AI SEO vs. Traditional SEO: What I Am Actually Doing Differently in 2026

The debate about AI SEO versus traditional SEO is mostly a false dichotomy. The fundamentals of traditional SEO still apply. What has changed is the layer on top — the structural decisions that determine whether your content gets cited by AI systems versus just indexed by Google.

Here is what I kept, what I changed, and what I added when building SEO for four sites in 2026.

What I Kept

Page titles and H1 alignment. Your H1 should match (or closely mirror) your target keyword phrase. AI systems use heading structure as a primary signal for what a page is about. This has not changed.

Meta descriptions. Less important for click-through from Google AI Overviews (AI writes its own summary), but still critical for traditional SERP snippets and Open Graph previews on social.

Site speed and Core Web Vitals. Google's ranking systems still weight these. AI systems that pull from Google's index inherit these rankings. Vercel + Next.js with ISR handles this without extra effort.

Schema markup. BlogPosting, BreadcrumbList, FAQPage, HowTo — these remain the clearest signal you can send to any parsing system about what your content is.

Robots.txt access. If you block a bot, it cannot cite you. The key bots to allow in 2026: Googlebot, GPTBot, PerplexityBot, ClaudeBot, anthropic-ai. Blocking CCBot (Common Crawl, used for training data) while allowing the search bots is a reasonable balance.

What I Changed

Content structure. Traditional SEO rewarded comprehensive long-form content. AI SEO additionally rewards extractable passages — self-contained 40-to-60-word blocks that answer a specific question without requiring surrounding context. I now write with this in mind: every key claim should work as a standalone answer.

FAQ sections. I add a FAQ block to every post that targets a specific query. The questions are phrased as natural language searches ("How long does it take to..."). These sections get cited disproportionately in AI Overviews and Perplexity answers.

Heading phrasing. Traditional SEO: keyword in heading. AI SEO: phrase the heading as a question or as a statement that directly answers the likely search query. Not just "Benefits of Next.js" but "Why Next.js Outperforms WordPress for Content-Heavy Sites."

Authority signals per post. The Princeton GEO study found that adding statistics with citations (+37%), expert quotations (+30%), and authoritative tone (+25%) materially increased AI citation rates. I now include at least one cited statistic per post.

What I Added

sitemap.ts auto-generation. The sitemap generates from Supabase on every build cycle. Every published post is in the sitemap without manual work. AI crawlers parse sitemaps first.

Programmatic pages. A single template file generates pages for every city, every industry, every service combination. This is not spam — each page has unique templated content tailored to the combination. The SEO surface area multiplies without multiplying editorial effort.

llms.txt (under consideration). The emerging convention for AI context files — a plain-text summary of what your site is, who it is for, and links to key pages. Not yet supported by all AI systems but costs nothing to add.

Last-updated dates on all posts. AI systems weight freshness. Displaying and updating the dateModified field in schema markup and in the visible page signals that the content is current.

The Honest Assessment

The sites I built with this approach are three months old at the time of writing. I cannot yet report traffic numbers that prove causal impact. What I can say: all four sites are indexed and appearing in results for their target keywords faster than I expected, which suggests the structural choices are working.

The honest ceiling of AI SEO advice in 2026 is that the signals are noisier than traditional SEO metrics. Traditional SEO has 20 years of data. AI SEO has 18 months. Anyone claiming to have definitively cracked the formula is overstating their evidence.

What I am confident about: write for people, structure for extractability, cite your sources, keep content current, and do not block the crawlers. Everything else is iteration.

---

Resources

• —Google AI optimization guide — Google's official position on what matters for AI Overviews; essential reading
• —Princeton GEO study (KDD 2024) — the research measuring which content changes most improve AI citation rates; the source of the +37% statistics figure
• —llmstxt.org — the emerging convention for AI context files; a plain-text summary of your site for AI systems
• —Google Search Console — submit your sitemap here and monitor index coverage; the only official tool for tracking Google AI Overview presence

Related Posts

• —Auto Sitemap, Schema, and Robots.txt: The 3 Files Most Developers Forget — the technical implementation of the sitemap and schema layer described here
• —Programmatic SEO for a Local News Site: What Actually Moved the Needle — how programmatic pages applied AI SEO principles at scale
• —Case Study: Turning a WordPress News Site Into a Modern Platform in 48 Hours — the migration that made the AI SEO strategy possible

---

Want AI-optimized content and SEO infrastructure for your site? I build programmatic SEO systems, content pipelines, and schema layers as part of every project. See my services or get in touch.