Key Takeaways
No documented, verifiable case exists of an AI-narrated slideshow, faceless video, or AI avatar video earning a citation in Google AI Overviews, ChatGPT, Perplexity, or Gemini. PushLeads tracked 493 videos and scanned 17,000 citation links and found zero citations. YouTube captures roughly 21% of AI Overview citations, but the cited videos are human-presenter tutorials and hands-on demonstrations, not narrated slideshows. Transcript parsability gives slideshows structural access to retrieval pipelines, but structural access is not citation selection.
- YouTube accounts for approximately 21% of Google AI Overview citations, but cited content skews heavily toward how-to and demonstration formats, not AI-narrated slideshows.
- PushLeads tracked 493 videos and scanned 17,000 citation links across its own channel and client fleet and found zero AI citations attributable to slideshow-format videos.
- The only vendor claiming AI avatar videos earn citations at the same rate as human presenters is XLR8, which sells an AI video tool and provided no raw data or disclosed methodology.
- Otterly AI analyzed 100 million citation instances and found that 35% of cited channels had fewer than 10,000 subscribers, directly contradicting the 1,000-subscriber floor claim.
- Transcript-driven retrieval means a well-captioned slideshow is structurally parsable by ChatGPT, Perplexity, and Claude, but parsability does not determine citation selection.
- Google’s scaled content abuse policy, enforced from March 2026, targets templated, generic, high-volume output regardless of whether a human or AI produced it.
- AI referral traffic currently represents roughly 0.12% of total site traffic, meaning citation volume does not yet translate into meaningful business value for most contractors.
Do AI-narrated videos get cited by AI search engines?
No documented, verifiable case exists of an AI-narrated slideshow, a faceless video, or an AI avatar video being cited in Google AI Overviews, ChatGPT, Perplexity, or Gemini as of mid-2026. PushLeads scanned 17,000 citation links across 493 tracked videos and found zero citations attributable to that format. YouTube as a domain earns roughly 21% of Google AI Overview citations according to Ahrefs data covering three million US searches, but the videos selected within that share are overwhelmingly human-presenter tutorials, product demonstrations, and hands-on repair or installation content, not narrated slideshows over stock footage.
The distinction matters because vendor content marketing has blurred the line between what is theoretically possible and what the data actually shows. A slideshow video with a clean, accurate transcript is structurally parsable by text-crawling retrieval systems used by ChatGPT, Perplexity, and Claude. That parsability is a real, genuine advantage. However, parsability is an entry condition, not a selection criterion. The reason a specific video earns a citation is that it visibly and demonstrably answers the query. A narrated slideshow over stock footage cannot replicate a hands-on demonstration, and every large-scale study of cited videos reflects that gap.
The absence of evidence here is meaningful, not merely inconclusive. No named channel, no verifiable screenshot, and no independent replication supports the claim that AI-narrated video earns citations at scale. Understanding the difference between ranking on Google and getting cited by AI is the first step toward building a video strategy that actually moves the needle.
Want more customers from Google & AI search?
Get a free SEO audit of your site — see exactly what to fix first.
YouTube Dominates AI Citations, But Not Through Slideshow Videos
YouTube captures approximately 21.1% of citations in Google AI Overviews, making it the second-largest cited domain after Reddit, according to Ahrefs Brand Radar analysis of three million US queries in June 2026. BrightEdge reports a roughly 200x citation advantage for YouTube over Vimeo, its nearest video competitor. Peec AI’s analysis of 30 million sources across five engines also ranks YouTube second overall. These numbers are real and consistent across independent trackers.
However, a high-level domain citation figure tells you nothing about which types of videos within that domain are being selected. Every large-scale study that has looked at cited video format, rather than just cited domain, reaches the same conclusion. MP Digital’s Q1 2025 analysis of 10,000 Search Generative Experience overviews found that how-to citation rates were up 651% and visual demonstration citations were up 592% year over year. PushLeads reviewed its own top 50 cited channels across its client fleet and found the same pattern: step-by-step instructional videos, hands-on product comparisons, and visual repair or installation content.
What formats actually get cited within YouTube?
The formats that earn YouTube citations fall into three clear categories. First, step-by-step how-to tutorials where each step is visibly performed on camera. Second, product demonstrations that show hands-on use, comparison, or teardown. Third, visual repair and installation content, particularly in trades like plumbing, HVAC, roofing, and electrical work, where showing the physical task directly answers the searcher’s question. These formats share a common thread: the video itself is the evidence that the answer is correct.
Why does citation share not equal citation type?
YouTube’s 21% citation share is measured at the domain level. That figure aggregates every cited YouTube video regardless of format. When researchers or practitioners use that number to argue that any YouTube video has a 21% chance of citation, they are making a category error. The relevant question is not what percentage of AI Overview citations point to YouTube, but what percentage of AI Overview citations point to videos that look like the one being produced. The answer for AI-narrated slideshows, based on all available data, is zero verifiable cases.
Source Quality: Separating Primary Research from Vendor Marketing
The citation research landscape divides cleanly into two tiers, and treating both as equally credible produces bad strategy decisions. Tier 1 sources have disclosed methodologies, large sample sizes, and no direct financial stake in a particular content format winning. Tier 2 sources are vendor content marketing, often citing their own products or category, with small samples and no raw data.
