The Regulator's Paradox
The UK Competition and Markets Authority just forced Google to lay out its ranking rules—or at least, try. The deadline is six months. But don’t cancel your next SEO audit yet. Transparency sounds like a win, but it doesn’t rewrite the fundamental equation: if your content isn’t legible to a machine, no amount of open-source reasoning will surface it.
This isn’t the first time regulators have peeked behind the curtain. It won’t be the last. And yet here’s what hasn’t changed in twenty years: search engines still need your content to be visible, structured, and trustworthy—before they can even begin to decide what to show.
The New Rules, Same Old Problems
On June 17, the CMA issued binding directions under Google’s Strategic Market Status designation. The demand sounds straightforward—rank organic results by "objective and non-discriminatory criteria," including inside AI Overviews—but the real headline is how they’ll verify it. Google must provide businesses with advance notice of ranking changes and a documented process to appeal.
That last bit is the first real recourse most web professionals have ever had. No more guessing whether an algorithm update dropped your traffic or a contractor broke something in production.
But the work ahead still looks like this: you need content that exists outside a JavaScript evaluation, a clear page outline, and factual consistency across your entire site. The regulator asking Google to explain itself doesn’t magically fix broken client-side rendering or buried key information.
You’d be surprised how many sites pass a visual inspection but fail a machine audit. The CMA order doesn’t change that.
Rendering: Is Your Content Actually There?
The simplest test for machine-readability is to disable JavaScript and reload your page. If what’s left feels like a skeleton without skin, you’re not in the clear.
Most AI crawlers—including Google’s own—handle JavaScript poorly. They run a limited subset, prioritize speed over completeness, and often skip dynamic content entirely. That means anything inside an API call, a Vue render cycle, or a React hydration phase becomes invisible the moment an AI tries to synthesize an answer.
The fix isn’t fancy. It’s fundamental:
- Render core content in the initial HTML response.
- Use progressive enhancement so that early visitors get value before scripts finish loading.
- Audit your critical pages with the developer tools’ network tab filtered to document requests. If key content never appears in a raw HTML fetch, it doesn’t count.
Structure: Let the Machine Find Your Answers
A machine doesn’t read a page like you do. It scans for discrete claims, parses headings, and tries to extract reusable units of information.
Your job isn’t to make your content look pretty to humans—though that matters. Your job is to ensure those beautiful ideas can be lifted cleanly without losing meaning.
That means:
- Headings that signal distinct topics, not poetic flourishes.
- Paragraphs around one idea each, with topic sentences up front.
- Lists where appropriate, since machines love bullet points for citation extraction.
If your answers only resolve after reading the entire page top to bottom, an AI will skip you for sites that front-load their substance. This is especially true inside AI Overviews, where brevity wins and context is thin.
Verifiability: Build Trust Through Repetition
LLMs don’t trust a single assertion in a vacuum. They look for corroborating facts elsewhere on your site—or better yet, across trusted third-party sources.
That’s why a page that says "We’re the best in class for X" without naming competitors, evidence, or context rarely wins. It sounds like marketing boilerplate.
Machine-first sites repeat the same facts in different forms: FAQs, service pages, case studies, and even team bios should echo core claims consistently. This redundancy isn’t redundancy in the human sense; it’s signal reinforcement for an AI trying to decide whether your page is the right answer at all.
When Google or another engine eventually publishes its scoring rubric, verifiability will still win out over keyword stuffing because trust matters more than noise.
Stop Waiting for the Black Box to Open
Transparency would be nice. It would settle arguments, reduce confusion, and give sites a fair shot at appearing in AI answers.
But waiting for that moment to arrive is like waiting for rain during a drought—you’re handing control of your content strategy to someone else’s timeline.
The right move is to audit what a machine already sees on your site. Disable JavaScript, skim for headings, and check whether your core claims hold up across multiple pages.
Do that now. Not when the regulator’s deadline looms, not after a competitor shows up in an AI Overview and you have no idea why.
Because the machine doesn’t care how opaque or transparent the ranking system is. It only cares whether your content was legible in the first place.