Yes, if it is short, muted, behind a poster, and not the first thing the browser downloads. Among pages that have video, the median mobile home page ships 512 KB against an inner page’s 256 KB (HTTP Archive, 2025).
Do most websites have a video on the homepage?
No. 6.7% of mobile pages and 7.7% of desktop pages carried at least one video element (HTTP Archive, 2024).



6.7% of mobile pages and 7.7% of desktop pages had at least one video element in the June 2024 crawl (HTTP Archive, 2024).
The mobile figure is up 32%, in relative terms, from the 2022 crawl (HTTP Archive, 2024).
Desktop adoption ran a point above mobile, 7.7% against 6.7% in the same crawl (HTTP Archive, 2024).
None of these sources measures whether a video sells more. They measure what it costs the page to carry one.
How many bytes does a homepage video add?
On a phone, a median 512 KB among home pages that have video, double an inner page’s 256 KB. On desktop, 288 against 206 (HTTP Archive, 2025).
HTTP Archive, Web Almanac 2025, July 2025 crawl
View data
| Item | Value |
|---|---|
| Home, mobile | 512KB |
| Home, desktop | 288KB |
| Inner, mobile | 256KB |
| Inner, desktop | 206KB |
Read those as a ceiling: the chapter found video files downloaded more than once with different range headers, which inflates the bytes, looked worse on mobile, and is undocumented (HTTP Archive, 2025). Photos are the other half of the bill.
Median, desktop
288 KB on the home page206 KB on an inner page
Among pages that have video (HTTP Archive, 2025).
Median, mobile
512 KB on the home page256 KB on an inner page
Double, on the device most visitors use.
75th percentile, desktop
1,934 KB832 KB
The chapter puts the gap at 232.6% (HTTP Archive, 2025).
90th percentile, mobile
6,144 KB3,454 KB
The heaviest tenth of home pages that carry video.
View data
| Metric | Day zero → target |
|---|---|
| Median, desktop | 288 KB on the home page → 206 KB on an inner page |
| Median, mobile | 512 KB on the home page → 256 KB on an inner page |
| 75th percentile, desktop | 1,934 KB → 832 KB |
| 90th percentile, mobile | 6,144 KB → 3,454 KB |
Is a homepage video getting heavier every year?
Yes. The same measurement went from 410 to 512 KB on mobile home pages and 243 to 288 KB on desktop between the 2024 and 2025 crawls (HTTP Archive).
HTTP Archive, Web Almanac 2024 and 2025 Page Weight chapters, June 2024 and July 2025 crawls
View data
| Item | Value |
|---|---|
| Mobile, 2025 | 512KB |
| Mobile, 2024 | 410KB |
| Desktop, 2025 | 288KB |
| Desktop, 2024 | 243KB |
The chapter also puts video bytes on the median page up 28% in a year, 246 to 315 KB. Take it loosely: it never says which device, and its per device medians are 247 and 384 KB.



How long should the video be?
Short. More than half the videos on the web run under 30 seconds and nine out of ten under two minutes (HTTP Archive, 2024).
HTTP Archive, Web Almanac 2024 Media chapter, June 2024 crawl; only the mobile bars are labelled in the figure
View data
| Item | Value |
|---|---|
| 0 to 10 seconds | 21% |
| 10 to 30 seconds | 37% |
| 30 to 60 seconds | 21% |
| 60 to 120 seconds | 12% |
| Over 120 seconds | 9% |
21% of videos run 0 to 10 seconds and another 37% run 10 to 30 (HTTP Archive, 2024).
Only 9% of the videos found pass the 120 second mark (HTTP Archive, 2024).
The length of the video files found in the crawl. Not how long a visitor watches, and not what a longer cut does to your bookings.
How should the video be set up on the page?
Muted, behind a poster, in MP4, with no preload. Autoplay sits on 23% of video elements, muted on 8% and poster on 3% (HTTP Archive, 2024).
HTTP Archive, Web Almanac 2024 Media chapter, June 2024 crawl; the attribute present in the HTML, so a value set by script is not counted
View data
| Item | Value |
|---|---|
| autoplay | 23% |
| preload | 16% |
| playsinline | 14% |
| width | 10% |
| controls | 9% |
| src | 9% |
| muted | 8% |
| loop | 7% |
| crossorigin | 5% |
| poster | 3% |
- 01Ship it as MP4MP4 accounted for 68% of the video files found in the crawl (HTTP Archive, 2024). The next three extensions are the fragments of adaptive streaming, which is a different build with a different budget.
- 02Mute it, and keep it inlineMuted is what lets a looping clip run without hijacking the sound in a waiting room, and playsinline is what keeps a phone from opening it full screen. The attributes sit on 8% and 14% of video elements (HTTP Archive, 2024).
- 03Give it a posterWithout a poster the first frame is a black rectangle while the file loads. Only 3% of video elements declare one (HTTP Archive, 2024).
- 04Do not preload itPreload is on 16% of video elements and is down six percentage points since 2022 (HTTP Archive, 2024). Keep the clip out of what the page needs to answer the visitor, and out of what you measure your load time against.
Can someone who cannot hear the video still use it?
On almost no site. Caption track usage reached 0.4% of desktop pages and 0.2% of mobile pages in 2025, up from 0.1% (HTTP Archive, 2025).



