The Wire Clock

01What we found

1. 01

   The wire has exactly one rush hour, and it is 10:00 local time.

   9.2% of every release lands in that single hour, and 26.1% between 09:00 and 11:00. The whole stretch from 02:00–05:00 holds 3.5% combined. Give every source an equal vote instead of counting releases, and the peak does not move — so this is a property of the wire, not of its largest publishers.

2. 02

   A third of the wire is scheduled to a quarter-hour.

   33.5% of releases carry a timestamp of exactly :00, :15, :30 or :45 — 5.0 times what arbitrary minutes would produce. The top of the hour alone takes 19.6%. The wire is not a stream of events. It is a broadcast schedule, and it is denser than any newsroom's.

3. 03

   The Friday news dump does not exist. Friday does not shift late — it fails to start.

   Friday runs below the Monday–Thursday rate at almost every hour of the day. The morning loses 20.2% of its usual volume; the afternoon loses only 6.2%. That gap is the whole illusion: a day that shrinks unevenly feels like a day that shifts. In the United States there is no dump at any hour — 16:00 on a Friday, the slot the theory is named after, runs 38% below a normal weekday 16:00.

4. 04

   National publishing clocks are fourteen hours apart.

   Among the 16 countries whose peak hour survives both of our robustness tests, the busiest local hour runs from 07:00 (Quebec (Canada), Norway) to 21:00 (China). These are not different time zones — every figure here is already in the publisher's own local time. They are different working days.

5. 05

   Europe leads the world's newsflow for three hours a day.

   Counted by which region supplies the largest share of each UTC hour, Asia-Pacific leads 8, Europe/Middle East/Africa leads 3, and the Americas lead 13. The quietest hour on the planet is 23:00 UTC. A newsroom that staffs to a local clock is asleep for most of the wire.

02The day

Put every release in the local time of the organisation that issued it and the world's newswires turn out to share one working day. It starts sharply — the wire goes from 3.5% across four small hours to its 10:00 maximum in about five hours — holds a broad plateau through the afternoon, and drops off after 18:00.

The sharpness is the surprising part. Communications is a global, always-on industry that has been told for a decade to publish when its audience is awake, and the audience for a press release is worldwide. It publishes at ten in the morning anyway, wherever it is.

The dashed line on the chart below is the same distribution with every source counted once rather than every release counted once. It matters because one publisher supplies 16.8% of this window on its own. If the peak were that publisher's schedule, the two lines would separate. They do not.

03The week, and the myth in it

Thursday is the busiest day of the wire's week at 19.6% of volume; Friday is 16.6%, or 15.4% below Thursday. The weekend is not a slow period, it is a closed one: an average weekend day carries 20.8% of an average weekday.

That sets up the industry's most durable belief. The "Friday news dump" holds that organisations bury unwelcome news late on Friday, when reporters have filed and readers have left. It is repeated in media-training decks, in crisis-comms playbooks, and in newsroom folklore.

It is not in the data, and the reason it survives is a measurement error that is easy to make. Ask what share of Friday's releases went out after 15:00 and Friday does look later than midweek — because Friday's morning collapses hardest, the afternoon's share of a smaller day rises. Ask instead how many releases went out per Friday afternoon against per midweek afternoon, and the effect inverts. Friday is down at nearly every hour. It is down most between 08:00 and 10:00.

We tested all 23 sectors in the dataset on the same rate basis. None of them publishes meaningfully more into Friday evening than it does midweek once its largest single feed is removed. The dump is a habit the industry believes it has and the timestamps say it does not.

04The minute hand

Nothing in the wire's behaviour is as strange as its minutes. If publishing times were set by when work happened to finish, every minute of the hour would hold about 1.7% of releases. Instead the top of the hour holds 19.6%, and the four quarter-hours together hold 33.5%.

Some of that is machinery rather than intent — distribution platforms release in batches, and a batch stamps hundreds of releases with one time. The second-tallest bar in the chart below is exactly that: it is not a decision anyone made, it is one large platform's carousel firing at the same point past the hour, and with that platform removed the minute drops below what chance would give.

