The 3 Sites LLMs Trust Most. Kaleigh Moore on AI Search.

Kaleigh Moore has been writing for B2B SaaS for 12 years, reporting for Forbes and Vogue Business, and is now a Harvard graduate student studying LLM information retrieval. Here's what she sees coming for B2B brands trying to win AI search.

https://www.youtube.com/watch?v=WJbffU_bN9U

Key Takeaways

• LLMs pull from Reddit, YouTube, and LinkedIn first because community validation is what they trust.

• Video is the antidote to AI slop. People buy from people, and that signal can't be faked at scale.

• Paid ads keep getting more expensive. Video is a long-term play that captures AI citation white space.

• The brands moving now have a first-mover advantage. Models are constantly retraining on fresh content.

• The starting move is an AEO or GEO audit, not a campaign.

Why LLMs trust community-driven sites

The top three sources LLMs cite are Reddit, YouTube, and LinkedIn. Kaleigh saw this in a recent report covering ChatGPT Plus, Perplexity, and one other engine. The reason is simple: those platforms are community-driven and third-party. Brand-owned assets carry less weight because the model can't tell if you're just selling.

"It's essentially community-driven, right? It's third party, it's not brand-owned assets, this is community validation, which is what LLMs trust the most." — Kaleigh Moore

For B2B, that matters more than for most categories. People buying software are making large, multi-year decisions. They want to trust the company and the culture before they sign. Community signals on YouTube and LinkedIn give them that proof. If your brand isn't showing up there, the LLM doesn't have much to cite when a buyer asks about you.

This is why employee advocacy keeps coming up in the AI search conversation. A VP of product talking on LinkedIn is third-party signal in a way that a brand campaign isn't.

Video is the antidote to AI slop

The 2025 word of the year was "slop." That's not an accident. AI promised 100x text output and a lot of brands took the bait. The result was a flood of low-quality content that didn't move the needle on search or sales. The pendulum is now swinging back.

Video sits on the other side of that pendulum. It's harder to fake at scale. Deepfake CEO videos will get caught and will burn trust the moment they do. Real video, real subject matter experts, real conversations, that's the format AI search can lean on.

Video is also the best raw material for repurposing:

• One 15-minute interview becomes 3 to 5 short clips

• The audio becomes a podcast episode

• The transcript becomes a blog article and FAQ

• Each piece feeds back into LLM training data and citation surfaces

If a B2B company can only pick one investment, video covers more ground than any other format. It's not just a content piece, it's the input for everything else.

What to do when subject matter experts won't go on camera

The most common pushback Kaleigh hears is that internal experts don't have time. That's real. VPs of product and senior developers are busy, and a lot of them don't love the idea of being on LinkedIn. Two workarounds work today:

• Partner with B2B creators or thought leaders. Borrow their reach and their comfort on camera. Long-term partnerships with people who already publish in your space can stand in for the employee advocacy you can't get internally.

• Batch short audio answers. Pick a list of search terms your buyers are typing into ChatGPT. Have the SME read the question and the answer, audio only, with captions on screen. Each clip is one question, one answer, maybe 30 to 60 seconds. They can do five or ten in a single session.

Both routes solve the same problem: get the expert's voice and authority into a format the LLM can find and cite, without forcing them into a polished on-camera workflow they'll never sustain.

Why video is a survival play, not a marketing trend

Paid ads keep getting more expensive. Cost per impression is climbing in every B2B category. The math gets worse every quarter. Meanwhile, AI search is still wide open in most niches. There's white space available for citation rate and share of voice that won't be available in 18 months.

"There's a lot of white space here that's still left to be claimed when it comes to the AI visibility and citation rate stuff. So I think video is a really smart play. Again, not a quick win, but a smart long-term investment." — Kaleigh Moore

The brands moving now are seeding the next wave of LLM training. Models are constantly assessing new information. Brands that publish steady video content over the next 6 to 9 months will be cited more often as the models retrain. Brands that wait will find the slot already filled.

This is closer to a survival strategy than a marketing campaign. If competitors get cited and you don't, you're handing them every AI-driven discovery moment in your category.

How to start: audit your AI visibility

The first step is not a content calendar. It's an audit. There are AEO and GEO tools that show where your brand is cited, what your share of voice looks like across ChatGPT, Perplexity, and Gemini, and where your competitors are showing up that you're not.

That audit gives you a roadmap. You can see the questions your buyers are asking, the questions your competitors are answering, and the gaps you can fill. From there, build the video plan around those gaps. Not the other way around.

Kaleigh's advice for B2B brands: start with the audit, then go from there.

