Video SEO is the practice of making video content findable through search, both on Google and inside video platforms. The gap between publishing video and making it findable is wider than most teams expect: roughly 6.40% of desktop pages carry a video element, while only about 0.34% carry the VideoObject markup that makes a video eligible for video results. That is video outnumbering video markup by close to 19 to 1, measured across 16.2 million websites by the HTTP Archive Web Almanac.
Below are 50 statistics on how video actually performs in search, each one traced back to the organisation that produced it, with the year it was measured.
We produce explainer and product video for B2B SaaS and tech teams at Vidico, so we track this data to decide where video is worth making. Several of the most-quoted figures in this category turned out to be retired, and we have said so where that applies.
Key Takeaways
- Video appeared in 38.5% of tracked Google results on 20 July 2026, the lowest reading in 90 days and down from a 53.7% peak. Video presence in search is currently trending down, not up.
- Only about 0.34% of web pages use VideoObject markup, against 6.40% of desktop pages carrying video. The eligibility gap is the largest fixable problem in video SEO.
- Google requires the watch page to already rank before it considers the video for indexing. Video SEO is downstream of page authority, not a shortcut around it.
- YouTube is the most-cited video source in AI answers by a wide margin, and long-form video takes 94% of those citations while Shorts take 5.7%.
- Several of the most-repeated video SEO statistics are retired. The “53x more likely to rank” figure was retracted by its own author in 2012 and is still being republished with recent-looking dates.
How Often Video Actually Appears in Google Results
Video appears in roughly four out of ten tracked Google results, and that share has been falling through 2026.
1. Video results appeared in 38.5% of tracked Google search results on 20 July 2026. That reading comes from the MozCast Feature Graph, which tracks 10,000 hand-picked keywords daily across 20 industry categories and five US cities, page one only. Because the keyword set is hand-picked and skews commercial, this is not the same as 38.5% of all searches.
2. That 38.5% figure is the lowest of the preceding 90 days. MozCast’s 90-day mean across the same window was 46.8%, with a high of 53.7%. Video presence in results declined through the second and third quarters of 2026.
Video appeared in 38.5% of tracked Google results on 20 July 2026, the lowest reading in 90 days. Source: MozCast Feature Graph.
3. Video sits mid-pack among search features. In the same MozCast sample on the same day: Knowledge Graph 85.4%, related questions 84.7%, review stars 61.2%, verticals 49.5%, video 38.5%, sitelinks 24.4%, images 19.0%.
4. The video carousel triggered for roughly 40 million mobile queries against 8 million desktop queries. seoClarity’s Research Grid analysis covering July 2023 to June 2024 found mobile triggers the carousel around 400% more often than desktop. These are raw counts rather than rates, so they support the mobile-desktop gap but not a prevalence percentage.
5. Video carousel appearances grew roughly 75% year over year across the same seoClarity period.
6. 83% of results in Google’s video carousel came from YouTube. Perficient’s video ranking study, published December 2020, matched about 6,100 queries across Google and YouTube. This is the most recent study on the question with disclosed methodology.
7. Video carousels appeared less often alongside AI Overviews in late 2025 than they had in March of that year. Semrush’s AI Overviews study, published December 2025, tracked that decline while overall AI Overview presence rose.
YouTube as a Search and Discovery Engine
YouTube is a search destination in its own right, and the honest evidence for that is usage data rather than the “second-largest search engine” claim.
8. youtube.com is the second most-visited website in the world, behind google.com. That ranking comes from Similarweb’s top websites list using June 2026 data.
9. The average YouTube visit runs 18 minutes 55 seconds across 11.96 pages, with a 23.15% bounce rate. Similarweb recorded those engagement figures in the same June 2026 dataset.
10. 87% of YouTube users said the site is important for figuring out how to do things they have not done before, and 51% called it very important. Pew Research Center surveyed 4,594 US adults for this in 2018. It has not been repeated since, so treat it as the best available measure rather than a current one.
11. 84% of US adults use YouTube, and 33% visit several times a day. Pew reported this in its 2025 social media use survey, making YouTube the most widely used platform it tracks.