Tier 1 studies include Otterly AI’s March 2026 analysis of 100 million citation instances over 30 days, Ahrefs Brand Radar covering 463 keywords and three million US queries, SE Ranking’s December 2025 study of 50,000 searches, Peec AI’s analysis of 30 million sources across five engines, and Digital Applied’s review of 1,000 AI Overviews in April 2026. These studies have partial to full methodology disclosure and no vested interest in a specific video format performing well.
What is the XLR8 problem and why does it matter?
The only source claiming that AI avatar videos earn citations at the same rate as human-presenter videos is XLR8, a vendor that sells an AI video generation tool. In a July 2026 guide, XLR8 made three specific claims: every cited video in their sample came from a channel with fewer than 1,000 subscribers, AI avatar videos and human-presenter videos perform equally for citation, and their own near-zero-subscriber channel was outperforming established channels. The sample size was described as dozens of videos. No raw data, no methodology, and no third-party replication accompanied these claims. XLR8 sells the product that would benefit most if those claims were true. That is a significant conflict of interest, and the claims directly contradict Otterly AI’s far larger study on every subscriber-related finding.
Should the 1,000-subscriber floor be treated as a real threshold?
No. The 1,000-subscriber floor is almost certainly an artifact of XLR8’s small, self-interested sample rather than a real eligibility threshold. Otterly AI’s 100 million citation instance study found that 35% of cited channels had fewer than 10,000 subscribers, and 40% of cited videos had fewer than 1,000 views at the time of analysis. The median cited channel had fewer than 41 videos. Otterly AI’s explicit finding is that the best answer wins, not the biggest channel. Small channels with highly specific, well-structured content that directly answers a query are cited regularly. The relevant variable is answer quality and structural fit, not subscriber count.
How AI Retrieval Pipelines Actually Work, and Where Slideshows Fit
Understanding the technical retrieval pipeline clarifies exactly which variables determine citation eligibility and where the slideshow format has a genuine structural advantage versus where it falls short. The pipeline has three main stages, and the slideshow format performs differently at each one.
Text crawl is the first stage. ChatGPT, Perplexity, and Claude retrieve content by fetching the transcript, description, and chapter markers. These systems cannot watch video frames at all. A slideshow video is not structurally penalized at this stage because the retrieval system is operating purely on text. John Mueller has confirmed that Google uses transcripts and captions to understand video content, and that confirmation represents the strongest genuine structural point in favor of slideshow videos.
What advantage does a slideshow video actually have over other formats?
The one real structural advantage of a slideshow format is transcript-driven retrieval. For ChatGPT, Perplexity, and Claude, citation eligibility runs entirely through the text layer. A well-transcribed, chapter-organized slideshow is in principle as parsable as a human-presenter video. This is not a minor point. It means the format is not excluded from consideration by the retrieval systems that crawl text. The problem is that parsability is an entry condition, not a selection criterion. Getting into the eligibility queue and being selected from that queue are two separate events, and the slideshow format has no documented wins at the selection stage.
What happens at the frame sampling and timestamp stages?
Google’s Gemini stack can sample visual frames at approximately one frame per second, which means it can assess whether a video visually answers the query. Slideshows over stock footage do not perform well at this stage because the frames do not contain visual demonstrations of the task being answered. Timestamp and chapter structure is the third stage. Seventy-three percent of timestamp-level citations occur in Google AI Overviews and 27% in Google AI Mode. Clean chapter markers with accurate timestamps improve structural eligibility across all retrieval systems, and this is one area where slideshow producers can and should invest effort regardless of format decisions.
The Enforcement Risk: Scaled Content and Inauthentic Content Policies
Google’s scaled content abuse policy, published in March 2024 and enforced from March 2026, explicitly states that volume without value is a red line regardless of whether content is produced through automation, human effort, or some combination. The policy is method-neutral. An automated single-voice slideshow produced at scale lands directly within scope of this policy not because it is AI-generated but because it is templated and generic.
The pattern is familiar. Producing keyword-targeted slideshow videos with slight geographic or topical variation, for example a plumbing video for Asheville followed by an identical one for Charlotte, is structurally identical to the scaled thin-content strategies that Google penalized in 2011 and again in the Helpful Content Updates of 2022 through 2024. The enforcement mechanism has changed but the underlying principle has not: Google treats mass-produced, interchangeable content as an attempt to game the index rather than serve searchers.
What is YouTube’s inauthentic content policy and does it affect AI-voice slideshows?
YouTube’s monetization update of July 2025 and the enforcement wave reported in January 2026 targeted mass-produced or repetitive content. Whether a slideshow with an AI voice qualifies as repetitive content is a gray area, but the risk is real. YouTube’s altered or synthetic label is a transparency signal rather than an automatic penalty flag. A single 11 Labs voiceover on a slideshow generally does not trigger the disclosure requirement. The disclosure policy primarily targets realistic synthetic depictions of real people, so an AI voice without a face does not automatically require a disclosure label. However, producing hundreds of near-identical videos at scale increases the probability of enforcement action under the repetitive content clause.
How does the inauthentic content risk apply to home service contractors specifically?
Home service contractors who are tempted to produce dozens of geo-targeted slideshow videos, one for each service area, face the highest enforcement risk because the content is structurally identical across videos. A video titled “HVAC Repair in Asheville NC” and one titled “HVAC Repair in Hendersonville NC” produced from the same template are precisely the type of content both Google and YouTube have flagged in recent enforcement waves. The channel-level demonetization or removal risk may not be worth the speculative citation benefit, particularly for a contractor who depends on their YouTube presence for both search visibility and customer trust.
