Is YouTube the secret to your AI visibility?

For over a decade, we’ve encouraged clients to host their content on their website and point all other content to their site to benefit from domain authority, but we are now facing a world where visibility is built away from your domain, through journalism, user-generated content and YouTube videos. A year ago, we published an article entitled “Your website is your digital front door”. Today our message is that it’s time to head out.

You may have seen TV presenters leaving broadcast for YouTube-first formats. Holly Willoughby launched her lifestyle channel, Holly Willoughby Together, in July 2026. Alan Titchmarsh started Gardening with Alan Titchmarsh in 2025. In October 2026, physicist and BBC presenter Brian Cox signed a YouTube deal with Sony. Brands are doing the same. In September 2026, John Lewis launched Gift List, a vodcast hosted by Angela Scanlon, on YouTube. They stated that one aim was to make the retailer and its products more visible to AI chatbots. So, should other businesses follow suit?

AI search looks beyond your website

Google’s position is that good GEO (Generative Engine Optimisation) is simply good SEO (Search Engine Optimisation). Its May 2026 guide says optimising for generative AI search is still SEO. Yet the sources AI tools cite often differ from the top-ranking Google results.

A University of Toronto study found that AI search engines show a strong bias towards earned media, meaning third-party, authoritative sources, over brand-owned and social content. 

Muck Rack’s study, What is AI Reading?, points the same way. Its May 2026 edition found that earned media makes up 84% of AI citations across ChatGPT, Claude and Gemini. That includes journalism, academic research, government sources and encyclopaedic sites.

Early experiments support this. Stacker and Scrunch tested eight stories across five AI platforms. When the stories were syndicated across third-party news sites, the rate at which AI cited them rose from 8% to 34%. A larger follow-up in March 2026, covering 87 stories, found a median lift of 239%.

Keeping all your content on your own website, however strong its authority, may limit how often AI finds and cites it.

Why is YouTube boosting brands’ AI search visibility?

If AI prefers third-party sources like news articles, why would a brand’s own YouTube channel help its position?

The answer to this lies in which websites AI cites the most. One industry index of 680 million citations found that 15 domains account for around 68% of AI citations, led by Reddit, Wikipedia, YouTube and LinkedIn (5W Public Relations, 2026). Other studies find citations are a little more spread out. But the same platforms are topping every chart.

Most of these platforms are hard for a brand to use directly. You can’t write your own Wikipedia page, and Reddit communities are wary of marketing. YouTube is different. It’s a third-party platform that AI already trusts, and you control what you publish there.

BrightEdge found that YouTube is cited around 200 times more than any other video platform across ChatGPT, Perplexity and Google’s AI search. It was also the most-cited domain in Google AI Overviews (data from May 2024 to September 2025). This means that there’s a significant argument in favour of companies building up their YouTube content in 2026 and 2027.

YouTube itself isn’t a new opportunity as it has long been regarded as the second-largest search engine after Google. AI search has given businesses another reason to take it seriously.

This feels counterintuitive to many of us who have spent years encouraging our clients to keep everything on their website for maximum domain authority, but times are certainly changing. AI has broken down siloes in many areas, but one key one is the way that it pulls in content from across the digital landscape to bring answers to its users.

How to get started with YouTube content

Marketing managers around the world make a collective shudder when a stakeholder says “Why aren’t we on [insert trendy platform here]?” It’s a question that signals that your workload is going to increase, with new content to create, a new comments feed to monitor and new performance metrics to report on.

However, in the current digital landscape, marketing managers might want to see YouTube a little differently. Creating a presence on YouTube might prevent the other shudder-worthy question of “Why aren’t we appearing in ChatGPT results?”

It’s easy to see why many businesses hold back from video. Not everyone is comfortable in front of a camera, and video takes time, budget and resources that are often already stretched.

Getting started with YouTube content doesn’t need to be a stressful endeavour. To work out what content to put on YouTube, we go back to a few of Google’s foundational SEO best practice fundamentals:

1. Provide a unique point of view – don’t recycle content from others, or produce something that AI could have done for you.
2. Create non-commodity content – provide an experienced take that goes beyond common knowledge.
3. Keep the quality high – ensure your video, audio and script are all good quality as audiences often bounce from low-quality content.

