Engine Room Reflections: SEO in the AI Era
By the Cat Media Team – cross-functional, colour-coded, and perpetually curious
Engine Room Reflections is our ongoing editorial series where the Cat Media team shares ideas, insights, and forward-looking commentary across marketing, RevOps, customer experience, and technology. Each article reflects our cross-functional perspective, grounded in real client work, driven by data, and built inside HubSpot.
In this edition, we explore how SEO is evolving in the era of AI-driven discovery — and what teams need to change to remain visible in a world of single-answer search.
From Search Engines to Answer Engines
Earlier this year, LinkedIn and X lit up with headlines:
“HubSpot lost 80% of its blog traffic.”
“SEO is over.”
“Google is no longer the gateway.”
It was sensational — and naturally, attention-grabbing. But within our team, we took a different approach. We examined the claims, looked at the data, and discussed what this means for the future of marketing, sales, service, and operations.
Because SEO hasn’t vanished, it’s shifting.
Today’s discovery channels look different. Platforms like ChatGPT, Copilot, Perplexity, and other AI-driven interfaces are shaping how people ask questions and receive recommendations. They don’t list 10 options. They deliver one answer.
And being included in that answer depends on more than keywords.
“If you’re not part of the LLM’s knowledge base, you don’t exist.”
That quote captured our attention in a recent thread, and it neatly sums up what’s changing. (Source: @andreessenhorowitz on X, but also a dozen others in the AI-first SEO space).
Search used to reward backlinks, keywords, and DA. But LLMs learn through associations, patterns, and trust signals from a wide array of sources: Product Hunt, GitHub, Twitter, Reddit, forums, support docs, and even internal copy.
To be discoverable now means to be present in the right data loops — visible where the model learns, not just where people search.
Our Perspective: Four Colours, One Table
We’re a cross-functional team, each bringing a unique perspective rooted in our roles and personalities. Inspired by Surrounded by Idiots by Thomas Erikson (super recommended), we embrace our colour codes to reflect how we show up in the work.
🔴 Sales (Red – direct, results-driven, focused on outcomes)
AI-powered search shortens the decision cycle. When someone asks Copilot, “What CRM works best for B2B teams?” and your brand appears, that’s not a browser window — it’s a potential buyer, ready to act.
We're guiding clients to document and track these moments — attribution doesn’t just come from forms anymore. It comes from being mentioned as the default choice in an answer.
This is where deals are won — before your competitors realise the question was even asked.
🟡 Marketing (Yellow – creative, optimistic, future-focused)
This shift is full of possibility. We’re moving away from endless keyword-stuffed content and focusing instead on clarity, connection, and consistency.
We’re creating structured, answer-first content designed to train LLMs, not trick algorithms. Our focus is now on Product Hunt launches, social posts, FAQs, and docs — all places where LLMs pick up on associations. If your tool is for “media automation for developers,” that phrase should appear clearly and in all the right places across platforms.
Creativity thrives when it’s rooted in clarity.
🟢 Customer Success (Green – thoughtful, empathetic, people-first)
The search journey doesn’t end at the homepage. Customers turn to AI to troubleshoot, explore features, and validate choices, long after they’ve signed up.
We’re building support experiences that are both scalable and specific. Our onboarding programmes are delivered by RevOps specialists trained in the HubSpot Knowledge Base and Academy and RevOps frameworks and strategies, ensuring expert guidance from day one.
We’re also launching the Custom Expert, a conversational AI assistant trained on each client’s portal, documentation, and processes. It supports customers, internal teams, and new hires alike, offering fast, accurate answers in your business language.
Support becomes proactive, continuous, and embedded into the day-to-day.
🔵 RevOps (Blue – analytical, detail-oriented, structure-focused)
In operations, structure isn’t optional — it’s essential.
We’ve built our internal platform to support customers, furball, to automate the data hygiene that underpins everything else, and we're soon launching self-service. This includes:
-
Contact and company deduplication
-
Property standardisation and formatting
-
Multi-language classification and job title clean-up
-
AI-powered enrichment and inference
This is about data accuracy and making your portal AI-ready.
We’re also launching the RevOps Expert, a dedicated AI assistant trained on your specific HubSpot portal, naming conventions, workflows, and processes. Designed for admins, ops managers, and analysts, it serves as an operational co-pilot, answering questions, supporting workflow design, and reducing the load on technical teams. We’re also developing a Dev Expert, focused on supporting technical teams working with custom integrations, programmable automation, and advanced data flows inside HubSpot.
All part of the furball app suite, built for scale, speed, and strategic clarity.
The HubSpot View: What Actually Happened?
HubSpot addressed the SEO conversation directly in their blog post on traffic decline. While traffic did shift, the narrative around decline lacked context.
“The way people search is evolving, and that changes how we think about content. But trust, clarity, and usefulness are still the foundations.”
We agree. SEO isn't gone. It’s simply adapting to a different type of intelligence.
Their guide to SEO across the customer journey is also a valuable reminder: intent varies by stage, and AI search tools are now embedded in every stage, from consideration to onboarding.
What We’re Building Toward
We’re rethinking what discoverability really means. Our content, data, and infrastructure strategies are now tied to a single idea:
Show up in the answer.
This is what we’re doing:
-
Publishing answer-first content across key LLM-influencing platforms
-
Clarifying entity associations everywhere (website, docs, community, social)
-
Structuring every sentence, page, and property for clarity
-
Connecting our CRM and content workflows through HubSpot
-
Building expert systems — both human and AI — to reinforce that clarity
We’re not here to chase traffic. We’re here to build presence, trust, and signal.
We’ll keep testing, refining, and sharing what we learn. Until then, we’ll be in the engine room.
— The Cat Media Team
Frequently Asked Questions (SEO in the AI Era)
What is AI-powered SEO?
AI-powered SEO focuses on making content discoverable by large language models (LLMs) like ChatGPT, Copilot, and Perplexity. Rather than optimising for keyword rankings in Google, it prioritises clear language, structured data, and entity relationships across platforms and content types.
Start by grading your AI Search with HubSpot.
How is SEO changing in 2025?
Search behaviour is shifting from traditional search engines to conversational interfaces. AI tools increasingly recommend single answers based on confidence and clarity, rather than showing ranked lists of links. This places greater value on structured, answer-first content and cross-platform consistency. Read more.
What is entity SEO?
Entity SEO refers to the practice of helping AI models connect your brand, product, or service with a specific concept or function. Instead of chasing individual keywords, you train the model to associate your brand with a particular solution through consistent messaging across web, docs, social, and communities.
How can I make my content visible to LLMs?
To show up in AI-generated answers, focus on:
-
Writing concise, structured, answer-first content
-
Repeating your product’s purpose clearly across channels
-
Participating in platforms that train LLMs (Reddit, GitHub, LinkedIn, etc.)
-
Optimising internal docs and support content, not just blogs
What is furball?
Furball is Cat Media’s internal app suite for HubSpot users. It includes tools for deduplication, data formatting, multi-language classification, and AI assistants like the Custom Expert, RevOps Expert, and Dev Expert — all designed to support RevOps and front-line teams. Join the waitlist.
Share your Thoughts or Ask Us Anything!
.png)