Frequently Asked Questions

What sites do LLMs cite most for B2B information?

Reddit, YouTube, and LinkedIn are the top three sites LLMs pull from for B2B research. They're community-driven and third-party, which is the kind of signal LLMs treat as trustworthy. Brand-owned content carries less weight because it lacks community validation.

Why is video important for AI search visibility?

Video proves humanity, which is the antidote to AI slop. People buy from people, and that signal is hard to fake at scale. Video also gives B2B brands the most repurposable raw material, since one interview becomes shorts, audio, transcripts, and blog posts that all feed AI search.

How can B2B companies get subject matter experts on video without disrupting their schedule?

Two workarounds work today. Partner with established B2B creators who already publish on camera, or batch short one-question, one-answer audio clips where the SME reads the question and answer with captions on screen. Both let you publish expert content without forcing busy employees into a full on-camera workflow.

How long does it take to show up in AI search results?

This is a 6 to 9 month investment, not a quick win. LLMs are constantly retraining on new content, so brands that publish consistent video content now will be cited more often as models update. The brands waiting on results in week one are measuring the wrong thing.

Where should B2B brands start with AI visibility?

Start with an AEO or GEO audit. Tools exist that show your citation rate, share of voice, and where competitors are appearing in AI search results. The audit reveals the gaps. Build the content plan from those gaps, not from a generic calendar.

Watch the shorts

Each short answers one specific question from the interview:

• How does video help you show up in AI search?

• How do LLMs decide which sources to trust?

• How long does it take to show up in AI search results?

• What is a good B2B video strategy?

Full Interview Transcript

Dane: Hello everyone, my name is Dane Fredrickson. I am a B2B video expert and I am joined today by Kaleigh Moore, who is a thought leader on LinkedIn regarding subject matter experts getting out there to create the content that's going to get you found in AI search, among other things. What else do we need to know about you, Kaleigh?

Kaleigh: So I have been in content marketing for B2B SaaS for the last 12 years. I've worked with companies like Stripe, Shopify Plus, Amazon Business. I've also been a journalist reporting for places like Forbes, Vogue Business, AdWeek, pretty much all the major business publications. I'm currently a Harvard graduate student where I'm studying philosophy and ethics, specifically LLM information retrieval. So coming at this from a lot of different angles. But yeah, lots changing with AI and I feel like right place, right time.

Dane: Yeah, it's good to be talking about this with you right now. I think the wind is blowing in our direction right now with the AEO, GEO, whatever we're going to call that thing. And then now we have like a new layer, this agentic search need, which I forget what they're calling that. So it's alphabet soup. It doesn't really matter.

Kaleigh: I know, I can't keep it straight, really.

Dane: Yeah, but really what we're talking about is like we need to get found by people and by robots. And those are different disciplines with a lot of different overlap. And I've been coming at this from the video angle. That's what I do. And you've been coming at it from your angle, but I think there's a sweet spot right here in the middle where we're talking about getting subject matter experts on camera to record video content, as well as the transcripts and the text content. That's all going to feed back to the people in the machines.

Kaleigh: Yes.

Dane: That's what I wanted to talk about. What's your sense about like, where are we now in the adoption? Let's just say B2B SaaS. Like these are people that don't really like marketing to start with, but I think now we're getting the data that is sort of like pretty obvious. It's a need to have now, not a nice to have. Is that sound right to you?

Kaleigh: Yeah. Yeah, I mean, I think a lot of B2B buyers are already adopting AI search engines as tools for discovering and learning about products that they want to buy. The stance that I'm taking on this is that people like to buy from people, so I think this is what you're talking about as well. Getting people on video, leveraging your humanity in this world of AI, that is one of the most advantageous things you can do. And so in my case, I'm talking about people who are internal at companies, VP of product, developers, people with a lot of hands-on product experience, I think those are really smart play at getting them on LinkedIn, building kind of the thought leadership mentality in those types of employees, and earning AI visibility through those efforts. So that's where I'm focused. I think video is a huge piece of this, so I wanna hear your perspective too. Why do you think video is so important to all this AI search visibility conversation?