axe-core rates a video without captions a critical failure, tagged WCAG 2.0 level A, success criterion 1.2.2 (axe-core 4.14.1).
A moderate finding, and one axe-core flags for manual review: audio that starts on its own for more than three seconds with no control to stop or mute it (axe-core 4.14.1).
Visitors who are deaf or hard of hearing first (HTTP Archive, 2025), and anyone watching with the sound off in a truck or a waiting room.
A tool’s reading of a criterion is a tool’s reading. Ask counsel what applies to you.
So: a short muted clip behind a poster, below the sentence that answers why someone called. The home page is where video costs the most.
- 01Answer the question in text
- 02Keep the clip under 30 seconds
- 03Mute it and add a poster
- 04Caption it if it speaks
Questions owners ask
Does a video on the homepage help?
No figure in these sources measures that. What they measure is the cost: among pages that have video, the median mobile home page ships 512 KB of it, double an inner page (HTTP Archive, 2025).
How big should a homepage video be?
Smaller than the median, which is 512 KB on mobile and 288 KB on desktop among pages that have video (HTTP Archive, 2025). The crawl also found video files downloaded more than once with different range headers, undocumented, which inflates those numbers.
Should the video autoplay?
Only muted. The autoplay attribute sits on 23% of video elements and muted on 8% (HTTP Archive, 2024). axe-core flags a video that autoplays audio for more than three seconds with no control to stop or mute it (axe-core 4.14.1).
Do I need captions on a homepage video?
If it has speech, yes. axe-core rates a video without captions a critical failure, WCAG 2.0 level A, criterion 1.2.2 (axe-core 4.14.1). Almost nobody does: 0.4% of desktop pages ship a caption track (HTTP Archive, 2025). Not legal advice.
Sources
- HTTP Archive, Web Almanac 2025, Page Weight chapter (July 2025 crawl, millions of sites; home page and inner page measured per site; chapter source read from the public content repo of the Almanac, commit 1e6f734, on November 21, 2026) (2025)
- HTTP Archive, Web Almanac 2025, Accessibility chapter (July 2025 crawl; caption track usage, with the chapter's own note that HLS serves captions without the track element, so the figure understates) (2025)
- HTTP Archive, Web Almanac 2024, Media chapter (June 2024 crawl; video element adoption, durations, file extensions and attribute usage; only the mobile bars are labelled in those figures) (2024)
- HTTP Archive, Web Almanac 2024, Page Weight chapter (June 2024 crawl; the 2024 half of the home page against inner page comparison) (2024)
- Deque axe-core 4.14.1, rules video-caption and no-autoplay-audio (audit engine rule definitions, not a measurement; read from the v4.14.1 tag on November 21, 2026) (2026)
- W3C, Understanding Success Criterion 1.2.2 Captions (Prerecorded) (read from the w3c/wcag repository, main branch, on November 21, 2026) (2026)
The first meeting starts with findings, not promises.The free Business Health Report reads your business and shows you what it found, in days.
Get Your Free Business Health Report