12. 90% of US teens aged 13 to 17 use YouTube, and 73% use it daily. Pew published these figures in February 2025.
13. 32% of US adults regularly get news from YouTube, up from 23% in 2020. Pew recorded that increase in its 2025 news platform research.
14. More than 20 million videos are uploaded to YouTube every day. That figure comes from the YouTube press page as of January 2026. It replaces the widely quoted “500 hours every minute,” which YouTube no longer publishes.
15. YouTube Shorts averages more than 200 billion daily views. YouTube reported this in its 2026 leadership letter.
16. More than 20 million people used YouTube’s in-video AI “Ask” feature in December alone. YouTube disclosed this in the same 2026 letter. No methodology was published, but it is first-party and directly search-adjacent.
17. YouTube advertising revenue reached $9,883 million in the first quarter of 2026, up 11% year over year. Alphabet’s quarterly earnings release reports YouTube ads as its only broken-out YouTube line. Subscription revenue sits elsewhere, so any “YouTube total revenue” figure is an outside estimate.
18. YouTube holds 13.4% of all US television watch time, the largest share of any single distributor. Nielsen’s The Gauge measured this for April 2026. Nielsen notes The Gauge does not reflect its currency TV ratings, and a methodology update is planned for late 2026.
Is YouTube the Second-Largest Search Engine?
Not according to any current measurement. The claim traces to comScore’s November 2008 US search rankings, where a combined “YouTube and all other” bucket exceeded Yahoo. It was never YouTube measured on its own, it covered the US only, and comScore retired the series years ago. No comparable measurement has been maintained since. YouTube’s role in search is well supported by Similarweb ranking it the second most-visited site globally and by Pew finding 84% of US adults use it. The specific “second-largest search engine” claim is not.
Video Schema, Structured Data, and Indexing
Video markup is the least covered part of this topic and the widest gap in it. Video outnumbers video markup on the web by roughly 19 to 1.
19. Only about 0.34% of web pages use VideoObject structured data. The HTTP Archive Web Almanac measured this in its structured data analysis. The Web Almanac is community-run and non-commercial, publishes its full methodology and SQL, and carries a DOI.
20. 6.40% of desktop pages and 5.68% of mobile pages contained a video element, per the Web Almanac’s SEO analysis, up from 5.87% and 5.13% the previous year. Set against the 0.34% markup figure, video outnumbers video markup by roughly 19 to 1. Because this counts the <video> element, YouTube iframe embeds may not be captured, so real video presence is likely higher and the gap likely wider.
6.40% of desktop pages carry a video element, while only 0.34% carry the VideoObject markup that makes a video eligible for video results. Source: HTTP Archive Web Almanac.
21. 50% of pages now contain structured data of some kind, up from 48% on desktop and 49% on mobile the year before. The Web Almanac measured 16,213,084 websites in its July 2025 crawl.
22. JSON-LD appears on 43% of home pages, against Microdata at 17% and RDFa at 1%, per the Web Almanac’s 2025 crawl. Only 2% of crawls deliver structured data through JavaScript rather than raw HTML.
23. VideoObject does not appear anywhere in the Web Almanac’s list of most-used schema types. The top types are WebSite at 37%, SearchAction at 28%, Organization at 26.7%, and WebPage at 25%.
24. Google requires that “the indexed watch page must be performing well in Search before its video can be considered for indexing.” That is verbatim from Google Search Central’s video documentation. It reframes video SEO as downstream of page authority rather than a way around it.
25. Google defines a watch page as one whose “main purpose is to show users a single video.” That definition sits in Google Search Central’s video documentation. A page where video sits alongside substantial text and other elements does not qualify.
26. VideoObject has exactly three required properties: name, thumbnailUrl, and uploadDate. Google’s video structured data reference lists contentUrl, description, duration, and regionsAllowed among the recommended additions, with contentUrl preferred over embedUrl.
27. Google states plainly that it “doesn’t guarantee that adding markup will result in a specific video feature in Search results.” Markup establishes eligibility, not placement.