PushLeads First-Party Data: 493 Videos, 17,000 Citation Links, Zero Citations
PushLeads tracked 493 videos and scanned 17,000 citation links across its own channel and client fleet and found zero citations attributable to the slideshow video format as of mid-2026. The channel itself had 677 subscribers, approximately 161 public videos, and around 82,000 total views at the time of analysis. These figures put the channel well within the range that Otterly AI identifies as citation-eligible based on subscriber count and view count alone.
The zero-citation result is significant precisely because it cannot be explained by channel size. Under Otterly AI’s model, channels with fewer than 1,000 subscribers and videos with fewer than 1,000 views are cited regularly when their content matches the query with high specificity. The PushLeads channel fell short not because of its size but because most of the 493 tracked videos were broad, generic topic videos covering questions like what is SEO and how does SEO work, rather than hyper-specific, single-question-focused content.
What does the first-party data show about YouTube citation share overall?
Across PushLeads’ client fleet of approximately 45 local service businesses, YouTube showed up in roughly 50% of AI citation appearances, which is higher than the 21% domain share figure from Ahrefs but consistent with BrightEdge’s finding of around 60% and SE Ranking’s finding of around 82% for video content in AI Overviews. ChatGPT drove approximately 4% of YouTube citations, consistent with Otterly AI’s finding that ChatGPT rarely uses YouTube as a citation source. The practical implication is that video content primarily serves Google AI Overview visibility rather than ChatGPT or Perplexity visibility.
Why does zero citations across 493 videos matter more than the overall YouTube citation share?
The overall YouTube citation share of 21% is a real and meaningful figure for strategy planning, but it does not predict what will happen to any specific video format. The PushLeads first-party data provides a format-specific data point that the large third-party studies cannot, because those studies aggregate all YouTube citations without distinguishing by video format. Zero citations across 493 videos, in a channel size range that the largest available study identifies as citation-eligible, is the most direct available evidence that slideshow format alone is not sufficient for citation selection. For a deeper look at what structures actually drive citations, the contractor’s guide to structuring FAQ pages for AI citations covers the underlying content architecture principles that apply across formats.
The Traffic Reality: Do AI Citations Translate to Business Value Today?
AI referral traffic represents approximately 0.12% of total site traffic across Ahrefs’ study of 35,000 websites, which is roughly equivalent to the traffic share driven by Reddit. Steve Huffman, Reddit’s CEO, has stated publicly that chatbots are not a major traffic driver today. Lily Ray, writing in Search Engine Land, noted that AI Overview citations consistently underperformed even the blue links near the bottom of the search results page for click-through rate during the first half of 2026.
That does not mean AI citation strategy is irrelevant. AI usage has grown from approximately 5% of search interactions to roughly 50% over the course of 2025 and into 2026, and that trajectory points toward a future where citation visibility will matter much more than it does today. The strategic question for a home service contractor in 2026 is whether to invest heavily now in a format that earns zero documented citations, in the hope that citation behavior will change and that format will benefit, or to invest in formats that already earn citations and build from a documented foundation.
Does citation share differ by AI engine in ways that affect contractor strategy?
Yes, significantly. YouTube earns the majority of its citation share through Google AI Overviews, not through ChatGPT or Perplexity. For contractors whose customers primarily use Google Search, which accounts for roughly 90 to 95% of all search activity, video content strategy should be optimized for Google AI Overview citation patterns. Those patterns strongly favor demonstration content, clean transcripts, and timestamp structure. For contractors who also want to appear in ChatGPT responses, video content is largely irrelevant. ChatGPT drives only about 4% of YouTube citations, and the primary citation sources for ChatGPT are text-based, including Reddit threads, structured FAQ pages, and authoritative editorial content. Getting cited by ChatGPT in 2026 requires a fundamentally different content approach than getting cited by Google AI Overviews.
What is the real business case for video content if not AI citations?
Video content delivers genuine business value through channels other than AI citation. On-page engagement increases when a relevant video is embedded alongside text content. YouTube search and conventional Google video search results provide traffic independent of AI Overviews. Content repurposing across formats, turning a video transcript into a blog post, a FAQ section, or social clips, multiplies the reach of a single production effort. These are not AI citation benefits, but they are real and measurable. The honest framing for a contractor is that video content is worth producing for these reasons, and that AI citation visibility may become an additional benefit as retrieval systems evolve, but that it is not a reliable primary goal today.
The Two-Stage Recommendation for Contractors Considering Video Strategy
Stage one is to stop counting on AI-narrated slideshow videos as AI citation assets. The evidence for that format earning citations is not merely weak. It is absent. There is a meaningful difference between weak evidence, which would be a few citations from small samples, and no evidence, which is the current state. No named channel, no verifiable screenshot, and no independently replicated study supports the claim that AI-narrated slideshows earn AI citations. PushLeads’ own 493-video, 17,000-citation-link audit supports the same conclusion. Slideshow videos may still have value for on-page engagement, YouTube search visibility, and content repurposing, but those are not citation benefits.
