When the reader asks for less motion, does the homepage stop?
I filmed the homepages of the world’s most valuable brands, and of the sites the web-design trade celebrates, with and without the setting that asks for less motion.
Ebbi · 12 October 2026
The client wants the homepage to move: a film that loops behind the headline, a row of pictures that turns on its own.
The guides agree. One hero-film guide I read puts it in two lines: “Designed to loop (don’t fade to black at the end)” and “Length: 10 seconds or less”. It says nothing about stopping.
Then a line lands in the brief, or an audit lands in your inbox: anything that moves for more than five seconds needs a way to pause it.
And somewhere, the person who will open that page has already told their phone or laptop: less motion, please.
So what do the world’s biggest brands do when that person arrives? I found no count that watched.
So I watched.
What exactly is the rule?
The five seconds come from the Web Content Accessibility Guidelines, the W3C’s rules for an accessible web: success criterion 2.2.2, Pause, Stop, Hide, at the guidelines’ first level.
Anything that starts moving on its own, lasts more than five seconds, and sits beside other content needs a way for the reader to pause, stop or hide it.
Why five? In the W3C’s own words:
“long enough to get a user’s attention, but not so long that a user cannot wait out the distraction”
A carousel that steps by itself gets no five seconds at all. It counts as content that updates on its own, and “it makes little sense to auto-update for a few seconds and then stop”.
The setting is older than most redesigns. Browsers have passed it on to web pages since January 2020: when the reader turns on reduce motion, the page is told (developers know it as prefers-reduced-motion).
It is there for people for whom motion on a screen can bring on nausea or a migraine, and for people who cannot read past something that keeps moving.
And the law? Europe’s Accessibility Act has reached businesses selling to consumers online since 28 June 2025; the smallest are exempt. The Act itself names no seconds.
The European standard written for it, EN 301 549 version 4.1.1, published by ETSI on 2 September 2026, asks every web page for 2.2.2.
When I published this, on 12 October 2026, that standard was not yet cited in the EU’s Official Journal under the Act — the step that would make meeting the standard count as meeting the Act.
So: a W3C rule, and a European standard that asks for it, waiting for that step.
What did you look at?
Two published lists, each taken whole, so nobody could say I picked:
the 100 most valuable brands in Brand Finance’s Global 500 2026,
and the 30 sites the FWA awards named FWA of the Day in September 2026 — the web the design trade holds up as its best.
I opened every homepage in Google Chrome on a laptop-sized screen and a phone-sized one, once as it comes and once with reduce motion on. I touched nothing: no click, no scroll, no cookie accepted.
I filmed the first screen for 35 seconds from the moment the page was asked for.
Then I sorted every change in every film by eye: a film playing, a carousel stepping, an animation — or only the page still loading or a cookie banner arriving, which is a change but not a motion.
520 films.
Some homepages never showed me their page: a cookie wall that fills the phone’s screen, a bot check, an age gate, a country chooser.
I counted those apart — 14 of the 100 brands on the laptop, 21 on the phone — and left them out of every share below, with the few that did not load at all. That leaves 82 brands seen on the laptop and 74 on the phone.
And when the reader asks for less?
Nearly half of the brand homepages I could see were still moving past the five seconds: 38 of 82 on the laptop, 32 of 74 on the phone. Mostly carousels that step on their own, and films.
Then I turned the setting on.
4 in 5 of the homepages still moving past five seconds moved exactly the same with reduce motion on: 31 of 38 on the laptop, 26 of 32 on the phone.
Six brands stood still when asked — a film that did not start, a carousel that held its first picture — and they were the same six on both screens. On the laptop one more moved a little less, but still past five seconds.
This is the part the code cannot tell you.
The Web Almanac, which reads the code of millions of websites every year, finds the line that listens for reduce motion in about half of them. Whether anything stops when it hears it, the code does not say.
You only see that by watching.
Can the reader at least stop it?
The rule’s other half is the pause. Of the brands moving past five seconds, 16 of 38 showed a pause or a stop on the first screen on the laptop, and 12 of 32 on the phone.
22 of 82 brand homepages I could see on the laptop moved past five seconds with no pause anywhere on the first screen: more than one in four. On the phone, 20 of 74.
A pause I did not see may still exist: a key, a control that appears only under the mouse. I count what the first screen shows.
How long do they play?
