If you care about content performance in 2025, what happened at SEO Week matters.

Sure, there were big names, bold slides, and tactical tips. But it was also a clear line in the sand: the future of SEO is no longer about those little blue links on page one, but about language models, influence, and system design.

There was no shortage of memorable moments:

• Aleyda Solis debuted new AI tools for refreshing PDPs—showing exactly how to make old content competitive again.
• Mike King dropped bars and insights, rapping on stage after a panel on retrieval architecture.
• And yes, AirOps kept the crowd caffeinated with a coffee booth that sparked more than a few LLM debates before 9 a.m.

But beneath the flash was substance. The smartest teams are rethinking how they align to LLMs, shift KPIs away from traffic alone, to winning on brand and with great content.

This is your cheat sheet for what’s next.

We'll get into the 7 biggest shifts, and how your team can stay ahead. But first the really good stuff: 

Aleyda Solis showed off AirOps as a new wave of tools to enhance and enrich content

A big afterparty with Busta Rhymes and iPullRank’s Mike King showing off: 

And AirOps serving up a great cappucino

Okay, now let’s dive into the real takeaways to improve your SEO:

1. SEO isn’t a checklist anymore. It’s a system.

Modern SEO is about building a system that reflects how search engines and language models interpret and retrieve information. That means proving relevance through structure, semantics, and context—not just keywords and links. Mike King, the emcee and host of the event from iPullRank, introduced the concept of relevance engineering. He encouraged SEO professionals to think more like systems architects. His framework focused on content chunking, semantic triples, and vector-based content relationships to better align with retrieval models like RAG.

Jeff Coyle from Site Improve shared an AI Editorial Excellence Checklist that outlined what it really takes to create content that performs across both human and AI audiences. It includes elements like developmental editing, semantic analysis, fact-checking, and AI-human collaboration. His emphasis was on rigorous editorial standards, not volume.

2. LLM visibility is the new SEO metric

It is no longer enough to rank in Google. To be discoverable, your brand needs to show up in the tools your audience uses to get answers, including ChatGPT, Perplexity, Gemini, Claude, and DeepSeek.

“We must analyze how LLMs associate entities with our brand and influence them like we do Google.” — Dan Petrovic

Dan introduced a probing method to measure brand presence inside language models. His approach examined both how a brand is linked to key entities, and how often those entities return the brand in association. This creates a feedback loop that influences which sources LLMs retrieve, reference, and recommend.

JR Oakes shared benchmark data showing that LLMs increasingly surface mid-tier or niche brands—not just the ones with the most traffic. He found that strong performance in traditional search often correlates with stronger LLM visibility, especially when pages are well-structured and linked.

“Having a better performing site for traditional search potentially leads to better visibility outcomes for LLMs as well.” — JR Oakes

Crystal Carter emphasized that tools like DeepSeek reward expert-driven, timely content. She explained that DeepSeek cites more sources per query than ChatGPT or Perplexity and favors pages that support “chain of thought” reasoning—especially those with clear authority, structured facts, and multimedia elements.

What helps boost LLM visibility:

• Structured, fact-rich pages (FAQs, About, expert bios)
  
  
• Multimedia content and UGC
  
  
• Content that answers “why” and “how,” not just “what”
  
  
• Strong traditional SEO signals: schema, links, internal structure

3. AI Overviews are lowering clicks. Your influence matters more than your ranking.

AI Overviews are cannibalizing organic traffic. Multiple speakers shared data showing that even position one is no longer a guarantee of visibility.

“The future does not look good for those reliant on organic traffic.” — Rand Fishkin

Rand Fishkin from SparkToro made the case for influence-first content strategies. Rather than trying to win the click, you need to win the recommendation. This means optimizing for brand citations, offsite mentions, and structured content that gets quoted or summarized by AI systems.

Here’s a slide from his presentation: 

Nick Eubanks added that these Overviews are directly responsible for reducing SEO-attributed traffic. He warned that even highly visible rankings are no longer reliable performance indicators. 