Dane: Yeah. I come at it from, you know, obviously a lot of bias. I know intuitively a lot of the benefits video provides outside of the AI visibility part. Like, let's just put all those other things to the side for a moment and just focus on the AI search result thing. It's like, you've got this amazing ability with video to repurpose the images, the audio and the transcripts. Just off the cuff, it's like the best format for harvesting. If you could only do one thing, I wouldn't be investing in a writer right now. I'd be investing in someone who's going to help you harvest this goldmine that you've got of your subject matter experts and turn that into video content, transcript content that gets repurposed into blogs. So that's like the big thing I've been advocating for a while. And I think we're at this moment now when, at the end of 2025, the word of the year was slop. And such a perfect moment for calling out the promise of AI did not deliver on the 100x output of like, let's just write our way out of this and slop our way to visibility. That didn't work. So now you're seeing this backlash. I think the pendulum is swinging the other way, or maybe we're at an inflection point where you got to choose, are you going to go that way and just automate your way through text only? Or are you going to actually do quality content, which I think includes video and invest in that path? Talk about an antidote to slop. I mean, I think that's really what we're talking about here is like the human connection. This is still, I would say impossible to consistently. You might be able to do some deep fake kind of stuff here where you're like pulling it off for a little while. Like, hey, my CEO said this and it's actually a robot. We're not gonna buy into that stuff. You're gonna have a significant backlash against these AI avatars. People that are trying to do that. I think it's a huge mistake. It's gonna eat your trust immediately.

Kaleigh: I do too.

Dane: So that's sort of long-winded answer how I'm seeing this right now. I think the challenge for us, you and I is to convince these leaders that the subject matter experts need to be compelled, incentivized to go on camera. Cause a lot of people don't want to do that. You know, they're shy, they're too busy. They don't think that this webcam thing is good enough quality. There's all kinds of excuses, but like, if we start to see this as like a need to have, like a matter of survival, if you need to show up in AI search results at all, I don't see how you do this without including video because your competitors are going to, and if you don't step up to the video plate as well, you're sort of conceding that market or that visibility that's gonna come from that. And I just think of it as like coverage. You're not covering that square. It's like you're driving a three-wheeled car. That's not gonna get you where you need to be. So I don't know, does that resonate to you in this B2B SaaS space? Are these people buying that yet? Are we still trying to convince?

Kaleigh: You know, the workaround and the pushback that I get when I say employees need to be on video and on LinkedIn specifically is that they don't have time for that. And so I think another creative workaround and supplemental piece to all of this is that there are B2B influencers, creators, people who are thought leaders within their subject matter expertise who can be a stand in for maybe the companies that don't have the willingness or have the employee buy-in. So there are other ways to kind of borrow leverage here and to, again, this needs to be thoughtful, needs to be authentic, it needs to not just be a cash grab or an impression play. But long-term partnerships with established creators, I think, are a nice workaround for all this. And again, they're great at video, they're great on camera. This is work that I'm already doing for myself and seeing pretty good results. So there are a lot of different ways you can tackle this beast, but I do think video needs to be a piece of the puzzle.

Dane: Yeah, and there's a new way of doing this, or maybe not that new, but I've been thinking about a new approach here. If you don't have the ability to get your SMEs on camera and you don't want them to spend their time writing this stuff, what we could potentially do is come up with that targeted list of search terms that people are searching for. They really concise one question, one answer, not these long rambling pieces. And if you could get subject matter expert to just read the question and read the answer in audio only and then do graphics maybe just the captions or some sort of like not on camera solution you could probably do those in in a batch where it's basically give you this ammunition for getting found without having to you get them on camera and then if you incentivize them to just take these one little videos and post those on their LinkedIn account. I think you could do that in a matter of seconds really. So like it's something I think it needs to be tested, but I think this is one it's got to be like part of the solution for some of these companies because I just run into this all the time. I'm too busy. I don't want to go on camera. I'm shy. I don't like my camera. I don't like my audio like all that stuff. So this feels like a good work around that we can test.

Kaleigh: Yeah, I love that.

Dane: I'm curious if you have a sense of like, who is doing video for AEO search correctly now? Is there anyone that comes to mind or like a thought leader who, you know, other than us, that's like advocating for a video approach in this way? I'm not really seeing that yet. And the companies that I am seeing that are doing some of this right, I don't think they're doing all of it right. Like HubSpot came to mind. I think they've done a lot over the years. They've got multiple channels on YouTube. I think they're kind of doing that part of this, but that little like one question, one answer thing I just talked about, I don't think that they're quite doing that yet unless I missed it. So sorry, I'm rambling, but like are you seeing anyone that's kind of doing this the right way?

Kaleigh: I've seen a few brands who are tackling this through a lot of personality driven content. So it's not necessarily a direct AI search play, but it is the humanity piece. So they'll do person on the street interviews at a conference. I know AirOps did that. They did a trivia game. Gong has a great LinkedIn presence where they're incorporating a lot of video. I do think there's an experiment here that needs to be done as far as what's the right mix on first pure. What are the questions our audience is asking and how can we answer those with video? But also, how can we showcase our company culture and how can we get our employees building rapport with people through that personality driven content? So I think there could be an interesting mix here where you're ticking all the same boxes but through different mediums and platforms and channels for doing that.

