Episode 35: The State of AI Search & What Top Brands Are Doing Differently

Key Takeaways

• Blogs outperform product pages as AI entry points
• Brands must consider the growing importance of YouTube and video in training AI models
• There are new AI-era KPIs: Share of Voice, mentions per prompt, and ROI per prompt
• Brands should now map prompts to revenue using real customer insights
• Industries like travel and hospitality face the biggest disruption risk

Projections reveal that 60 to 70% of search traffic could shift to AI-powered channels within the next 8 to 18 months. This makes AI search one of the fastest-growing customer acquisition channels we’ve seen in years.

In this episode of the Local Marketing Beat podcast, host Christian Hustle sits down with David Zeledon, Head of Growth at Athena HQ, to break down the findings from Athena’s State of AI Search 2026 report. David shares why industry leaders are being included in AI-generated answers 56% of the time — while the average brand sits at just 17% — and what multi-location brands can do to close that gap.

‍

‍

Timestamps

00:00 The rise of AI search and traffic decline predictions

00:58 What top brands do differently in AI search

03:08 From clicks to influence: the new funnel

05:55 New KPIs for AI search success

08:45 YouTube, context, and AI training data

12:24 Why blogs are the #1 AI entry point

14:02 Industries most at risk in AI search

15:09 GEO strategy for CMOs in 2026

17:06 Final takeaways and where to follow

The Shift from Keywords to Answers

David Zeledon: “The top brands that are winning have had two things really in common. One is they’ve taken content as a strategic approach, and they leverage it across different channels and surfaces. AI systems are trying to gather those ingredients from all of those surfaces. And then they actually create the brand’s lexicon, their universe, from these touch points, from these citations, from these mentions.”

David explains that leading brands treat content as a cross-channel asset — not something siloed on a blog. They distribute it across social, PR, community forums, and their website. AI systems then pull from all of those surfaces to build a picture of who the brand is, what it offers, and whether customers are happy. External sources like reviews, mentions, and third-party editorial content are becoming critical decision points for how AI evaluates and recommends businesses.

For multi-location brands, this means every touchpoint matters. Your listings, your reviews, your social posts, your PR coverage — AI is reading all of it. The brands already investing in these areas are seeing compound returns now that AI search has entered the picture.

Beyond Zero Click: the New KPIs That Matter

David Zeledon: “Beyond the zero click, I think you start to hone in on where you can start to provide the answers for people. Understanding where you can embed yourself across the journeys offsite to then eventually have that traffic come back to you through brand recall, through being in other answer engine recommendations, comparisons — those types of prompt behaviors that really lead to a sales capture or a conversion capture.”

Zero-click search is nothing new, but AI has accelerated it dramatically. AI systems can synthesize hundreds of documents and dozens of keywords from a single prompt query, delivering a direct answer without the user ever visiting a source. For brands, the implication is clear: Traditional click-based metrics no longer capture the full picture.

The AthenaHQ report introduces new GEO metrics including Share of Voice in AI, mentions per prompt, and ROI per prompt. David suggests that brands work backward from actual customer conversations — (why did customers choose you?) — and map those reasons to the prompts where AI is (or isn’t) recommending them.

This thinking aligns with how Uberall approaches Location Performance Optimization (LPO) — connecting visibility, reputation, engagement, and conversion into one measurable framework. Even as clicks decline, brand recall and recommendation frequency become the signals worth tracking.

YouTube: the Underestimated AI Training Ground

David Zeledon: “When you combine the power of video and content, it creates this context universe that’s really rich. YouTube is a really great, rich resource to train models on, but also to understand how your brand is being perceived across the industry. It’s still the number-two search engine in the world.”

Much of the GEO conversation has focused on Reddit, Wikipedia, and traditional web content as AI training sources. But David highlights that LLMs are increasingly leveraging YouTube transcripts to learn about brands and industries — and most businesses haven’t caught on yet.

AI systems read video transcripts and use them to build richer context about what brands do, how they serve customers, and how they’re perceived. David shares examples from the food and CPG space: brands that use detailed descriptors (aromas, textures, vibes) in both written and video content are significantly outperforming competitors in AI-generated recommendations.

For local and multi-location brands, this is a signal to invest in video content that goes beyond surface-level promotion. Think about what questions your customers are asking, and answer them on camera. Those transcripts become another surface for AI to discover and cite your brand.

Blog Pages: the Number-One AI Entry Point

David Zeledon: “Blogs have had this abstraction value to businesses where even if you’re writing about a competitor, you kind of have to have some type of ethics behind it. Category pages and product pages tend to be pretty salesy, and AI systems stay away from those.”

The Athena report found that blog pages account for 44% of all AI-sourced entry points — making them the single most important content type for AI visibility. Product pages and category pages, by contrast, are largely ignored by AI systems because of their promotional tone.

This reinforces something we’ve been seeing with GEO optimization: AI rewards content that informs rather than sells. Brands that organize their content hubs into clear topics, themes, and subcategories — supported by rich FAQs and links to technical documentation — give AI systems the structured, trustworthy material they need to generate accurate recommendations.

David recommends thinking of blog content as the foundation of your AI visibility strategy. It’s where AI looks first, and it’s where brands can demonstrate authority on the topics that matter most to their customers.

Which Industries Are Most at Risk?

David Zeledon: “Travel and hospitality comes to mind. When you think about AI search and those surfaces, it’s going to be really pertinent to become a part of that booking answer. You run the risk of not being seen anymore, not being found. When you think about brand recall, being in the answer is really powerful because it brings a certain component of trust to your business.”

David flags travel and hospitality as the vertical facing the most immediate disruption from AI search. Companies have already built booking tools around AI interfaces, and the brands that don’t show up in those recommendation flows risk becoming invisible to high-intent travelers.

But the risk extends beyond hospitality. Any industry with a high-consideration purchase cycle — SaaS, financial services, automotive, healthcare — is vulnerable if competitors start capturing AI recommendations first. The report shows that the gap between category leaders and average brands in AI visibility is already substantial, and it’s only widening.

For brands operating in these spaces, the question isn’t whether to invest in a GEO strategy — it’s how quickly you can get started.

‍

The Three-Step Strategy for AI Search

Asked what he’d recommend to a CMO with a strong SEO team looking to reallocate budget toward GEO, David outlined three priorities:

1. Align your content, digital marketing, comms, and PR strategies to a single North Star that reflects where the market is heading. The brands winning in AI search aren’t running separate playbooks for each channel — they’re building a unified narrative that AI systems can pull from anywhere.
2. Build out your prompt universe. Understand the questions, themes, and topics that drive growth for your business. Map the prompts your customers are using in AI tools and identify where your brand is — and isn’t — showing up.
3. Activate a GEO workflow to find compounding gaps and opportunities. Look at where your competitors are winning and go after it. Find low-hanging opportunities to dominate. Then feed that intelligence back into your content strategy across your website, blogs, PR, social media, product marketing, and schema/structure.

And, as Christian adds: don’t forget YouTube. The data shows it’s an increasingly powerful surface for AI discovery that most brands are overlooking.