Foundation
The problem isn't your website. It's what's underneath it.
You redid your branding. Redesigned the website. Rewrote the copy. Published the blog posts, optimised the images, and fixed the mobile layout. You did everything you were told to do. And it still doesn't rank.
The enquiries still don't come—your website is not converting despite the work you've put in. Your competitors—with outdated design, worse products, scrappy copy, and less experience—still appear above you in search. And now, when someone asks ChatGPT who to hire for what you do, your name doesn't come up at all.
Your first instinct is to burn it all down. A complete rebrand or pivot—or quit altogether. When the emotional wave is over, you end up doing more of the same. More content. A better headline. Another web builder. Another agency.
The problem with all of that is that you're addressing the wrong layer.
The part of your website you can't see
Every website has two layers.
The visible layer is what you and your visitors see: the design, the copy, the images, the navigation, the pages. This is where most website work happens—and where most website problems are diagnosed.
The invisible layer is the website infrastructure underneath: the schema markup that tells search engines and AI engines what your content means, the entity signals that tell them who you are, the internal linking architecture that tells them what matters, the crawl configuration that tells them what they're allowed to read, and the code structure that determines whether they can read it at all.
Human visitors never see the infrastructure layer. But the engines that determine whether those visitors find you in the first place see almost nothing else. They don't notice the third time you redesign your hero image. Or change your font. Frankly, it doesn't matter.
When a business has a website not working as it should—when the site doesn't rank, doesn't convert, doesn't appear in AI-generated answers—the problem is almost always in the invisible layer. Not the visible one.
This is the distinction that most website advice misses entirely. It focuses on the surface because the surface is what you can see. But patching the surface of a site with broken infrastructure is like repainting a house with a cracked foundation. It looks better. It doesn't fix anything, and it will still fall apart eventually; you just might not know it's happening at the time.
What infrastructure actually means
Infrastructure is a broad word. In the context of a website, it means five specific things.
Schema markup. Structured data—added to your site as JSON-LD code—that tells AI engines and search engines not just what your pages say but what they mean. Without schema, an engine reading your homepage sees words. With schema, it understands that you're a person, this is your business, these are the services you offer, this is where you're based, this is what you specialise in. Schema is the difference between being read and being understood.
Entity signals. A consistent, recognisable identity across your site, your schema, and anywhere else you appear online. Search engines and AI engines build confidence in a source by recognising it repeatedly across multiple signals—your name, your title, your location, your area of expertise, your links to and from credible external sources. Inconsistent or absent entity signals mean engines can't build a confident picture of who you are. And websites that answer engines aren't confident about don't get cited at all.
Crawl architecture. Whether the engines that need to read your site actually can. This includes your robots.txt file—which controls what crawlers are allowed to access—your sitemap.xml, your internal linking structure, and, critically, whether your site delivers readable content in the initial server response or buries it inside JavaScript that AI crawlers don't execute. A significant portion of websites are accidentally blocking the crawlers that matter most.
Content architecture. How your content is structured at a technical level—not whether it's well-written, but whether it's formatted in a way that allows AI engines to extract and cite specific answers from it. AI engines favour content that leads each section with a direct, self-contained answer. A site full of excellent, well-researched content structured for human reading will consistently lose citations to thinner content structured for machine extraction.
Site performance. Whether your site loads fast enough that visitors don't leave before they've read a word—and fast enough that search engines don't penalise you for making their users wait. Core Web Vitals are a ranking factor, and they're a function of infrastructure, not design.
These five things are entirely invisible to your visitors. None of them shows up in a screenshot. None of them is addressed by a redesign. And all of them have a direct, measurable effect on whether your site gets found.
Why this problem is getting worse
For twenty years, the stakes of getting infrastructure wrong were moderate; the playing field was fairly level. A site with missing schema and poor crawl architecture could still rank if the content was good enough and it had enough backlinks. The gap between a well-infrastructured site and a poorly-infrastructured one was real but bridgeable.
In 2026, the stakes have changed significantly.
Google has announced what it describes as the biggest change to Search in over 25 years. AI Overviews—synthesised answers generated by AI, delivered at the top of search results—now reach more than 2.5 billion monthly users. ChatGPT's search functionality reaches hundreds of millions more. Perplexity, Claude, and the growing category of AI assistants are fielding questions that would previously have gone to Google and delivering single synthesised answers with citations rather than a list of links to choose from.
In today's environment, infrastructure is no longer a technical nice-to-have. It's the prerequisite for visibility.
AI engines don't browse websites the way humans do. They crawl them, extract structured content, and build a model of what a source knows and who it is. They cite sources they can read clearly, understand confidently, and extract precise answers from. A site without schema, with blocked crawlers, with JavaScript-dependent content, with no entity signals—that site is structurally invisible to AI engines, regardless of how good the content is.
