How creator marketing builds brand visibility in AI search

TL;DR

• AI tools like ChatGPT, Claude, and Gemini are reshaping how consumers discover and evaluate products, and 90 to 95% of what gets cited in those answers comes from third-party sources like creator content and community conversations, not from brand-owned pages.

• Share of model (how often and how prominently your brand appears across relevant prompts in major LLMs) is the foundational AEO metric every marketer should be tracking right now. A healthy share of model starts at 30% or above for your highest-priority prompts.

• Effective AEO strategy requires two distinct types of creators: citable authorities who build long-form expert content, and community validators who generate the social proof and consensus signals that LLMs treat as trust indicators.

AI search visibility, AEO, GEO, LLM search: whatever you prefer to call it, you’ve likely heard how AI is rewriting the customer journey your marketing strategy was built around. 

People still discover products through social media, but the step that used to take them from discovery to Google is now increasingly taking them to a large language model (LLM) like Claude, ChatGPT, or Gemini instead. 

What these AI tools surface in response to consumers’ product and brand questions has almost nothing to do with your brand website, your blog, or anything you've put direct investment into owning.

That’s the challenge we set out to help brands solve in our latest installment of Made You Look, our monthly live conversation series built for marketers who want more than surface-level trend reports. 

In our latest expert session, Later's Social and Community Manager Kayla Monis sat down with Lyle Stevens, Later’s co-founder and Chief Strategy Officer at Later, and Carly Baron, Chief Commercial Officer at Advantice Health, to unpack what AEO (answer engine optimization) actually means for brands, how to measure where you stand today, and what it takes to close the gap through a creator marketing strategy built specifically for the AI era.

Read on for what they shared, or watch the full expert session on demand.

1. Understand how the customer journey has already shifted

The first thing Lyle made clear is that the shift in consumer behavior isn't coming; it's already here. Product discovery still happens primarily in social media, on TikTok, Instagram, and YouTube, but the next step in the journey has changed. Consumers who would previously have taken their curiosity from social to Google are now increasingly turning to AI instead. "55% of consumers are now using AI specifically in tandem with their product discovery on social media," Lyle noted, citing recent data, "and one in two have actually made a purchase based on what they've used AI to research."

That shift has real implications for where brand visibility actually matters. We already know that AI is influencing purchase decisions. The question is whether your brand shows up when it does.

2. Stop treating AEO like an extension of your SEO strategy

One of the most consistent patterns Lyle sees across the brands he works with is an instinct to port over SEO knowledge and apply it to AI search. The logic is understandable, but the execution is largely misdirected. "Most of the content that is being cited by these LLMs is coming from third-party sources," he explained. "To the tune of 85 to 90% of what gets cited is coming [from] elsewhere; it's not coming from your own brand."

Your brand's own website, blog, and product pages tend to show up third, fourth, or fifth in LLM responses to category-level questions. The content that ranks first is creator validation and community consensus, typically hosted on YouTube, LinkedIn, Reddit, and Substack. Investing primarily in owned content while those platforms do the heavy lifting for your competitors is a meaningful strategic gap.

3. Track share of model before you try to move it

Before any AI search strategy can be optimized, brands need a baseline. Lyle described share of model as the AEO equivalent of share of voice in social media, measuring how much of the conversation happening in your category across major LLMs your brand owns. He breaks it into three nested metrics: 

• Share of model (does your brand appear in the conversation at all around relevant prompts?)

• Mention rate (how do you show up against a specific prompt?)

• Citation rate (when your brand is mentioned, is it accompanied by a source URL, or just referenced in the text?).

"A healthy share of model is usually north of 30% to start with," Lyle explained. "It means you're in the conversation and you're owning a third of it." He tracks this across Gemini, Claude, and ChatGPT as the three primary LLMs, noting that Perplexity essentially aggregates from all three, so strong performance on that set translates broadly.

4. Audit your prompts the way you once audited your keywords

The practical first step for establishing a share of model baseline is to bridge your existing SEO data with the prompt patterns your consumers are actually using in LLMs. Lyle recommends starting with your highest-traffic SEO keywords and working from there to construct a prompt list of 10 to 20 queries across four categories: buying intent prompts, competitor comparison prompts, feature or RTB-related prompts, and broad categorical prompts.

Carly described how Advantice Health approached this in practice. "We kind of reverse engineered it," she said. "Going back to the prompts, we all understand our consumers, so we thought about what types of questions they might be asking, and created a prompt list we thought would make the most sense for each of our brands." She acknowledged that Advantice Health is still working toward a fully quantified share of model view, but the prompt-based audit gave them enough signal to understand where they stood, including some meaningful surprises in what was already working organically on forums like Reddit.

5. Recognize the two creator types that drive AI search results

Lyle introduced a framework that's become central to how Later approaches AEO strategy: the distinction between citable authorities and community validators. These two creator types address different elements of what LLMs evaluate when constructing a response.

Citable authorities are creators, professionals, or credentialed experts who produce long-form content with enough depth and authority that LLMs treat it as a reliable source. They typically anchor their expertise in video, long-form written content, or a body of work that establishes credibility in a specific category. 

Community validators, on the other hand, build the surrounding consensus: the internet forum threads, product reviews, and community conversations that signal to an LLM that a particular perspective or recommendation has broad, real-world backing. "It's sort of like the wisdom of the crowd," Lyle explained, "mass perspective and opinion around the product or the brand." Both types are necessary. Neither works as well in isolation.

