SEO starts with URL architecture
Before writing metadata, every search intent should have one primary URL. Duplicate routes, uncontrolled parameters and thin pages split signals across multiple addresses.
App Router routes should reflect the real content model, with a self-referencing canonical and unique metadata for every indexable page.
Use the Metadata API as a system
Static routes can export metadata directly while dynamic routes should use generateMetadata from trusted content data.
- Unique titles
- Intent-specific descriptions
- Absolute canonicals
- Localized Open Graph
- Noindex for non-public pages
Reliable sitemaps and modification dates
Using the current date for every URL is not meaningful. lastModified should come from the actual content or record update date.
Language versions should be linked bidirectionally and only when they are genuine equivalents.
Structured data must match visible content
Use BlogPosting for articles, Service for service pages, BreadcrumbList for navigation paths and FAQPage only for questions visible on the page.
Structured data supports understanding; it does not replace useful content.
Validate after launch
Run a full crawl, inspect Search Console, validate rich results and review errors after release. Technical SEO must be rechecked whenever routes, CMS behavior or content models change.
Planning a serious digital product?
Share the current problem and expected outcome. We will suggest a practical architecture and phased delivery path.
Start a project