But the quarter-hours are not machinery. They are the visible edge of an industry that books its announcements the way television books its slots — and a reminder that "when a company announced something" is usually a calendar entry, not an event.

05The map and the relay

Countries keep genuinely different clocks, and the differences are not time zones — everything below is already local. Some national wires are extraordinarily compressed: the tightest keeps a fifth of its entire output inside one hour. Others spread across the working day and into the evening.

Two cautions govern this section, and both are in the data rather than the footnotes. First, a country's peak can belong to a single publisher: where one service supplies most of a country's volume, its release schedule IS the national curve. We test every country's peak by dropping its largest source and by giving every source an equal vote, and mark the 9 rows that fail. Second, a release's country in this archive is frequently the country of the feed rather than of the story — a US distributor's foreign-language channel is tagged to the market it serves, not the desk that filed it. Read the map as where the wire is operated, not where the news happened.

Stacked by UTC hour, the day becomes a relay. Asia-Pacific holds the small hours, Europe takes a narrow morning window, and the Americas hold the rest — more than half the clock. For anyone monitoring the wire, that is the operational finding: there is no hour in which nothing is published, and no single working day that covers it.

06Questions this answers

Three things practitioners ask about timing, answered from the timestamps rather than from received wisdom.

Is there a best time to send a press release?

There is a most CROWDED time, which is not the same thing. 9.2% of all releases go out in the 10:00 hour and 26.1% between 09:00 and 11:00 local. If you publish then you are competing with the densest part of the wire; if you publish outside it you are competing with less but reaching a desk that may not be staffed. This report measures supply, not attention — it can tell you where the crowd is, not whether standing outside it works.

Should we bury bad news on Friday afternoon?

The timestamps say the industry does not, whatever it believes. Friday runs below midweek at nearly every hour, and in the United States 16:00 on Friday is 38% below a normal weekday 16:00 — one of the emptiest slots of the week rather than a busy one. A release sent into it is unusually visible, not unusually hidden.

Does the wire ever stop?

No. The quietest hour of the global day is 23:00 UTC, and it is quiet, not empty. Weekend days run at 20.8% of a weekday. Any monitoring set-up scoped to one country's business hours is missing the majority of the wire by construction.

07Method

This report covers 1 July – 31 August 2026 — 62 complete days. It begins from every press release PPN World archived in that window from 1,423 sources, and reports 268,373 of them for hour-of-day analysis and 247,286 for local-time analysis. The difference between those numbers is three exclusions, each made to stop the report measuring itself.

NO USABLE TIMESTAMP — 7,702 releases (2.59%). When a feed supplies no parseable publication time, our parser records the moment we fetched it and marks that substitution. Counting those would have plotted our own ingestion schedule and labelled it the world's.

DATE-ONLY SOURCES — 21,589 releases (7.25%) from 132 sources. A feed that publishes a bare date gets pinned to a fixed hour. One government feed contributed over two thousand releases stamped, every one of them, at the same minute. These sources are excluded from hour and minute figures and kept for weekday and volume, where their data is real.

NO TIME ZONE — releases from countries without a representative zone in our map are excluded from local-time figures only.

Local time uses one representative IANA zone per country, DST included. For multi-zone countries that zone is where the publishing actually originates — the United States is mapped to Eastern, not to a national average. Weekday and hour comparisons are computed as RATES (releases per Friday, per Monday–Thursday day) rather than shares, because a share cannot distinguish a busier afternoon from a quieter morning.

Every country row clears a floor of 1,000 localisable releases from at least three distinct sources, and every peak is stress-tested twice: once with the country's largest source removed, once with every source weighted equally. Peaks that fail either test are marked. Sector significance is a two-proportion z-test with a Benjamini–Hochberg correction across all sectors tested; without that correction six sectors would appear significant by chance alone.