For a brand like Advantice Health, which spans categories from skin care to baby care to postpartum health, that mix matters especially. "Those are exactly the conversations that people are going to LLMs for," Carly noted, "because it's like your doctor, it's your expert, it's like the friend you want to talk to when you need that really important information."

6. Brief creators around problems, not products

One of the clearest through-lines in the conversation was how the approach to creator briefing has to shift in an AI search visibility context. The most LLM-friendly creator content starts the way a search query starts: with a question or a problem rather than a brand mention. "Ultimately, the most engaging and authentic content that we're seeing from our creators is content that starts with what feels like a very natural question or problem that the creator is then going to talk about and solve for you," Carly said.

Lyle framed this the same way from a strategic standpoint. "I always start at the problem set, not at the product solution," he explained. "Your consumer seeking information about the problem that they're trying to solve with your brand is usually the first place to start." When creators frame content around the questions their audiences are already asking, that content is more likely to surface in the prompts consumers are typing into LLMs, creating the virtuous cycle that makes AEO compound over time.

Carly added that Advantice Health has had to calibrate how much structure they bring to creator briefs. Too much direction produces content that feels scripted and earns lower engagement. "If we brief a dermatologist and they're too cookie-cutter, it doesn't get a ton of engagement in general," she explained. For example, a dermatologist they recently worked with for a product in their Triple Paste eczema line had a child with eczema, which gave her a personal and specific lens that made the content more resonant and more authentic, a meaningfully different result than a clinical-style product walkthrough would have produced.

7. Treat reach and AI brand visibility as connected, not competing

A question that came up from the audience cut to a tension many brand marketers are feeling right now: do you need separate influencer programs for reach and engagement versus AI search visibility, or can a single strategy serve both? Lyle's answer was that they're complementary rather than competing, though they do require different creator profiles and different briefing approaches.

"You need reach and awareness first and foremost for folks to even discover your product and even start asking questions in LLMs," he said. "You don't want to just shut it off." The opportunity is in finding creators who can do both: those whose content has enough virality to build awareness and enough authority to earn LLM citations. He described this as "a two for one" and positioned it as upside in the strategy rather than a design principle. Carly added that working with Later has allowed Advantice Health to brief multiple content lengths to the same creators, extracting both long-form authority content and shorter-form awareness content from a single engagement.

8. Prioritize YouTube and Substack alongside forums like Reddit

Reddit is the most-cited third-party source in LLMs for consumer product categories, and that reputation is well-earned. Lyle pushed back gently, though, on the instinct to treat Reddit as the only channel worth optimizing. YouTube is growing its citation rate faster than any other platform, up over 300% year over year in terms of how frequently it's cited by LLMs, Lyle noted. Substack and blogs together account for roughly 10% of all LLM citations and have seen a meaningful resurgence as the AEO conversation has matured.

"I would look at YouTube. I would also not sleep on Substack and blogs," Lyle said. "Bloggers have had a resurgence in the last year because of this. They were like the OG influencer back in the day." His broader point was about diversification: no single platform should be the sole pillar of an AEO creator strategy, both because the citation landscape shifts and because over-indexing on one channel creates fragility. Carly reinforced this from the brand side, noting that the channel weighting that dominated conversations a year ago (Reddit, Reddit, Reddit) has already started to evolve.

9. Start smaller than you think you need to

A recurring theme across the Q&A portion of the session was how brands can apply this framework when they're working with limited creator budgets. Both Lyle and Carly made the same point from different angles: the scale of creator investment needed to have an impact on AI search visibility is more modest than most marketers assume.

"You don't need hundreds of creators to move the needle," Lyle said. "We're talking a dozen or a couple dozen, max." Carly pointed to Advantice Health's own trajectory as evidence. A year ago, the company worked with roughly 50 creators across five major brands. By the end of this year, they're on track to work with more than ten times that, but the results started coming before the scale did.

For brands at the earliest stage, Lyle's recommendation was to look inward first. Existing customers who already love a product and have any kind of social following can be powerful early advocates, and an elevated brand experience or early product access can get the ball moving without significant spend. Carly added that a DM, a sample, or a handwritten note to someone who's already mentioning your brand organically can create a long-term ambassador relationship at almost no cost.

10. Treat LLMs as the most influential channel your brand isn't briefing yet

The most resonant reframe from the session came near the end, when Carly offered a single way to think about where AI sits in the broader marketing mix. "Think of your LLMs as the most mega influencer that's out there," she said, "and make sure they're educated with the right info." Lyle had made a similar observation earlier when he noted that one in three product discovery moments now happens in AI, and that 50% of those interactions lead to a purchase.

The implication, as Lyle framed it, is that waiting is a competitive liability. LLMs train on historical data, and brands that establish credible third-party signals now will benefit from that positioning until the next model version retrains. "You want to be at the bleeding edge of this," he said. "You don't want to be a laggard in thinking about AEO because you will just see too much space for your competitors."

What LLM search really measures

LLMs are doing what consumers have always done: looking for signals that a recommendation is credible, consistent, and comes from voices they can believe. Creator content, when it's briefed around real problems and delivered through authentic voices, provides exactly those signals. The brands that understand this will build AEO into their influencer strategy from the ground up, rather than treating it as an add-on after the fact. Want to hear more from Lyle and Carly? Watch the full session now on demand.

If you want to go deeper on any of this, including how to structure a creator program that builds LLM visibility alongside traditional influencer KPIs, schedule a session with our creator AEO strategy team.