Dane: Yeah, it's something that I've been building for some time now. Past couple of years, I've been realizing that B2B companies need help with video. They're not equipped to come up with a good strategy for the best ways to go about production of this at scale, sustainably, make those tough choices about how good does this need to be really? Or is this a live action thing, people on camera? Or is this an animated explainer video? All these people in decision making roles that are not video experts, I don't know how they're gonna be able to do that. So I'm trying to help these companies like roll out a video strategy. And I think this GEO angle is the closest I've been to actually getting buy-in that this is a need to have, not a nice to have. And I'm wondering like, what is that slam dunk moment for you? Like, is there a piece of data or study that we can be like, hey, it's time, like does something stand out to you?

Kaleigh: Yeah, yeah, I just read a report this morning that said I think Reddit, YouTube, and LinkedIn are the top three sites that LLMs are pulling from. And again, this is a selective bucket. I think it was OpenAI or ChatGPT Plus, and then Perplexity, and I can't remember the other engine. But those are the top three platforms because they are so, it's essentially community driven, right? It's third party, it's not brand owned assets, this is community validation, which is what LLMs trust the most. And that to me is a clear case that moving forward, brands have to have this, especially B2B where people are making really big purchasing decisions about software, these are large scale investments, this is not something you make with a click one time. You really need to trust the company and the culture that they have there that they're gonna be around for a long time and again, that's another piece where the employee advocacy comes in. You're building trust and rapport for the long term. And so I think there's a lot of different ways you can solve for this.

Dane: Yeah, I'm curious too, like the people in the companies that get what we're talking about and that are sort of bought into this, they're gonna have to probably convince upwards, right? There's some sort of financial arguments and data argument. They're like, a skeptical CEO needs to be convinced. Like, have you had experiences like doing that? Like, cause I feel like that's one of the headwinds we're kind of up against here is like, yeah, we get it. But how do we slam dunk this point?

Kaleigh: Yeah, I think one of the big things to touch on in this conversation is that paid ads are getting so expensive and cost per impression is getting very high. And so if you look at the creator angle where what's the conversion rate, what's the return on investment for those types of engagements or video, what's the, again, it's not a quick win, it's a long-term play. So that's a little bit of the education piece that needs to happen as well. Just talking about, let's think six to nine months down the road and where we need to go and start making steps in that direction because the paid ads, it's just not sustainable. It's getting too expensive, it's getting too competitive. There's a lot of white space here that's still left to be claimed when it comes to the AI visibility and citation rate stuff. So I think video is a really smart play. Again, not a quick win, but a smart long-term investment towards kind of capturing that white space.

Dane: Yeah, and I think it's also not just investment, it's like a survival strategy for seeding the future LLMs that are going to be trained on this stuff. Like the people that are doing that now, I think they're going to have a big leg up as a first mover advantage thing, I think.

Kaleigh: Right, right. Mm-hmm, for sure. I think so too, I think so too. I think the models are constantly assessing new information, so again, this is not something you do once and then never touch again. It's something you have to be investing in long term. It has to be ongoing, but yes, the people who are starting now, like I said, are capturing this white space, and so it's only gonna become more competitive from here on out. There's never been a better case, I think, to start right now, start today.

Dane: Yeah. So as we wrap up, I'm curious if there's anything that comes to mind as like a next step for someone watching this video, they're like, yeah, I'm buying into this. Now what? What's the next step for them? And I'm just going to throw out, maybe it's an audit of what you have and what's working, what's not. And then you can kind of build from there.

Kaleigh: Yes. Yep. Yeah, there are a lot of tools you can do an AEO or GEO assessment through to just kind of see what your standing is, where you stand amongst your competitors as far as citation rate, share of voice, lots of tools that do that. I think that's a great place to start. And again, you can kind of look at what you're showing up for and what your competitors are showing for, combine the two and find where the gaps are. So that's kind of gonna give you a roadmap for where to start. And that's what I would encourage all B2B brands to do, start with the audit and go from there.

Dane: Okay, that's great advice. Thanks so much for taking the time and sharing your insights today. If people need to get ahold of you, can they find you?

Kaleigh: KaleighMoore.com or on LinkedIn where I spend way too much time.

Dane: Awesome. All right, well, thanks again. Take care.

Kaleigh: Thank you.