28. Google restricted video thumbnails to pages where video is the main content, announced 13 April 2023 and extended to Video mode on 4 December 2023. The December announcement added a Search Console reason string, “Video is not the main content of the page,” and told publishers to expect fewer indexed videos. Google published no impact figures. Blog posts with a supporting video are the exact pattern this change excludes.
Teams building this into their workflow usually pair it with structured video content creation so every asset ships with schema and captions built in.
Video and Click-Through Behaviour
A video block on the results page lifts the top result modestly. Whether your own thumbnail helps is unmeasured.
29. The first organic result earned 16.90% click-through when a video block was on the page, against 14.61% when it was not. Advanced Web Ranking’s SERP feature data for May 2026 puts that at a 2.29 point gain, or about 16% in relative terms, drawn from 113.7 million impressions on the video series.
Click-through at positions one and two, with and without a video block on the results page. Source: Advanced Web Ranking, May 2026.
30. The second position moved the other way: 8.45% with a video block against 9.24% without. Advanced Web Ranking’s data shows the lift concentrating on position one and slightly suppressing position two.
31. From position four down, the difference is noise. Positions four through ten differ by hundredths of a percentage point in the same dataset.
32. Sistrix found the same pattern independently. Its click-through study, built on more than 80 million keywords of mobile search data, reports that when video integrations appear, “more users click on the first organic result than average, while the click rates on the remaining rankings remain roughly the same.”
33. No study isolates the effect of a video thumbnail on a single result’s click-through rate. We looked specifically for one. Every claim of the “thumbnails lift CTR 30 to 40%” variety traces to vendor marketing with no sample size or method. The direction may also run backwards: queries that trigger video blocks are a different intent mix than queries that do not.
Captions, Transcripts, and What They Actually Do
Captions have a strong evidence base for engagement and accessibility. Their SEO benefit is asserted far more often than it is documented.
34. Neither Google Search Central nor YouTube Help documents captions or transcripts as an input to video indexing or ranking. We checked Google’s video documentation, its video structured data reference, and YouTube’s own help articles on captions. YouTube’s stated benefits are accessibility and reach to viewers in other languages. The “captions boost rankings” claim traces to captioning vendors, not to either platform.
35. 80% of people said they are more likely to watch a video to the end when captions are available, and 80% of caption users reported no hearing impairment. The Verizon Media and Publicis Media study surveyed 5,616 US adults aged 18 to 54 in 2019. These are stated intentions rather than measured behaviour, and both commissioning parties sold video advertising.
36. 69% said they watch video with sound off in public, and 25% do so in private. Verizon Media and Publicis Media reported this in the same 2019 research. It is the properly sourced replacement for the “85% of Facebook video is watched without sound” claim, which Facebook never published.
37. Captions raised ad recall 8% and ad memory quality 10% for mobile viewers, but pairing captions with brand logos lowered ad memory by 12%. That negative finding, from the same Verizon Media and Publicis research, is almost never quoted alongside the positive ones.
38. 36% of US adults always or usually use subtitles, rising to 56% among under-35s and falling to 20% among over-55s. CivicScience polled 1,774 people in April 2024.
Video in AI Search
AI answers cite video heavily, but almost all of it is YouTube, and almost all the measurement comes from vendors selling AI visibility tools.
39. AI Overviews triggered for 6.49% of queries in January 2025, rising to 15.69% by November 2025. Semrush’s AI Overviews study tracked more than 10 million keywords across that period. Semrush sells AI search tooling, so treat the magnitude as directional.
40. Video carousels appeared on 25.65% of results pages that also carried an AI Overview, per the same Semrush study, typically featuring YouTube and Reddit. Worth being precise: this counts a separate feature on the same page, not video shown inside the AI Overview.
41. 5.54% of all AI citations went to social and video domains, and YouTube took 31.8% of those. Otterly.ai analysed more than 100 million citation instances over 30 days, published March 2026. By platform: Perplexity 38.7%, Google AI Overviews 36.6%, AI Mode 19.6%, ChatGPT 4.4%.