Stage two is to run one clean, parsable test over 60 to 90 days. Because transcript-driven retrieval is real, it is worth testing whether addressing one specific question per video, with a manually verified transcript, clean chapter timestamps every 10 seconds, and a structured description of at least 300 words with entities and relevant hashtags, produces any citation appearances. Produce 10 to 15 videos meeting all criteria and track citation appearance over six weeks. The test should use YouTube’s own Analytics alongside a citation monitoring tool to separate YouTube search visibility from AI citation visibility. Understanding how local service businesses get cited by Google AI Overviews provides the structural framework that makes this kind of test interpretable.
Does switching to a human presenter solve the underlying problem?
Partially, but not automatically. The data does not say AI video never works and human video always does. It says that videos that demonstrate value and match retrieval structure are driving citations. A human presenter helps when the presenter enables real physical demonstration, establishes experience signals through visible expertise, and combines a real face with a hands-on visual demonstration and clean transcript. That combination is what most cited YouTube videos look like. A talking-head presenter without demonstration content is not meaningfully better than a well-structured slideshow for citation purposes. The real differentiator is demonstration content, transcript accuracy, and topical specificity, not the presence of a human face alone.
What content structure variables matter most across all video formats?
Three variables matter across all formats. First, topical specificity: one question per video, answered directly in the first 60 seconds. Second, transcript accuracy: a manually verified transcript is significantly more reliable than YouTube’s auto-generated captions, which contain errors that degrade retrieval quality. Third, chapter structure with timestamps at roughly 10-second intervals for the key steps or claims in the video, so that retrieval systems can surface the specific segment that answers a query rather than the video as a whole. These variables apply equally to human-presenter videos and slideshow videos, and they are the variables that a contractor can control immediately without a major format change.
Frequently Asked Questions
Is there any documented case of an AI-narrated slideshow video earning an AI citation?
No. As of mid-2026, no named channel, no verifiable screenshot, and no independently replicated study has produced a confirmed case of an AI-narrated slideshow, faceless video, or AI avatar video being cited in Google AI Overviews, ChatGPT, Perplexity, or Gemini. The only source claiming otherwise is XLR8, a vendor that sells AI video tools, and that claim has not been replicated by any independent researcher using disclosed methodology.
Does a YouTube channel need 1,000 subscribers to get cited by AI search?
No. The 1,000-subscriber floor is a claim made exclusively by XLR8 based on a sample of dozens of videos with no disclosed methodology. Otterly AI’s analysis of 100 million citation instances found that 35% of cited channels had fewer than 10,000 subscribers and 40% of cited videos had fewer than 1,000 views. The median cited channel had fewer than 41 videos. Otterly AI’s conclusion is that the best answer wins, not the biggest channel. Subscriber count is not a documented eligibility threshold for AI citation.
Why does YouTube capture 21% of AI Overview citations if slideshow videos earn zero?
YouTube’s 21% citation share is a domain-level figure that aggregates all cited YouTube videos regardless of format. The studies producing that figure, primarily from Ahrefs, BrightEdge, and Otterly AI, do not break down citations by video format. Every study that has examined cited video format finds that cited content is overwhelmingly how-to tutorials, hands-on demonstrations, and visual repair or installation content. The 21% figure is real. Applying it to slideshow-format videos specifically is a category error.
Does a clean transcript make a slideshow video eligible for AI citation?
A clean, accurately transcribed slideshow video is structurally parsable by the text-crawling retrieval systems used by ChatGPT, Perplexity, and Claude. That parsability is the one genuine structural advantage of the slideshow format, and John Mueller has confirmed that Google uses transcripts and captions to understand video content. However, parsability determines retrieval eligibility, not citation selection. A video that is readable by an AI retrieval system can enter the candidate pool but still not be selected, because selection depends on whether the video demonstrably and visually answers the query.
What does Google’s scaled content abuse policy mean for contractors producing slideshow videos?
Google’s scaled content abuse policy, enforced from March 2026, explicitly targets volume without value regardless of whether content is produced by humans, AI, or a combination. Producing large numbers of geo-targeted or keyword-targeted slideshow videos from the same template is precisely the pattern the policy describes. The policy is method-neutral, meaning the fact that a human recorded the narration does not protect a templated, interchangeable video from enforcement. Contractors should treat mass-produced slideshow production as carrying real enforcement risk at both the Google Search and YouTube levels.
How much traffic do AI citations actually drive to websites today?
Very little. Ahrefs’ study of 35,000 websites found that combined referral traffic from ChatGPT, Perplexity, and Gemini represents approximately 0.12% of total site traffic, roughly equivalent to Reddit’s share. Lily Ray’s research published in Search Engine Land found that AI Overview citations consistently produce fewer clicks than blue links near the bottom of the search results page. AI citation visibility is a forward-looking investment given the rapid growth in AI search usage, but it does not yet translate into measurable traffic volume for most small business websites.
What is the difference between weak evidence and absent evidence for AI slideshow citation?
Weak evidence would mean a small number of confirmed, verifiable cases exist but the sample is too small to generalize. Absent evidence means no confirmed, verifiable case has been produced at all, no named channel, no screenshot, no independently replicated study. For AI-narrated slideshow citation, the evidence is absent, not merely weak. That distinction matters for strategy because absent evidence suggests the phenomenon either does not occur or occurs so rarely that it cannot be detected even in studies covering 100 million citation instances.
Should a home service contractor invest in video content if AI citations are uncertain?