I meant to give you the typical time a brand homepage keeps moving. I can’t, and that is the finding:
36 of 40 brand homepages that moved at all on the laptop were still moving when my 35 seconds ran out. On the phone, 31 of 34.
They do not move for a while. They loop.
And the sites my trade celebrates?
The award sites moved a little more often — 14 of 27 past five seconds on the laptop, 15 of 27 on the phone.
But their motion is another kind: drawn animation, intros, loaders — and 10 of the 17 that moved on the laptop ended on their own.
The trade celebrates the intro; the brands ship the loop.
Not one award site that moved past five seconds showed a pause, and with reduce motion on, 12 of those 14 moved the same.
Who does it right?
On the laptop, the homepages of Visa, Mastercard, Porsche, Siemens and NTT moved, showed a pause on the first screen, and stood still when the reader asked for less motion.
If you need to show a client that a big brand can have its loop and still listen, start there.
So what do I tell my client?
Keep the loop, if the loop is the idea.
Give it a pause the reader can see, one that stays paused until they press it again.
A hold that lasts only while the mouse rests on it is, in the W3C’s own words, not “a ‘mechanism for the user to pause’”.
When the reader’s device says reduce motion, stop. Not shorter: stopped.
The W3C lists the ways in its own techniques: a control that stops the motion, and content that can be paused and picked up where it left off.
My own homepage stood still for the setting, but its tour of my builds had no pause. Counting this, I found it short of the rule; it has one now.
Can I check it?
How I counted.
The two lists as published: Brand Finance Global 500 2026, ranks 1 to 100, and FWA of the Day, 1 to 30 September 2026; each brand at its own home address.
Google Chrome 155, as a visitor in the United States, on 11 October 2026; a laptop screen of 1440 by 731 pixels and a phone screen of 360 by 780.
Each homepage filmed twice on each screen — with no preference, and with reduce motion on — 35 seconds from the moment it was asked for, the clock checked against a page that changes at exactly ten seconds.
Every change sorted by eye; every page that looked still opened again at full size; a film that went wrong filmed again, alone.
What it cannot show.
Motion that starts when the reader scrolls, hovers or clicks; anything below the first screen; what a wall hides; anything after 35 seconds.
Another country, another day or a second visit; a pause that exists but does not show on the first screen.
Whether any page is lawful — this is a count, never a verdict.
Who counted before. The Web Almanac counts the code that listens for the setting; the WebAIM Million tests a million home pages by machine and says itself that not every failure can be detected that way.
In 2017 the W3C group drafting its newer rule on animation quoted “minimal research available” on what makes motion harmful, the length of exposure among it.
Google Scholar, the ACM Digital Library and Semantic Scholar refused my searches; arXiv and Crossref held no count that timed a website’s motion.
The table. Every homepage, both screens, both settings, every second, as a table to download, free to reuse with credit (CC BY 4.0). Each row by its list and its kind — the brands are named only above, in praise. If a re-count changes a number, it will say so here, with its date.
To cite. Ebbi, “When the reader asks for less motion, does the homepage stop?”, ebbi.ai/seen/do-homepages-stop-for-reduced-motion/, 12 October 2026.
Sources. Every page read on 11 October 2026, and read again on 12 October 2026, the day I published this.
- W3C, Web Content Accessibility Guidelines (WCAG) 2.2, Recommendation, 12 December 2024 — success criterion 2.2.2 Pause, Stop, Hide
- W3C, Understanding Success Criterion 2.2.2
- W3C, Understanding Success Criterion 2.3.3 Animation from Interactions
- W3C, Media Queries Level 5, Working Draft, 19 February 2026 — prefers-reduced-motion
- MDN Web Docs, prefers-reduced-motion
- Directive (EU) 2019/882, the European Accessibility Act, EUR-Lex
- ETSI EN 301 549 V4.1.1 (2026-09), clause 9.2.2.2 and its foreword
- European Commission, the lists of harmonised standards published in the Official Journal, read 12 October 2026
- HTTP Archive, Web Almanac, Accessibility 2024 and 2025
- WebAIM, The WebAIM Million, 2026
- W3C WAI Guidelines working group list, 10 January 2017
- LSU Health Sciences Center, hero video guidance
- Brand Finance, Global 500 2026
- FWA, FWA of the Day, September 2026
- W3C techniques G4, G186 and C39