And the search behaviour shift is happening fast enough that a site that was adequately visible two years ago may now be significantly less discoverable than it was—not because anything changed on the site, but because the landscape around it changed entirely.
The redesign trap
The most common response to a site that isn't performing is a redesign.
It's understandable. A redesign is visible, tangible, and produces a result you can show people. You can point to the new homepage and say, “We fixed it.” It feels good. You look better, more up to date. Many small business owners are now using AI-assisted coding tools to build their own sites from scratch—genuinely excited by how quickly they can produce something that looks polished and professional.
But redesigns address the visible layer. They don't touch the infrastructure layer. And because the visible layer is rarely where the actual problem lives, most redesigns produce sites that look significantly better and perform almost identically.
The agencies and freelancers who run those redesigns aren't being dishonest. They're doing what they're hired to do. But most web design and development work is sold and scoped at the surface level—design, copy, UX—without a serious audit of the infrastructure underneath. The infrastructure either gets carried over from the old site (bringing all the original problems with it) or gets rebuilt from scratch without anyone having formally specified what it needed to do.
The result is a new coat of paint on the same foundation.
The alternative is to audit the infrastructure before deciding what needs to change—and to build whatever comes next with the infrastructure designed correctly from the start, not retrofitted at the end.
What fixing it actually looks like
Getting the infrastructure right is not a design project. It's not a content project. It's not an SEO project in the traditional sense. It's a systems, engineering and architecture project—and it has a specific, diagnosable scope.
A proper infrastructure fix involves:
Auditing what exists—crawl access, schema, entity signals, content architecture, internal linking, performance—against what AI engines and search engines actually need. Identifying the specific gaps and their relative priority.
Specifying what needs to change—in enough technical detail that the work can be executed precisely, by anyone. A schema graph. A crawl architecture specification. A content architecture with keyword mapping, answer-first structures, and FAQ blocks scoped per page.
Building it correctly—or handing a specification to whoever will. The infrastructure layer is not complex to implement when it's been correctly designed. Most of it is JSON-LD added to existing pages, configuration changes, and restructured content. It doesn't require a rebuild. It requires knowing exactly what to change.
The businesses that get this right—that invest in auditing and fixing the infrastructure layer—consistently outperform larger, better-funded competitors whose sites look better but are built on broken foundations. Because the engines that determine who gets found don't see the design. They see the infrastructure.
Where to start
If you recognise your site in any of this—good traffic, few enquiries; website not ranking despite your expertise; invisible in AI-generated answers; a redesign that didn't move the needle—the starting point is a proper diagnostic.
Not a generic site infrastructure audit from a tool that checks title tags. A structured review of the website infrastructure layer: schema, entity signals, crawl architecture, content structure, performance. A clear picture of what's broken, what it's costing you, and exactly what needs to change.
That's what the Infrastructure Audit is. A half or full-day—remote or in person in San Miguel de Allende—that produces a prioritised fix list you can act on immediately, in language you can hand to any developer.
Most engagements begin there.
Frequently asked questions
What is website infrastructure?
Website infrastructure is the technical layer underneath your visible content—the schema markup, entity signals, site architecture, internal linking structure, crawl configuration, and code foundation that determines whether search engines and AI engines can understand and cite your site. It's invisible to human visitors but determines almost everything about whether your site gets found.
Why is my website not ranking despite having good content?
Most commonly because the infrastructure underneath the content is broken or absent. Good content on a site with missing schema, poor internal linking, JavaScript-blocked crawlability, or no entity signals will consistently underperform content of equal or lesser quality on a technically sound site. Content is visible. Infrastructure is what makes it findable.
What is the difference between a website problem and an infrastructure problem?
A website problem is surface-level: the design looks dated, the copy isn't converting, the UX is confusing. An infrastructure problem sits underneath: the site can't be properly crawled, the schema is absent or wrong, the entity signals are inconsistent, AI engines can't extract and cite the content. Infrastructure problems are invisible to human visitors but highly visible to the engines that determine whether those visitors find you in the first place.
How do I know if I have a website infrastructure problem?
The clearest signals: your site gets traffic but not enquiries; you've redesigned or rewritten the site and rankings didn't move; competitors with worse content outrank you; you're invisible in AI-generated answers despite being an authority in your field. A site infrastructure audit—a structured review of schema, crawlability, entity signals, content architecture, and technical health—is the only reliable way to diagnose it.
What does a site infrastructure audit cover?
A site infrastructure audit reviews: how AI and search engine crawlers can access and read your content; whether schema markup (JSON-LD) is present, correct, and complete; entity signals—whether Google and AI engines have a confident, consistent picture of who the business is; content architecture—whether pages are structured to answer the questions your buyers are actually asking; internal linking—whether authority flows correctly through the site; and Core Web Vitals—whether the site loads fast enough not to be penalised.
If you recognise your site in any of this, we should probably talk. You don't need to have it figured out before we speak—that's what the Audit is for.
Get in touch →