The 500-Word Rule Is Wrong
The finding
The median press release runs between 613 and 879 words depending on the service carrying it — substantially longer than the 400 to 500 words almost every PR guide recommends. On no service in the study did a majority of releases fall inside that classic band.
The advice is not slightly off. It describes a minority of what actually gets published.
Length by service
| Service | Releases | Median words | In the 400–600 band | Over 1,000 words |
|---|---|---|---|---|
| Service C | 43,802 | 879 | 9.8% | 35.0% |
| Service E | 19,053 | 749 | 23.9% | 29.8% |
| Service F | 2,510 | 649 | 19.6% | 23.3% |
| Service D | 9,793 | 642 | 23.7% | 24.0% |
| Service A | 50,802 | 621 | 30.9% | 16.4% |
| Service B | 3,361 | 613 | 39.8% | 7.7% |
The spread suggests length is shaped as much by platform norms and submission templates as by editorial judgement. A sender who moves the same announcement between two services will often see it published at noticeably different lengths.
Where the 500-word rule came from
The number is not arbitrary, it is just obsolete. It descends from the era when releases arrived on paper and a page of double-spaced typescript held roughly 250 words, so two pages was the practical limit for a fax machine and an editor's patience.
Nothing about that constraint survives. Releases are read on screens, syndicated into feeds, parsed by software and increasingly summarised by AI assistants before a human sees them. None of those readers has a page limit.
What has survived is attention, and attention was never governed by word count. It is governed by how quickly the reader learns whether the story concerns them. A tight 900-word release that answers that in the first sentence outperforms a padded 450-word one that takes four paragraphs to arrive.
The rule persists because it is easy to teach and easy to check. Front-loading is harder to teach and much harder to check, which is why the industry kept the wrong advice.
Does length actually matter?
Length is a proxy, not a virtue. A long release is usually longer because it carries more boilerplate, more legal disclaimer and more product detail — not because it tells a better story.
What a journalist needs sits in the first two paragraphs: what happened, who it happened to, why it is worth anyone's attention. Everything after that is reference material for the minority who keep reading.
The practical read is not that you should write longer. It is that the 500-word rule is folklore rather than an industry norm, and that padding to hit a word count adds nothing a reporter wants.
What to do about it
- Front-load ruthlessly. Assume the reader stops after paragraph two, because most do.
- Do not pad to a target. No editor has ever run a story because the release reached 500 words.
- Put reference detail below the fold — specifications, financials, boilerplate. Length there costs nothing.
- Cut the second boilerplate paragraph. It is the most common source of avoidable length.
Where the 500-word rule came from
The number is not arbitrary, it is just obsolete. It descends from the era when releases arrived on paper and a page of double-spaced typescript held roughly 250 words, so two pages was the practical limit for a fax machine and an editor's patience.
Nothing about that constraint survives. Releases are read on screens, syndicated into feeds, parsed by software and increasingly summarised by AI assistants before a human sees them. None of those readers has a page limit.
What has survived is attention, and attention was never governed by word count. It is governed by how quickly the reader learns whether the story concerns them. A tight 900-word release that answers that in the first sentence outperforms a padded 450-word one that takes four paragraphs to arrive.
The rule persists because it is easy to teach and easy to check. Front-loading is harder to teach and much harder to check, which is why the industry kept the wrong advice.
Methodology
These findings come from an analysis of 134,126 press releases studied between 7 May and 15 August 2026 across ten distribution services. Body text was available for 129,742 of them, and only services carrying more than 2,000 releases are reported.
Percentages are calculated within each service rather than across the whole corpus, so they are unaffected by how many releases each service contributed. Services are reported anonymously and the same letter refers to the same service throughout; the spread between them is the finding, not the ranking.
Some services display contact details in a masked form rather than plain text; masked details are counted as present. A publication date is not recorded for every release, so analyses that depend on dates or times state their own smaller sample size. Figures last verified 15 August 2026. You are welcome to cite these findings with attribution to PR Gun.
Ready to distribute your press release?
Get your news in front of thousands of journalists with PR Gun's distribution network.