The Technical Stack Behind SEO, AEO & GEO, Part 1: Sitemaps, JavaScript & CSS

Most conversations about search visibility start with content: keywords, headlines, backlinks, the things a marketer can point to and say "this is the strategy." But underneath every one of those strategies sits a much less glamorous layer — the technical plumbing that decides whether a search engine, an AI crawler, or a language model can actually see, parse, and trust your site in the first place.

That layer is made up of things most business owners never think about: your sitemap, how your JavaScript and CSS are structured, what your robots.txt file allows and blocks, whether your pages carry schema markup, whether you have an llms.txt file at all, and — increasingly relevant, and increasingly misunderstood — how your cookie and consent setup behaves when a bot, not a human, loads the page. This is Part 1 of a three-part breakdown of that layer. Here, we cover the sitemap, JavaScript, and CSS — the foundation that determines whether a page can be reached and rendered at all. Part 2 covers robots.txt, schema markup, and llms.txt. Part 3 covers cookies, and pulls all six pieces together into a single checklist.

Get this layer wrong, and it doesn't matter how good your content is. A search engine can't rank a page it can't crawl. An AI answer engine can't cite a fact it can't extract. A generative engine can't recommend a business whose technical foundation actively hides information from it. This is the difference between SEO (getting ranked by traditional search), AEO (getting quoted by answer engines like ChatGPT, Gemini, and Google AI Overviews), and GEO (Generative Engine Optimization — being surfaced, summarized, and recommended inside AI-generated responses more broadly) all resting on the same unglamorous foundation.

Three Different Games, One Shared Foundation

SEO, AEO, and GEO are often described as separate disciplines, and in terms of what "success" looks like, they are. SEO success is a ranking position. AEO success is being the source an AI quotes verbatim. GEO success is being woven into a generated answer, a recommendation list, or an AI shopping assistant's suggestion — sometimes without a click or a citation at all.

But all three depend on the same prerequisite: something has to be able to access your content before it can rank it, quote it, or synthesize it. That access is governed entirely by the technical layer. A content strategy without a technical foundation is a beautifully written letter with no address on the envelope.

The Sitemap: How Discovery Actually Starts

A sitemap is the most literal expression of "here is what exists on my site." It's a list of URLs, handed directly to search engines, that says: crawl these, here's when each one last changed, here's roughly how important each one is relative to the others.

An automated, self-updating sitemap — one that reads each page's real last-modified date off the file system rather than relying on someone remembering to update an XML file by hand — has a compounding effect over time:

• Faster discovery of new and updated content. The moment a page changes, its lastmod date changes with it. Search engines use that signal to prioritize re-crawling, so updates get reflected in the index faster instead of sitting stale for weeks.
• No orphaned or missing URLs. A hand-maintained sitemap drifts out of sync with the real site — pages get added, forgotten, or removed, and the sitemap doesn't notice. An automated one is generated from the same source of truth as the site itself, so what's listed and what exists never diverge.
• Multilingual and multi-template scaling without manual duplication. On a site with several language versions of every page, a sitemap driven by a shared slug list means adding one new page automatically produces five correct sitemap entries — one per language — instead of five chances to make a typo or forget one.
• Crawl budget efficiency. Search engines allocate a finite amount of crawling attention to any given site. A clean, accurate sitemap tells crawlers exactly where the real content lives, so that budget goes toward pages worth indexing instead of being spent rediscovering the same structure over and over.

For AEO and GEO specifically, the sitemap matters even more than it used to, because AI crawlers (GPTBot, ClaudeBot, PerplexityBot, Google-Extended, and others) are now crawling sites independently of traditional search indexing. A sitemap that's accurate and current is one of the fastest ways to get new content in front of those systems without waiting for an organic re-crawl.

JavaScript: The Silent SEO, AEO, and GEO Killer

Modern websites lean heavily on JavaScript for interactivity, but JavaScript is where a huge share of invisible SEO and AEO failures happen. The core problem: many crawlers — and nearly all AI crawlers — either don't execute JavaScript at all, or execute it with limited patience and limited budget. If your key content, headings, or answers only appear after a script runs, a large share of the systems trying to read your page may never see that content in the first place.

This matters differently across the three disciplines:

• SEO impact: Google can render JavaScript, but it does so as a second-pass process, on a delay, and with a finite rendering budget. Heavy, render-dependent pages get indexed slower and less reliably than pages with content present in the initial HTML.
• AEO impact: Answer engines need to extract a precise, quotable fact. If that fact is injected client-side after page load, many AI crawlers will simply never encounter it — meaning content that's perfectly visible to a human visitor is functionally invisible to the system trying to cite it.
• GEO impact: Generative engines building a broader understanding of "what this business offers" or "what this page is about" are often working from a lightweight, JavaScript-free fetch of the page. If your value proposition lives inside a JS-rendered component, it may be entirely absent from the version of your page these systems actually process.

The practical fix isn't "avoid JavaScript" — it's making sure the content that matters for search, answers, and AI summarization exists in the server-rendered HTML, with JavaScript layered on top for interactivity rather than as the delivery mechanism for the substance of the page.

CSS: More Than Looks — Core Web Vitals and Crawl Efficiency

CSS rarely gets blamed for search visibility problems, but it plays two distinct roles that touch all three disciplines. First, render-blocking CSS slows down how quickly a page becomes usable, which directly affects Core Web Vitals — a ranking factor for SEO and a trust signal that feeds into how favorably AI systems weigh a source when multiple candidates are otherwise similar.

Second, and less obvious: bloated, unoptimized CSS and asset delivery slows down every crawl, human or automated. A crawler working through a large site with a finite time budget per page will get through fewer pages, less often, if every page is heavier than it needs to be. Multiply that across a multilingual site with hundreds of URLs, and unoptimized CSS quietly caps how much of your site ever gets seen at all.

Practical priorities: minimize render-blocking stylesheets, load critical CSS inline for above-the-fold content, defer the rest, and cache aggressively so repeat crawls (which happen constantly) don't re-fetch the same bytes unnecessarily.

Coming Up in Part 2

Sitemaps, JavaScript, and CSS get a system to your content and let it render correctly. They don't tell that system what any of it means, or whether it's even allowed to use it. Part 2 covers robots.txt — and why it now has to think about AI crawlers specifically — schema markup, and llms.txt, the newest file in this stack, built for how language models orient themselves around a site.