42. Long-form video took 94% of YouTube’s AI citations, with Shorts at 5.7%. Otterly also found views, likes and subscriber counts showed almost no relationship to how often a video was cited, at a correlation of roughly -0.03. Popularity is not what gets a video cited.
Long-form video takes 94% of YouTube’s AI citations, with Shorts at 5.7%. Source: Otterly.ai, March 2026.
43. YouTube was the single most-cited source in health AI Overviews, at 4.43% of all citations. SE Ranking analysed 50,807 healthcare prompts producing 465,823 citations in December 2025. YouTube ranked first in AI citations while ranking only 11th in organic results for the same queries. Two limits matter: German-language queries, and health topics only.
A caution worth carrying. Different vendors report materially different YouTube citation shares using different methods, and we found no way to reconcile them. Cite one, name it, and never average them together.
Video Consumption and Ad Spend
US digital video ad revenue reached $78.0 billion in 2025. The market data is the best-measured part of this topic, because auditors and panel measurement produce it rather than marketing surveys.
44. US digital video ad revenue reached $78.0 billion in 2025, up 25.4% year over year. The IAB and PwC Internet Advertising Revenue Report, published April 2026, surveys ad sellers and has PwC verify actual booked revenue. It is the strongest measurement in this article.
45. US video ad revenue over five years: $39.5B, $47.1B, $52.1B, $62.1B, then $78.0B. The IAB and PwC Internet Advertising Revenue Report tracks that series across 2021 through 2025.
US digital video ad revenue, 2021 to 2025. Source: IAB and PwC Internet Advertising Revenue Report, April 2026.
46. Total US internet ad revenue hit $294.6 billion in 2025, with search at $114.2 billion. In the same IAB and PwC report, search grew 11.0%, decelerating from 15.9% the year before. Video grew more than twice as fast as search.
47. US digital video ad spend is forecast to pass $80 billion in 2026 and to exceed 60% of all TV and video ad spend for the first time, at 61% digital against 39% linear. IAB published that ad spend outlook in May 2026. It is a forecast built on a buyer survey, so it belongs in framing rather than in hard claims.
48. Video accounted for around 75% of all mobile data traffic at the end of 2025. The Ericsson Mobility Report from June 2026 also put total global mobile data traffic at 210 exabytes per month, up 22% year over year. This covers mobile networks only and should never be generalised into “75% of internet traffic is video.”
49. Video is the largest single application category on fixed networks at 39% of downstream traffic. Sandvine’s Global Internet Phenomena Report measured this in 2024. Earlier editions reported around 65% using a broader definition, so the two figures are not a trend line.
50. 91.1% of adult internet users watched online video in the past seven days, the top-ranked media activity in DataReportal’s Digital Global Overview Report, published October 2025 with data from GWI. This is self-reported survey data, so read it as directional.
How B2B Teams Measure Video
Search impact is rarely what B2B teams actually measure video on, which is a large part of why clean video SEO data is so scarce.
In our 2026 State of Creative in Tech report, a survey of more than 230 B2B tech marketing leaders, respondents said they judge creative success primarily on engagement and views, at 52.1%. Revenue influenced followed at 36.3%, then conversion rate or cost per lead at 35.0%. Measurement and attribution ranked as the second biggest constraint on their creative work at 14.1%, behind only budget at 27.4%.
That matches what practitioners say directly. Across the SEO threads we reviewed on whether video improves rankings, not one contained a before-and-after export, a controlled test, or a screenshot. One commenter put the problem cleanly: teams change five things at once and credit whichever one they liked most.
Two more findings from the same survey are worth reading next to the 19-to-1 gap between video and video markup. Only about one in three teams has a fully reusable asset system, and just 6.8% say they lack enough proof material, meaning roughly 93% have proof they cannot activate. Both point at the same thing this article keeps running into. Video gets made. It does not get made findable.
These are industry findings about how B2B tech marketing teams operate. They describe the market, not our own results.
For statistics on how video performs as a marketing channel rather than a search channel, we cover that separately in our video marketing statistics roundup.
FAQs
Is video content good for SEO?