What this cannot tell you

• This is a census of what PPN World collects, not a random sample of the world's press releases. It is weighted toward sources with machine-readable feeds and away from those without, and toward English, Japanese, German and French.
• Volume is not comparable across the window. Our source panel grew during it, so this report makes no claim about whether the wire got busier — only about how the releases in it are distributed across the clock, which a growing panel does not distort in the same way.
• A release's country is often the country of the feed rather than of the news. Around two in five carry a country inherited from their source's home market.
• Sector labels are assigned by a classification cascade whose accuracy is currently unmeasured. Sector figures in this report are the least certain thing in it, and the Friday result is stated as a null — no sector shows an effect — which is the direction least sensitive to that uncertainty.
• Publication time is not attention. Nothing here measures whether anyone read the release.

08The data

Everything in this report is downloadable. The CSV is tidy long-format — one row per dimension, key, bucket and value — which is the shape that goes straight into a notebook or a spreadsheet pivot without reshaping.

Embed a chart

Every figure can be embedded on your own site. The embed carries the attribution with it, so crediting is automatic, and it renders without JavaScript.

<iframe src="https://www.ppnworld.com/report/wire-clock/embed/day" width="100%" height="460" style="border:0" loading="lazy" title="The wire has one rush hour, and it is 10:00 — PPN World"></iframe>
<p>Source: <a href="https://www.ppnworld.com/report/wire-clock">PPN World — The Wire Clock</a> (CC BY 4.0)</p>

<iframe src="https://www.ppnworld.com/report/wire-clock/embed/week" width="100%" height="460" style="border:0" loading="lazy" title="The working week, hour by hour — PPN World"></iframe>
<p>Source: <a href="https://www.ppnworld.com/report/wire-clock">PPN World — The Wire Clock</a> (CC BY 4.0)</p>

<iframe src="https://www.ppnworld.com/report/wire-clock/embed/countries" width="100%" height="760" style="border:0" loading="lazy" title="Every country keeps its own clock — PPN World"></iframe>
<p>Source: <a href="https://www.ppnworld.com/report/wire-clock">PPN World — The Wire Clock</a> (CC BY 4.0)</p>

<iframe src="https://www.ppnworld.com/report/wire-clock/embed/friday" width="100%" height="460" style="border:0" loading="lazy" title="Friday does not dump — it fails to start — PPN World"></iframe>
<p>Source: <a href="https://www.ppnworld.com/report/wire-clock">PPN World — The Wire Clock</a> (CC BY 4.0)</p>

<iframe src="https://www.ppnworld.com/report/wire-clock/embed/minute" width="100%" height="460" style="border:0" loading="lazy" title="A third of the wire is scheduled to the quarter-hour — PPN World"></iframe>
<p>Source: <a href="https://www.ppnworld.com/report/wire-clock">PPN World — The Wire Clock</a> (CC BY 4.0)</p>

<iframe src="https://www.ppnworld.com/report/wire-clock/embed/relay" width="100%" height="460" style="border:0" loading="lazy" title="The 24-hour handover — PPN World"></iframe>
<p>Source: <a href="https://www.ppnworld.com/report/wire-clock">PPN World — The Wire Clock</a> (CC BY 4.0)</p>

How to cite this

If you use these figures, a link back to this page is the attribution we ask for. Suggested citation:

PPN World (2026). The Wire Clock, Edition 1: press release publication timing, 2026-07-01 to 2026-08-31. https://www.ppnworld.com/report/wire-clock

Licence

The data and charts in this report are published under Creative Commons Attribution 4.0 (CC BY 4.0). You may republish, redraw, remix and use them commercially, including in a paid publication, provided you credit PPN World and link to this report. The underlying press releases remain the property of their issuers. CC BY 4.0

Corrections

1 September 2026 — data correction; no finding changed. The first extraction paged the archive by timestamp alone, which silently skips rows sharing the boundary timestamp at the end of a page. That is not a rare edge in this archive: more than half of it is stamped on an exact minute, and feeds publishing a bare date put thousands of items on one timestamp. The loss was therefore concentrated rather than spread — a spot check found one government source short by 23% of its rows while most sources had lost none. The extraction now pages one day at a time and reconciles every single day against an exact database count, so a short pull fails the build instead of quietly becoming a report. Every figure on this page is rebuilt from that verified-complete extraction. We re-checked each headline against it: the peak hour and its share, the morning block, the regional split, the weekday shape and the Friday result are all unchanged. Future corrections will be listed here with their date rather than edited quietly into the page.