Yes, but with realistic expectations about which benefits are documented. Video content delivers measurable value through on-page engagement, YouTube search traffic, and conventional Google video results. Those benefits are real and independent of AI citation. AI citation through YouTube is a real possibility for demonstration-format content based on Otterly AI’s findings, but it requires topically specific, single-question-per-video content with clean transcripts and timestamp structure. Slideshow-only format carries both enforcement risk and zero documented citation benefit, making human-presenter demonstration content the more defensible investment for most contractors.
Start With What the Data Actually Supports
The PushLeads audit of 493 videos and 17,000 scanned citation links produced zero AI citations for slideshow-format video content. The largest independent studies covering 100 million citation instances agree that small channels with highly specific, well-structured demonstration content earn citations regularly. The format variables that matter are topical specificity, manual transcript accuracy, clean timestamp chapters, and visual demonstration of the answer. Those variables are achievable for any home service contractor in any trade.
Before investing further in video production, check whether your existing content meets the structural criteria that retrieval systems actually use. One video that directly demonstrates how to diagnose a leaking pipe, reset a tripped breaker, or identify a failing HVAC capacitor, filmed by a real technician with a clean transcript and accurate chapter markers, is a stronger citation candidate than 50 narrated slideshows on the same topic. The evidence points clearly in that direction. A good starting point is reviewing how local service businesses get cited by Google AI Overviews and applying those structural principles to your next video production. If you want a hands-on review of your current content structure and citation eligibility, contact PushLeads directly to schedule an audit.
Key moments
- 0:00 Do AI-narrated videos get cited?
- 0:37 The short answer, and why I pulled the product
- 1:25 The verdict in three bullets
- 3:45 YouTube owns 21% of AI Overview citations
- 4:36 What actually gets cited: demonstration, not narration
- 5:53 Tier 1 vs Tier 2 evidence: who disclosed a methodology
- 7:13 The single-vendor claim with zero replication
- 8:34 The 1,000-subscriber floor is almost certainly false
- 10:01 How AI engines actually retrieve video
- 11:19 The slideshow's one genuine advantage: transcript-driven retrieval
- 12:52 Enforcement risk: scaled content abuse is method-neutral
- 14:48 AI disclosure labels are a red herring
- 15:24 My first-party data: 493 videos, zero citations
- 17:25 Does ChatGPT cite YouTube?
- 18:18 Traffic reality check: citation is not revenue
- 20:50 Public vs unlisted vs private citation eligibility
- 21:22 Recommendations: stage 1, stop counting on it
- 22:42 Stage 2: one falsifiable 60-90 day test
- 23:32 The human presenter question
- 25:08 Key caveats: absence of evidence is not evidence of absence
Full transcript
The complete spoken content of the video above, in text, with timestamps that jump to that moment. This transcript is the exact audio of the published cut, transcribed from the render rather than lifted from auto-captions.
0:00 Hey, I'm Jeremy with push leads helping small businesses get more leads without struggle and frustration and today We're looking at a really interesting topic that I've been personally researching Do AI generated videos actually get cited by answer engines by chatbots by AI's and this all goes back to Sort of an idea that I came up with what and that's all goes back to an idea that I came up with which is basically that You know if we could make a PowerPoint slide show with an actual AI voice like a real, you know human voice But an AI version of it is that enough to get AI citations and the short answer is no It's not enough.
0:39 There's no evidence for that, but let's dig in and Throughout this video you'll see references to push vid Which is sort of the version one of this product that created and that and that product basically is essentially a slideshow video product So no, there's so there's definitely a good reason to have a slideshow video product The question is is putting it on YouTube.
1:04 Will it give you citations? So we started doing this and then later on retracted these because it looks like there could be a possibility that Google could consider it abuse So I'm not saying that you shouldn't do slideshow videos, but I was I'm but I am saying you need to be super careful So let's look at what the data says about it and jump into it So the verdict in three bullets is that there's zero documented cases There's no first -hand verifiable account of an AI generated narrated slideshow a face this video or an AI avatar video being chided Being cited in Google's AI overview perplexity chat GDP Gemini, and
1:44 so it doesn't actually exist YouTube is of course cited But in our videos which we just started doing about a month ago at the beginning of August or so 2026 we're not being cited, but obviously we just start there the couple videos were cited in but not very many Again sort of my picture is that I've been in and out of YouTube for a long time really jump fully into it because You know essentially there's just so much traffic on the shorts and there's so many videos that do exactly what I do There's so many companies that do exactly what I do So what's the point and then I think when I looked at my own fleet of 45 clients that we're doing SEO for
2:25 mostly service -based businesses. I started realizing that there's one competitor that kept showing up over and over again That's competing for my clients visibility and that was YouTube it showed up 50 % of the time So YouTube has a domain captures about 21 % of Google's AI overview citations But the overwhelming videos are the human presenter the hands -on demonstration content or maybe videos like this where it's an actual video of me talking Walking through a slideshow.
2:56 This is real me not AI version of me. This is not Hagen. Right? So anyway, it's human first and Google and YouTube seems to prefer human first content and then our own Data has you know zero citations as we talked about it So small channels can be cited according to Otterly AI's model where small channels can be cited So it points to that this will work doing videos will work, but the reality is it's gonna take some time.
