the same record is carried by the wire, the inbox, the social post and the newsroom shelf in different lengths and different furniture — the three families of instruments doing their different work. The specimen’s error is to treat distribution as photocopying. Each version failed in its own way — the post truncated the fact, the email buried it, the thread lost it — while the one thing that should survive every version intact, the checked fact itself, is the thing a hurried sender most often edits by accident. Tailor the carriage. Never tailor the record.

The Checklist, Corrected

The seven-point checklist that usually accompanies articles like this one survives the autopsy in corrected form — the same seven checks, each now carrying the discipline that owns it:

• Newsworthiness first. The editor’s test, applied by the sender before the desk applies it: news to the readers, timely in their calendar, checkable.
• The five questions answered in the lead. Who, what, where, when, why — because a release that cannot say what happened has no fact to publish.
• Every sentence translatable. Read it to a stranger; what cannot be restated in plain words is not finished.
• Headline and form in the desk’s order. The format is a checking device, not decoration.
• Proofread in two passes. One for the record — names, dates, figures — and one for cosmetics, in that order of importance.
• A named human to ask. The correction channel, printed on the document itself.
• Timing chosen, frequency rationed. The date is part of the record; the eleventh release of the month is a confession.

The list is deliberately unglamorous. Nothing on it requires a platform, a subscription or a specialist; every item can be performed by the person who wrote the document, in minutes, with the document as the only tool. That is precisely what makes its omission diagnostic. A release that fails these checks did not lack resources. It lacked the pause.

What the Autopsy Shows

Ten specimens, one pattern. Every bad release in this collection was once a good intention with a skipped check: the fact not yet established, the sentence not yet translated, the claim not yet audited, the public not yet named, the channel not yet considered. None of the skips saved anything. The correction, where one was needed, cost more than the pause would have; the burial cost the story; and the archive kept both.

There is a defensive reason to study these specimens closely: desks remember senders. A newsroom’s working judgement of an organisation is assembled from exactly such documents — the release that wasted its time, the claim it could not verify, the correction it had to chase. The specimen habit, repeated, becomes the sender’s standing in the one room where standing for accuracy is actually decided. The repair is not genius; it is reading the document once as its least friendly reader would — the date against the calendar, the claim against the evidence, the sentence against a stranger’s understanding — before anyone else has to.

That is the practical moral of bad press release examples, and it is a hopeful one. The checks are cheap, known and quick — a reader who has finished this autopsy owns all seven. What separates the release that works from the specimen is not budget, not access and not flair. It is whether the discipline was finished before the document went out.