4. Reporting needs to connect inputs to ROI

Executives do not care about rankings or traffic. They want to understand where resources are going and what impact those efforts will create. That requires better reporting and new metrics.

“What gets you promoted is reporting that answers: where did my money go?” — Tom Critchlow

Tom Critchlow, the EVP of Growth from Raptive, outlined a shift away from lagging indicators like traffic and toward operational input metrics. These include things like schema coverage, freshness audits, expert-reviewed content, and internal link density. These metrics give teams better visibility into the effectiveness of their SEO efforts before outcomes hit dashboards.

He also recommended reporting with narrative. Dashboards alone do not tell the story. Content leaders should aim to explain what was worked on, why it matters, and how it connects to business goals. These are the kinds of reports that secure more resources and more trust.

5. Search is about journeys, not keywords

Google is no longer just answering questions—it’s modeling entire user journeys. Showing up for one keyword is not enough. You need to appear across the full arc of discovery.

Cindy Krum of MobileMoxie explained how Google’s MUM model powers this shift. MUM connects search behaviors—clicks, filters, follow-up queries—into cohesive journeys. Google uses a technique called query fan-out, triggering multiple related searches in the background to anticipate what users will want next. This allows Google to serve more results, more ads, and more AI Overviews with less processing cost.

“Filters equal journeys equal conversations.” — Cindy Krum

‍

In practice, this means that ranking for a keyword is only one part of the equation. To stay visible, your content must align with Google’s understanding of what comes next. Fajr Muhammad reframed the funnel as a flywheel, where awareness content fuels both retention and referrals, especially when it’s modular and interconnected.

The best-performing teams today are:

• Mapping intent across People Also Ask, related searches, Reddit threads, and journey-based filters
  
  
• Creating multi-modal clusters around each stage of discovery
  
  
• Building for pre-search and post-click experiences that guide users and AI systems alike
  
  

If your content supports the journey, Google is more likely to include you in the results—whether that’s a link, a snippet, or an AI-generated summary.

6. Off-site is on-brand

Clicks are fading in importance. What matters now is attention, interaction, and shareability.

“Win on Reddit or YouTube and you’ll win in search.” — Ross Simmonds

Ross Simmonds explained that platforms like Reddit, YouTube, and TikTok are driving discovery. Content that performs well there often shows up in search and LLMs faster than blog posts.

Modern engagement looks like:

• Content that sparks discussion
  
  
• Interactive tools and explainers
  
  
• Assets built for SEO and social sharing

Your best-performing content might not live on your site. It needs to earn distribution to earn visibility.

7. PR, brand, and SEO are merging and the machines are watching

Your About page, team bios, and FAQs are no longer just branding tools. They are training data. If you want LLMs to understand and recommend your brand, you need to teach them.

“LLMs pull from About pages, FAQs, and PR content. These aren’t soft brand pieces anymore—they’re your first impression to the models.” — Lexi Mills

Lexi Mills of Shift6 explained how LLMs are trained on structured editorial content. That includes company descriptions, expert quotes, and brand-owned metadata. She also stressed the importance of clear, structured, and trusted sources that can get cited inside GenAI outputs.

James Cadwallader from Profound introduced “Agent Experience” (AX) as a new lens for understanding how LLMs interact with and surface content.

Instead of focusing only on ranking in traditional search engines, James urged SEO leaders to optimize for how AI agents retrieve, evaluate, and present brand information. He described this as an evolution of UX not for users, but for agents.

The teams that are winning in this space are those who treat brand as infrastructure. They structure bios with credentials, publish expert content on core topics, and maintain a public knowledge graph that LLMs can interpret and retrieve from.

How AirOps can help your team with SEO

At AirOps, we help content and SEO leaders build systems that work for modern search and AI-powered discovery. Our platform gives teams the tools to:

• Create structured workflows that match how search and LLMs evaluate content
• Build content with semantic clarity, entity alignment, and high editorial standards
• Automate repeatable tasks with AI workflows to enhance your productivity

If you are rethinking your content stack for 2025, AirOps can help you scale smarter.

👉 Schedule a call to see how AirOps supports AI-native content teams.