Beauty discovery has moved off the search bar.
Google's AI overviews now cite 15 sources per answer on average, and it’s 16.53 for Fashion and Beauty queries specifically (SE Ranking, 2026). ChatGPT cites between 10 and 17 (Ahrefs, 2026). More than 60% of beauty consumers now begin their journey through an AI diagnostic, chat, or LLM – not a search bar (BeautyMatter, 2026).
If your brand isn't in those citations, you don't exist.
That verdict came from Show off Your Stack: AI as the Leverage Layer, moderated by Alexa Lombardo, Brand Strategist and Founder of CADRE. The panel paired Emily Rose Campbell, Head of Performance at Iced Media – who runs search and discovery for beauty brands across Google, Reddit, TikTok, and now LLMs – with Grace Clarke, former Head of Community at Shopify, now teaching brands how to actually operationalize AI. Their insights are recapped below.
[Register for the next Show Off Your Stack event]
Emily opened with the line that every CMO needs to hear: "Advocacy is actually the number one most important thing when it comes to driving AI visibility for your brand."
LLMs don't surface brands by domain authority; they surface them by community signal: Reddit threads, YouTube tutorials, retailer reviews on Ulta and Sephora, best-of lists from Allure and Byrdie, and PR coverage. That's what gets cited.
"These LLMs are crawling the web for community signals to drive visibility," Emily said.
For beauty specifically, the top citation sources are consistent: Reddit, YouTube, retailer PDPs, and editorial PR.
The implication is major for legacy brands. If your growth model was paid media plus a polished corporate site, you have nothing for the LLMs to cite. Authority in the AI era is earned. It lives in the communities, creators, and Advocates outside your channels. The brands that didn't invest in those communities are struggling.
An Iced Media client illustrates this. Their website drives 17% of citations, which is strong by D2C standards. Meanwhile, their competitor drove 4% of their citations from their website. But the competitor was still winning overall because the other 96% came from off-site sources including PR, Reddit, and YouTube.
You can't optimize your way out of this.
That's exactly what Brand Advocacy at scale produces: third-party content, third-party reviews, third-party endorsement distributed across the channels LLMs trust. Brands running an advocate-first operating model have been building the AI visibility foundation for years without knowing it. Brands still running channel-first marketing are finding out what the gap costs.
One of the best points of reflection came from Grace: "AI is not the end game. It's leverage. To what end?"
That question reframes the AI conversation. AI is a multiplier, not a fix. Leverage it with a weak business and you scale the weakness. Leverage it with a strong one and you scale the strength.
Emily put a finer point on it. The brands winning AI search visibility aren't winning by spamming content. They're winning by being deliberate about what they "deserve to be famous for" by building topical authority around hero products where they already have community equity. Then they use AI to scale: blog content (which accounts for ~40% of D2C citations), FAQ pages, internal linking, and structured data.
Grace also clarified how AI should fit into your business: AI doesn't run your business, rather you teach it to think like you. You feed it your Klaviyo data, Shopify data, call recordings, and strategy documents. You build SOPs and workflows. You audit your channels with it. None of it matters if the business underneath isn't built to scale.
"AI is amazing," Grace said, "but for it to work, you have to work for it."
Translate that into Brand Advocacy terms. AI is what makes advocate-first operating models scale. It leverages personalization, surfaces patterns in Advocate behavior, audits engagement, and identifies content gaps. But the foundation – the Advocates, the relationships, the community – has to exist first. AI doesn't generate Advocates. It amplifies what you've built.
[Learn more: Brand Advocacy Isn't a Channel. It's Infrastructure.]
Grace’s most actionable prediction: "I want every person on every team in the next month or two to have an agent that is personalized to them."
She means an always-on copilot – one that records calls, pulls data, queries it on demand, sits next to each person on the org chart. The hybrid org. Humans + Advocates + agents.
The math is uncomfortable. Grace was direct: "Do I think companies today could run with one or two people on top of an AI-powered operating system? Yeah, most I do."
She points to Meta's announced layoffs, heavily weighted toward management, as the early indicator. Small businesses, she argues, will have to navigate the same territory.
This isn't a layoff prediction. It's an organizational design challenge. The brands that win the next 12–24 months won't be the ones with the most headcount. They'll be the ones that orchestrate the largest network of humans, Advocates, and agents working together.
Brand Advocacy programs already operate this way. They scale brand-building across thousands of people who don't sit on payroll. Advocates are, in a real sense, the first agent layer brands ever had – human ones. Adding AI agents to that mix is the logical next move. The brands already running advocate-first will adopt the agent layer fastest, because the operating model is already distributed.
For a decade, Brand Advocacy was framed as nice-to-have: an organic growth lever next to paid, retail, and influencer.
AI just made that framing obsolete.
If LLMs are how half of consumers now discover products, and LLMs cite community signals, then the brands without a Brand Advocacy footprint aren't underperforming. They're invisible.
That's the verdict. AI doesn't replace Brand Advocacy. It exposes which brands built one and which didn't.
The leverage layer is real. So is the need for a system built on Advocacy. The brands that build advocate-first – with distributed authority, community-led content, and third-party endorsement at scale – now have the infrastructure the AI era rewards. The brands still running channel-first will spend the next 24 months trying to retrofit a strategy they should have built five years ago.
You can't AI your way to advocacy. But you can advocate your way to AI relevance.
Your Brand Advocacy Score tells you where you stand. It measures the depth, reach, and quality of your advocacy footprint. The exact signals LLMs are crawling for.
Find out if you're visible, or find out where to improve – while you still have time to do something about it.