3:27 So So let's get in so chapter one YouTube dominates AI citations, but you but not so So chapter one YouTube dominates AI citations, but not YouTube slideshow videos So independent trackers converge on a striking number YouTube captures about 21 .1 % of sites in Google's AI overview and that comes from H .ress As of June of 2026 looking at three million US searches and bright edge reports a 200x advantage over the nearest video competitor Which is Vimeo only having about 1 % 0 .01 % peak AI 30 million Source studies ranks
4:07 YouTube as number two behind Reddit So this so this does not mean that a high So what does this not mean a high -level domain citation tells you nothing about the types of videos that are being selected? So every study is really looking at the same pattern the how to the tutorial the visual demand and the product comparison format so So let's keep going So what actually gets cited it's demonstration not narration.
4:37 So every third party describing a character So every third party study looking at cited videos look at that came to the same conclusion So every third party study of character cited videos came to the same conclusion MP Digital's Q1 2025 analysis of 10 ,000 sge overviews reported that the how to Citations were up 651 % and the visual demonstration videos were up 592 % so My own top 50 cited channels also meets that pattern.
5:13 So it's how to tutorials which are step -by -step instructions It's product demonstrations, which is hands -on Comparisons and then of course visual repair and bit anything with installation construction and repair So an AI narrated slideshow over stock footage cannot replicate this experience We're big the transcript could be readable by an AI But there but it just means that the video itself is not gonna get citations Chapter 2 source quality who founded what chapter 2 source quality who funded what?
5:47 So we need to look at two studies here And this actually is a foundation of two tiers tier one is the primary research and disclosed methodology Which is utterly AI looked at a hundred million citation instances over 30 days six in March and this was in March of 2026 Atrath's brand radar looked at 463 keywords, which is four million URLs three million US queries and search E ranking German house study looked at 50 ,000 searches of December 2025 peak AI also looked at 30 million sources five engines and digital applied looked at a Thousand aios in April 2026 And so
6:27 this is a ton of data here and they all have partial to full methodology and then tier two is vendor content marketing Well our eight AI samples dozens No raw data just they sell their own video AI tool and bright edge Which sells AI catalysts and the and and their founder claims Google watches video Which is unconfirmed by Gush thought and then everything PR slash 5w Which is a synthesis of other studies Veed vid IQ and P digital ratty rant, etc So a ton of information that goes into it So let's look So let's
7:08 so let's look at the XL R8 problem, which is one vendor and there's been zero replication of it So the only source claiming that AI avatar videos get cited at the same rate as the human presenters is XL R8 That's a mouthful. This is a vendor who sells an AI video generation tool and And in their July 2026 guide that it made three specific claims It's said that it said that every sighted video came from a channel with less than a thousand subscribers in this AI avatar and human presenters work Quote both work equally and it has and it's own near zero subscriber channel
7:48 Outsights established channels. So this got So this guide sample is dozens of videos with no raw data and no disclosed methodology And of course they sell this product, right? So if they're selling an AI video product, of course, they're gonna put out stuff saying that AI video works It really kind of this whole thing really seems like a gray area honestly and this and then and it also Contradix the far larger and also directly contradicts the far larger otterly AI study on every subscriber finding So the verdict here is treat every XL R8 claim as an unverified single vendor marketing
8:28 It's a sole source of any kind of AI event AI avatars get cited and of course the thousand subscriber floors floor claims neither has been replicated by in by an independent self and that should send off alarm bells and Then we're gonna get into the thousands of scrap and then we're gonna get into the thousand subscriber floor Which is almost certainly false and otterly AI looked at a hundred million citations, which is the large study on YouTube AI citations and under 35 % and 35 % were under 10 ,000 subscribers So that was the actual number is 10 ,000 subscribers and then and then 40 %
9:08 were under a thousand views of the AI sighted videos Fewer had less than a thousand views at the time of the analysis and then the median channel had over had fewer than 41 videos So otterly's AI so otterly's so otterly AI is Thesis is the best answer wins not the biggest answer. So otterly's AI Expleased it so otterly AI's Explicit thesis is that the best answer wins not the biggest channel and the thousand subscriber floor is an artifact of XL of XL AR's tiny self -interested sample not a real
9:49 threshold So small channels do get so small channels do get cited regularly when their content matches the query So how exactly do AI engines ants? So chapter 3 retrieval mechanics how AI engines actually retrieve videos So it's important to understand the technical pipeline because this clarifies exactly what determines the citation Eligibility and of course where slideshows fit in that mix so the first thing we have is a text crawl where chat GDBV So the first thing we have is a text crawl where chat GBT perplexity and clod are flexing Where chat GPT perplexity
10:29 and clod are fetching the transcript description and chapters They cannot watch the frames only a slide and so therefore and so therefore a slide show is not structurally penalized Fraying sampling Google's Gemini stack can sample visual frames at about one frames per second But tellingly the but tellingly Gemini sites YouTube the lead of any engine which is point two percent Which doesn't make any sense to me because it's Google's product and then we third and then third we have timestamp Parsivit which is 73 % of all time stamp citations occur in Google's a I O 27 % in Google's AI mode
11:09 and clean chapters which is you know time stamps and of course Clean chapters with time stamps are structurally and we also have the slideshows number one genuine advantage And that's the ceiling what transcript driven retrieval means for slideshow for chat GPT perplexity and clod Citation runs entirely through the text label. So this is where a well transcribed chapter and slideshow is in principle as Parsible as a human presented video But John Mueller has and John Mueller has confirmed that Google uses transcripts and captions to understand video content So this alone is the strongest genuine point in