Video helps when the query already returns video, and it helps least when it is added to a page as decoration. Google requires the watch page to already be performing well in Search before it considers the video for indexing, so video amplifies existing page authority rather than substituting for it. The practical test practitioners use: search your target keyword, and if videos appear near the top, video is worth producing for that query.
Does adding a video to a blog post improve its rankings?
There is no controlled evidence that it does. Google’s April and December 2023 changes specifically excluded pages where video is not the main content, which is exactly what a blog post with a supporting video looks like. Advanced Web Ranking’s data shows a video block on the results page lifting the top result about 2.3 percentage points, but that measures the presence of a video feature on the page, not the effect of your embed.
How do I get my video into Google video results?
Use VideoObject markup with its three required properties, name, thumbnailUrl, and uploadDate, and host the video on a page whose main purpose is showing that single video. Google supports structured data, video sitemaps, and Open Graph equally. Google also states that markup does not guarantee any specific video feature. Only about 0.34% of pages currently carry VideoObject markup, so the eligibility bar is lower than the competition suggests.
Do captions improve video SEO?
Neither Google nor YouTube documents captions as a ranking or indexing input. We checked both platforms’ documentation directly. Captions have a solid evidence base for engagement and accessibility, with 80% of viewers in Verizon Media and Publicis Media’s research saying they are more likely to finish a captioned video, and they are a WCAG Level A requirement. Treat them as an engagement and accessibility measure with an unproven search effect.
Does YouTube get cited in AI search results?
Heavily, and disproportionately. Otterly.ai found YouTube takes 31.8% of the social and video citations in AI answers, and SE Ranking found YouTube was the single most-cited source in health-related AI Overviews at 4.43%, ranking first in AI citations while ranking only 11th organically. Long-form video takes 94% of those citations, and view counts show almost no relationship to citation frequency.
The Bottom Line
The most useful finding in this data is not a growth number. Video presence in Google results is currently falling, not rising, and the statistics most often used to argue the opposite were retired years ago. What has not changed is the gap between publishing video and making it findable: video outnumbers video markup on the web by roughly 19 to 1, and Google’s own documentation puts video indexing downstream of a page that already ranks.
That reframes the work. Video SEO means making the video you already have eligible for indexing, on pages built for it, targeting queries that actually return video.
At Vidico, we produce explainer and product video for B2B SaaS and tech teams that need search and YouTube to do real work. Our campaigns with Digital Ocean drove more than 3.7 million YouTube views at a lower cost per acquisition, and Hiver’s YouTube campaign generated over 20 million impressions and 80-plus leads at a 5.5% engagement rate.
If you want to see how your video presence compares against the brands you compete with, our Creative Intelligence Report covers 12 areas of analysis and comes back within 48 hours.
Sources
- MozCast Feature Graph
- seoClarity: Video Carousel SERP Feature Impact
- Perficient: YouTube Continues to Dominate Video Results in Google
- Similarweb: Top Websites
- Pew Research Center: Many Turn to YouTube for How-To Lessons
- Pew Research Center: Americans’ Social Media Use
- Pew Research Center: Facts About Americans and YouTube
- YouTube Press
- YouTube: The Future of YouTube
- Alphabet Quarterly Earnings Release
- Nielsen: The Gauge
- HTTP Archive Web Almanac: Structured Data
- HTTP Archive Web Almanac: SEO
- Google Search Central: Video SEO Best Practices
- Google Search Central: Video Structured Data
- Google Search Central: Video Is the Main Content
- Advanced Web Ranking: Google Organic CTR
- Sistrix: Why Almost Everything You Knew About Google CTR Is No Longer Valid
- Streaming Media: Verizon Media and Publicis Media Sound Off Study
- CivicScience: Young Americans and Subtitles
- Semrush: AI Overviews Study
- Otterly.ai: YouTube AI Citation Study
- SE Ranking: Health AI Overviews, YouTube vs Medical Sites
- IAB and PwC: Internet Advertising Revenue Report
- IAB: US Digital Video Ad Spend to Surpass $80B
- Ericsson Mobility Report: Mobile Traffic Update
- DataReportal: Digital Global Overview Report
- Forrester: The Easiest Way to a First Page Ranking on Google
