Published on December 18, 2025

SEO is Now a Retrieval Game

By Ben Murphy

If your content only works as a long-form read, it will struggle in AI-driven search.

You will often hear Google say that optimising for AI-driven search is “just SEO”.

That statement is technically true, but dangerously incomplete.

AI search is no longer deciding who ranks. It is deciding who is credible enough to be used as a source.

Current Update Alert: December 2025 Core Update

As of mid December 2025, Google is in the middle of a core update rollout that is aggressively filtering content quality.

We are seeing sites drop visibility, not because they broke rules, but because their content hits a word count without actually satisfying intent.

Pages built with unedited AI content, generic explanations, or no firsthand proof are being quietly removed from prominence. This is affecting both traditional rankings and AI driven visibility at the same time.

This update is not about punishment. It is about credibility.

The Part Google Doesn’t Say Out Loud

Traditional SEO was about pages.
AI-driven search is about passages.

In classic search, Google evaluated an entire page and decided where it should rank. In AI Overviews and generative results, Google retrieves small, self-contained chunks of information and assembles answers from them.

You can rank number one organically and still be invisible in AI results if your content cannot be easily extracted, understood, and trusted at a passage level.

This is why many strong-looking websites are losing visibility during this update without obvious technical issues.

Ranking Versus Citation: The Shift at a Glance

Traditional SEO and AI SEO solve different problems.

Traditional SEO aims to win a click.
AI SEO aims to be the source.

Here is the difference in practical terms.

Feature
Traditional SEO
AI SEO and GEO

Primary goal
Occupy a blue link position
Be the source of truth for the answer

Content unit
The entire page or URL
Self-contained passages or chunks

Main metric
Clicks and impressions
Share of model and citation rate

Format focus
Narrative flow
Direct answers and atomic retrieval

If your content only works as a long-form read, it will struggle in AI-driven search.

What AI Search Is Actually Filtering For

AI Overviews act like a source filter.

Generic advice does not survive this filter.
Vague service descriptions do not survive this filter.
Content that buries the answer does not survive this filter.

This is why Reddit threads, forum answers, and YouTube transcripts are frequently cited. They offer direct, experience based statements that AI systems can reuse confidently.

The Shift to GEO, Generative Engine Optimisation

Good SEO now requires a dual-track approach.

You still write for humans, but you must also structure content for retrieval.

At PunkFox, GEO means three practical changes.

Passage First Content Structure

AI does not read like a person. It retrieves answers.

Every major section of a page should begin with a direct answer of roughly forty to fifty words. That answer must stand on its own.

If the AI cannot extract a complete response from the opening sentences of a section, it will move on to a competitor who made the job easier.

This is one of the biggest visibility leaks we see during the current update.

Machine Readable Authority Signals

Headings are no longer just for readability. They act as intent markers.

Clear H2s and H3s tell AI systems what problem is being solved in each section. Combined with consistent language and semantic clarity, this dramatically improves retrieval.

Schema also matters again.

Organisation schema, Person schema, and FAQ schema help remove ambiguity around who you are, what you do, and why you are qualified to speak on a topic.

Authority is no longer implied. It is declared and verified.

From Keywords to Share of Model

Traditional SEO reporting focused on keywords and clicks.

AI-driven search introduces a new reality. Visibility without clicks.

That can sound uncomfortable until you understand where the conversion actually happens.

When an AI answer repeatedly references your brand as the expert, users do not click immediately. They remember you.

They then search for you by name.

This is where branded search grows, trust accelerates, and conversions occur. AI SEO is not replacing performance. It is becoming the top of the funnel that feeds it.

If the AI never mentions you, you are invisible, even if you technically rank well.

So Is Google Right When It Says “It’s Just SEO”?

Only if your definition of SEO has evolved.

If SEO means clarity, structure, proof, and authority, then yes. This is still SEO.

If SEO means chasing rankings with padded pages and hoping traffic follows, that version is being filtered out right now.

The December update is making that very clear.

AI search is not changing the rules. It is enforcing them with less tolerance than ever before.

The businesses that win are not chasing AI tricks. They are making it effortless for both humans and machines to understand exactly what they do and why they are credible.

That is where modern SEO now lives.

The natural next step is not more content. It is auditing what you already have to see whether it is built to rank, to be cited, or both.

A simple AI-ready content audit can reveal gaps in minutes that traditional SEO reviews often miss.

Ben Murphy

About The Author

Ben Murphy - Founder

Ben Murphy is an SEO specialist with over 15 years of hands-on experience helping businesses grow through transparent, data-driven search strategies, having launched and scaled one of Manchester’s leading SEO agencies before relocating to Perth in 2025 to bring his proven methodology to the Australian market. Known for long-term client retention, measurable results, and a partnership-first approach, Ben now leads PunkFox with a focus on delivering senior-level expertise, honest guidance, and sustainable organic growth for brands across Perth and beyond.