11:50 the favor of a slideshow only video And that's the only one with structural support which should give you red flags, right because Google and AI's prefer things a certain way and when you try to not and when you do things not their way That's considered gray hat or black hat. So again doing slideshow videos appears to be a gray hat sort of category So with the argument breaks down is Parsibility and select on the how to demonstrate his queries where YouTube actually earns its citations share I mean the reason a specific video is actually chosen it is because it visibly answers the aunt is because it visibly shows
12:30 the answer a Clean a clean transcript on a B role gets you the eligibility to queue But it does not win the citation selection over a channel that actually demonstrates the task So structural access doesn't mean citation success super important. Do you understand that? Chapter four the enforcement Ricity the real penalty vectors is scaled content and in authenticity So Google scaled content abuse policy that came out in March of in March of twenty So Google scaled con so Google scaled content abuse policy Which came out in March of twenty four and in force
13:10 March of twenty six and the explicitly the quote and it's Explicitly met the neutral no matter what content is produced through automation human efforts or some Combination the volume without value is a red line and Google views that is scaled content abuse So an automated single -voice slide should pop land at a scale is directly in scope of this Not not because it is considered AI made but it's because it's templated and generic and funny story I remember back in 2011 I had someone paying me to do slideshow videos that had keywords and they were targeting anyone searching for cable insulation or internet insulation
13:50 And so the videos were like able insulation Charlotte, North Carolina Cable Asheville North Carolina Internet Internet service provider and so we were cranking out some of these videos Thinking that I was thinking eventually Google is not going to like this and of course I was right so I don't that's so that was Like 15 17 years ago when I was kind of somewhat gray -hat. I don't do that at all now Obviously it makes you obviously you have to play the game and then and then YouTube's inauthentic content policy Which is July 2025 in January of 2026 is the enforcement wave and so YouTube's July 2025 Monetization update and reported an update
14:30 of January 2026 The this enforcement way targeted mass produced or repetitive content. So is slide shows that with an AI voice Sort of repetitive content. Well, it's a gray area, right? So rather guidelines have and then AI disclosure labels are a red herring So YouTube's alternate YouTube's altered or synthetic label is a transparency signal Not a penalty flag a single 11 labs voiceover on a slideshow generally does not trigger disclosure Disclosure requirements the policy targets realistic synthetic depictions of real people So if you're using hey gen
15:10 then you have to then you have to explain that it is AI generated But if it's the voice I you don't have to say it's AI generated and of course chapter five first party data So push leads which is our own age This is the most important data point in this entire analysis. So our current So our current channel stats are 677 subscribers around 161 public videos 82 ,000 or so 82 ,000 total views 100 and then we're tracking 493 tracked videos and of 17 ,000 scan citation links zero citations were found again I expect that
15:50 to start rising in 30 to 60 90 days I am starting to see YouTube search finding me and I am starting to see some visibility But it does take time when you start cranking out content and really I stopped doing content because I just didn't think that It was even valuable to Google. I had no idea that Google's AI was actually looking at all of these YouTube videos So why zero matters so much under autolies AI model channel size is not the barrier small channels with around 10k 5k or even less Than 1 ,000 subscribers are cited regularly, but there's but we have zero citations across 493 videos And that's because most of these videos have just been generic videos
16:31 like what is SEO? How does SEO work? They're not specifically answering specific questions However though some early data which is not in this PowerPoint is that when you make videos covering specific points Then you get visibility for that. So it really comes down to the targeting Versus and here's a look and here's a look at first -party data versus published benchmarks So we have 20 % of my so my YouTube citations share is around 20 % Versus H refs around 21 % and then everything PR 20 .9 % The YouTube actual citations and answers According to
17:11 my across my whole fleet is around 50 % which is higher than most domain share figures But that's consistent with bright -edged study of around 60 % and SE's ranking around 82 % And then how many and then how often is chat TPT setting YouTube? Essentially, it's identical to autolies AI finding that chat just drives around 4 % of YouTube citations So chat TPT barely uses YouTube So I always tell my clients if you want to show up for Google's a you need to have ongoing video content If you want to just show up for if you want to show up for chat TPT Then you want to show up on then you just then you want
17:51 to have reddit only Then you want to answer questions on reddit So the real question is is even making this content even worth it today, right? Because Google search really accounts for what 90 95 % of the entire equation the traditional search however, though AI usage has gone up from I think 5 to 50 % in the last year or so, so that's really gone up So chapter six the traffic reality check So chapter six the traffic reality check the citation share actually equal traffic Equal business value.
18:29 The answer is today. No, it doesn't So even we're having the citation case instructions disrupt The business case for chasing AI citations is actually somewhat Deserve somewhat So even where the citation case is strong the business case for chasing AI citations deserves some scrutiny here And the evidence is sobering we have AI traffic referral being tiny today, right?
18:57 In 27 28 and 30 it's going to be growing up short exponentially But a but href study across 35 ,000 sites found that chat to be perplexed and in gem night So i'm roughly the same amount of percentage of traffic as reddit to websites Which is like 0 .12 percent of total human traffic that and then reddit steve huffman says chatbots are not a major traffic driver today But one of the things we have to pay attention to is that Human behavior is changing more and more people are expecting automatic mode, right?