To be able to satisfy these pillars, ask yourself the following questions. This can form the basis of your YouTube content plan.

• What conversations and topics should my brand be speaking on?
• What problems can I solve for a user?
• Do I have the skills and expertise in-house to deliver the production, or is it more time- and cost-effective to bring in a partner?

Production quality still matters, particularly for events, broadcast and flagship series. Authenticity matters too. Seeing the real people behind a business and sharing their expertise in their own words is a strong trust signal for audiences and search engines alike.

Video doesn’t have to mean one person talking to a camera on their own. Roundtable discussions with your team, guest interviews and clips from existing podcasts, webinars and events can all work well. You can then cut shorter clips for YouTube Shorts and social media, so each recording goes further.

To get the most value out of YouTube, remember to use accurate captions. The best way to do this is to upload an edited .srt file. This is an important accessibility feature, as well as the way AI will ingest your content.

Use YouTube’s chapters to break your video up into digestible chunks. YouTube’s studio will generate automatic chapters on eligible videos, but we would advise selecting your own, and titling each chapter in a way that matches a user’s queries on the topic. For example, in a video all about building a time machine, your chapters might include “How to install a flux capacitor” and “Best car for a time machine: why I chose a DeLorean”.

YouTube thumbnails have become a bit of an art form in their own right. The days of simply choosing a still from the video are long gone, and designing a great thumbnail should be considered as important as choosing the video’s title. We’d recommend one clear focal point and bold minimal text. Instead of simply repeating your video title, use a pull quote to grab attention. Remember that this thumbnail will be scaled down to a tiny size on mobile devices, so bold simplicity is the best approach.

What should a content strategy look like in the age of AI?

Great marketing managers never look at one platform or piece of content in isolation, every piece of content feeds into a broader strategy. For brands trying to keep a website up to date and performing, across various social channels and now a fledgling YouTube channel, how can they direct energy to the right places in 2026?

Quality over quantity

The days of churning out content are behind us. High-quality content will cut through the noise and stand out against AI slop. Ensuring each piece is content-native is more effective than posting the same thing everywhere.

Find the efficiencies

Identify grunt work that can be automated or AI assisted. Marketers spending a significant portion of their time on reporting will be feeling burnt out. As an organisation, we cut hours out of our monthly reporting by building a tool that pulls in data from our owned, earned and paid marketing channels, leaving our humans to focus on creativity and strategy with the data to hand.

Remember your E-E-A-T principles

E-E-A-T is Google’s framework for judging the quality of web content. It stands for Experience, Expertise, Authoritativeness and Trustworthiness. Checking your content against these four pillars shows how search engines and people are likely to judge it.

You can also use it to triage ideas, so energy goes to the ones with the most potential. Video is especially good at showing Experience.

Follow the evidence

We are in uncharted territory. The rate of technological transformation is the fastest it has ever been, and so it’s easy to feel paralysed by decision-making. Our advice is to follow the numbers and make evidence-based decisions.

Conclusion

AI has changed where people find answers, and so where brands need to be found. Your website still matters, but it can’t do the job alone. Three fundamentals will help you adapt.

Think beyond your website. AI tools draw on the whole web, and they favour sources other than your own domain. Build a presence on the platforms AI already trusts, including YouTube, while your website remains the place people come to act. Remember, websites are still a great home for content you’re proud of. Embedding your videos on relevant pages can do wonders for engagement time and adding a transcript and video schema helps search engines understand what each video covers.

Go platform-native. Shape each piece of content for the place it lives. A YouTube video needs a clear title, accurate captions and a structure that answers real questions. A webinar recording cut into clips is a start, not a strategy.

Identify your remit. You can’t be everywhere, and you don’t need to be. Decide which topics your brand has the experience to speak on, and focus there. That’s where you’ll earn trust, from people and from AI.

AI search will keep changing. Brands that start testing now will have the evidence to make better decisions as it does. If you’d like help working out where to begin, talk to us.