The vibe coder's SEO checklist

Most AI page builders are excellent at the part of SEO that's visible — headings, readable copy, a clean layout. They're much less consistent about the part that's invisible: the metadata, markup, and crawl signals that never show up in the preview, but are exactly what Google checks first.

The dozen things worth checking before launch

• A unique <title> and meta description on every page, not one copied from the template.
• Open Graph and Twitter Card tags, so links don't render as a bare URL when shared.
• A single canonical URL per page — duplicate routes for the same content confuse crawlers.
• An actual sitemap.xml, kept up to date as pages are added.
• A robots.txt that isn't accidentally blocking the whole site (it happens more than you'd think).
• One <h1> per page, in a logical heading order underneath it.
• Descriptive alt text on images that carry meaning, not filenames.
• Internal links between related pages — crawlers follow links to find and rank content.
• A custom 404 page, so broken links don't dead-end into a blank error.
• HTTPS everywhere, with no mixed-content warnings.
• Fast enough Core Web Vitals — see the previous post on why Lighthouse alone won't tell you.
• Structured data (schema.org) where relevant — product, article, or FAQ markup.

The three items that trip people up

Canonical URLs matter most on sites that unintentionally serve the same content at two addresses — /pricing and /pricing/, or a marketing page reachable both with and without a trailing query string. Without a canonical tag pointing to one true version, a crawler can split ranking signal between duplicates instead of consolidating it on one page.

Structured data is the one builders skip most, because nothing visually changes when it's missing — the page looks identical with or without it. But it's the difference between a search engine guessing your page is probably about pricing, and a search engine knowing it's a Product with a specific price, currency, and availability, which is what unlocks rich results like star ratings or price snippets in search.

A custom 404 gets treated as cosmetic, but it has a real crawl-budget cost: a generic error page with no navigation is a dead end that wastes a crawler's visit and gives a lost human visitor nothing to click. A 404 with a search box and links back to key pages turns a bounce into a second chance.

Quick wins vs. the slow ones

• Minutes, not hours: title tags, meta descriptions, alt text, canonical tags, custom 404 — mostly copy changes.
• An afternoon: sitemap.xml generation, robots.txt review, Open Graph tags, internal linking pass.
• Ongoing: Core Web Vitals and structured data are never really "done" — they need re-checking as pages and content change.

Why this list gets skipped

None of it is hard. All of it is invisible in the builder's preview pane, so it's easy to ship a page that looks finished and is missing half the list. The fix for each item takes minutes; finding out which ones you're missing is the slow part.

Run an SEO scan and you get this exact checklist back per-page, ranked by how much it's likely to matter — rather than working through all twelve items on every page by hand.