19:29 And so this shift in human behavior is causing all of us in marketing to really pay attention to AI citations even though today It doesn't have much value hard to know AI's AI AI citations are underperforming compared to the blue rink AI's AI citations are underperforming comparing to the blue rink AI citations are underperforming comparing to the blue AI citations are underperforming So AI citations underperform compared to blue So AI oh citations underperform comparing to the blue rinks So AI oh citations
20:10 underperform the blue links, of course And lily and lily ray at search engine land said in my research from the first half of this year AI overuse citations constantly unperformed even compared to additional blue rinks near the bottom of the SERP And then citation volume is not the real constraint So youtube's citation share is real the business question is whether chasing that shit Especially through slideshow content that typically earns zero citations Is that the is that the highest use of the production value for a company?
20:43 And we can take a quick look at real versus unlisted versus private citation eligibility Public videos are indexed by youtube search and AI engines and around two to and they show up in around two to three Weeks and they're right and they're reliably and fully sideable Unlisted and private videos are sideable only if the URL is linked on a public page or on a playlist And of course private videos cannot be indexed or cited at all by search engines and having zero but and they have zero eligibility So what should you do from here?
21:17 So chapter seven recommendations What should you do from here? So stage one stop counting on an AI slideshow video as an AEO asset the evidence is absent Not merely weak There is a meaningful difference between weak evidence and no evidence So in AI slideshow video earning citations for an AI slideshow earning site the evidence So for an AI slideshow earning its earning AI citations the evidence is absent There's no no channel.
21:47 There's no name channel no ID and no screenshot has been produced by any source Our own videos are also showing that it doesn't work of course But our most of our content until the last month or so Has not really been hyper focused content answering questions and of course that's what we're doing now Unless i'm doing this type of video Which is more about looking at the data So what should you keep on doing?
22:10 Video earns us keep if it if it delivers So what should you keep on doing? Video earns us keep if it delivers on on page engagement human audience utility conventional youtube and video search and content repurposing across format These are not citation benefits, but they are real But but they do show that slideshow videos do have real business value Even if it's not going to actually drive citations So after stage one you're going to jump into stage two Which is run one clean possible test over the six over the next 60 to 90 days Because retrieval is transferable Test
22:51 whether one variable actually works produce 10 to 15 videos meeting all criteria and we can interact the citation experience And attract and track the citation appearance over six weeks You're going to look at the same one quick sure that you're answering one question per video You're going to be doing the verify transcript make sure it's clean manually checked Not googles not goot not youtube's your own transcript, which is much more accurate You need to have the proper you need to have the proper chapters with the start the time stamps and then 10 seconds each And you need to have a structured description which at least 3 ,000 words with entities and hashtags when So
23:32 on the human presenter question what the actual data says The data does not say AI video never works and human video always does It says that it says that videos that demonstrate value and structure are driving sites and it when it A human presenter helps when it enables real physical demonstration and establishes Eat signals the combination of a real face plus a visual demonstration plus clean transcripts Is the configuration that gets mostly aligned with what actually gets cited a talking head alone is not enough So a human presenter without a demonstration is obviously not is obviously not better than a well -structured demonstration slide show So that's
24:12 why we've been working on a version two of this that is actually human preserved along with whole much of other visuals That make it interesting because I mean in this video right now is in the raw recording is up to about 27 minutes I mean this final video might be 20 22 minutes The question is is someone going to watch this big long video some people will but most people won't I'm really making this video for my own documentation or anyone that wants to dig into this And of course having this video content on push leash As an inter -hidden page with the video schema does add value And the real differentiator here is demonstrating the value the transcript structure and the topical answer fit
24:52 These variables matter across all formats So pivoting to human presenter alone without demonstration content doesn't actually solve the underlying problem And the last thing to look at here is key caveats to hold in my there's hints An absence of evidence doesn't equal an evidence of absolute. There's no published study isolating AI generated versus human made videos Antitation data sets utterly AI bright edge and hrefs all lump youtube together AI citations could exist And simply just be undocumented But it's kind of a risk right you want to deal with a pedally or even having your channel demonetized
25:32 or potentially even having your channel You know removed. I don't know if it's worth it personally. I'm working on building my brand not destroying it Right. Um, so I made this slideshow product But I'm not I'm not really actively rolling it out on my main channel. I might do I might test it on some other channels at one point Cross vendor agreement is partially circular Mostly youtube dominance release And then number two the cross vendor agreement is partially circles Is and then number two the cross vendor agreement is partly circular most youtube dominates figures are tracing back to hrefs bright edge and otterley AI So your first party data
26:12 is much more trustworthy and then figures are a fast -moving window So AI O moved from gemini three in january 2026 and say to incitation so AI O moved from gemini three in january 2026 And citation behavior has been very volatile the key dates are audially AI and march of 2026 Hrefs domain leaderboard and june and june at 26 and se ranking in june And december of 2025 and of course we have that x l r eight study of july 2026 And then side fake doesn't equal revenue because AI referral traffic is 0 .12 percent of the total at this point So AI O citations are
26:52 underperforming and they're showing below So AI citations are underperforming below serp blue rings So AI citations underperform bottom of serp blue rings blue links for clicks So i'm jeremy ashram with push leads. I hope you found this valuable and thank you so
Keep Reading
Want more customers from Google & AI search?
Get a free SEO audit of your site — see exactly what to fix first.