WordPress to Next.js URL Structure: Keep the Addresses Google Already Trusts

Count the addresses before you count the pages

Ask an owner how big their site is and they will say forty pages and a blog. Ask a crawler and it will find several hundred URLs. WordPress is generous with addresses. Every post can be reached by its permalink, by its numeric ID, through each category and tag it belongs to, through a date archive, an author archive and at least one feed. None of this is a problem while WordPress is answering. It becomes a problem the day something else has to.

Here are the URL families a typical WordPress site exposes, whether or not anyone meant it to:

• Posts and pages at whatever the Settings, Permalinks screen says, for example /sample-post/ or /2024/03/sample-post/.
• ID links such as /?p=123 for posts and /?page_id=45 for pages. With pretty permalinks on, WordPress redirects these to the real address, so old emails and forum posts still work.
• Taxonomy archives at /category/news/ and /tag/pricing/, or under a custom base set in the Optional section of the Permalinks screen.
• Date and author archives such as /2024/03/ and /author/jane/, which exist even on sites that never link to them.
• Pagination at /page/2/ on the blog and every archive, plus /comment-page-2/ on posts with paged comments.
• Feeds at /feed/, /comments/feed/ and a /feed/ on the end of nearly every archive.
• Attachment pages, one per uploaded file, which still work on sites created before WordPress 6.4.

The inventory is the first deliverable of any WordPress to Next.js development project we run, because every later decision depends on it.

Six permalink settings and the route folder each one becomes

Next.js routes are folders. A folder named in square brackets is a dynamic segment, so app/[slug]/page.tsx answers to any single-segment URL. That makes most WordPress permalink settings easy to reproduce exactly, which is the cheapest way to keep rankings. The table shows each setting from the WordPress Permalinks screen and the route that matches it.

Two rows deserve a closer look. With Post name permalinks, posts and pages live side by side at the root, so the single [slug] route has to look up both and decide which template to render. With %category% in the structure, WordPress writes the full category path, parent and child, into the URL. A post filed under News and then Local lives at /news/local/sample-post/. A catch-all route handles that, but it must use the category that WordPress actually put in the address, which is not always the one the editor thinks of as primary.

Keep, change or retire: one decision per URL family

The default answer is keep. A URL that already ranks has history, links and trust attached to it, and an identical URL in the new build inherits all of it with no redirect at all. Change is a choice you make for a reason, and retirement is for addresses that never earned anything.

Changing a date-based structure to clean slugs is the most common temptation. It is legitimate, and a permanent redirect tells Google which address is now canonical. It also means a redirect for every post and a period where Google recrawls the lot. If you want it, do it during the migration rather than six months after, so the site goes through one reshuffle, not two. The SEO-safe migration guide covers what to watch in Search Console while it settles.

Trailing slashes: one character on every address

WordPress pretty permalinks usually end with a slash, because the structure tag is written /%postname%/. Next.js does the opposite by default. Its documentation says it redirects /about/ to /about unless you set trailingSlash: true in next.config, which flips the rule. That one setting decides whether every URL on the site stays put or picks up a redirect.

Match the old site. If WordPress served /services/roof-repair/, set trailingSlash: true and the new build serves the same string. The canonical tag, the XML sitemap, the internal links and the redirect map then all have to agree on that form. A site that links to /pricing in the menu, lists /pricing/ in the sitemap and declares a third version as canonical is sending Google three opinions about one page, and one of them through a redirect. If an old site was inconsistent, pick one form, make it the rule in code and let the framework redirect the other.

The third option

Athena CMS, with AI agents on staff.

Not a theme and not a blank Next.js repo. The content system we put behind every custom site: plain fields, a real preview, every version kept, and AI agents that handle the chores a plugin only nags you about.

• PPicassoImages
• AApolloSite analysis
• AAresCompetitor watch
• PPrometheusContent holes
• AAtlasPage depth and structure
• DDelphiAI search visibility

ZDAOAand 12 more agents. Two come with Athena; the rest are monthly add-ons.

Meet the agentsTake the Athena tour

PPicassoWeekly · ImagesAApolloWeekly · Site analysis

Archives, page two and the feeds that still have readers

Category archives are real pages to Google, and on content sites they often rank for broad terms the individual posts cannot. Rebuild the ones that earn traffic at the same address, with the same base. If the old site renamed its category base to /topics/ in the Optional settings, the new routes use /topics/ too. Give each archive an introduction worth reading, because a list of excerpts is a thin page in any framework.

Pagination is simpler than it looks. A route such as app/category/[slug]/page/[n] can be generated at build time, so /category/news/page/3/ exists as a static file. Each page in the series should carry its own canonical tag rather than pointing back to page one, so the posts linked from deeper pages stay discoverable. The blog migration guide goes further into archives, authors and comments.

Feeds are the family people forget. Feed readers, podcast apps, email newsletter tools that send new posts automatically, and partner sites can all be polling /feed/ for years. In Next.js a route handler can return the RSS XML at exactly that path. Category feeds can be rebuilt or redirected to the main feed. The comments feed can usually go.

The canonical decisions WordPress was making behind your back

WordPress runs a function called redirect_canonical on most requests. It sends ?p= links to pretty permalinks, adds or removes trailing slashes to match your setting, and, when a URL would otherwise 404, it tries to guess what you meant by matching the slug against existing posts. A link to /roof-repai/ might quietly land on /roof-repair/. Your site has been rescuing broken inbound links for years and nobody told you.

A Next.js build does none of that unless someone writes it. That is mostly a good thing, since guessing is how sites end up with strange duplicates, but it means the rescues have to become explicit. Before launch, pull the inbound links and the old server logs, find the requests WordPress was correcting, and add the ones with value to the redirect map. After launch, Argus in Athena watches for broken links so the stragglers show up as a list, not as a slow slide in traffic.

Canonical tags themselves move into code. Next.js lets each page declare its canonical URL in its metadata, generated from the same route data that builds the page, so it cannot drift away from the real address the way a plugin setting can. If your old SEO plugin had per-post canonical overrides, export them: those were deliberate choices and they belong in the map too. The WordPress vs Next.js SEO comparison explains why metadata in code holds up better over time.

A URL inventory in six moves

1. 1

   Pull every known URL

   Combine the XML sitemaps from the SEO plugin, a full crawl of the live site, the Pages report in Search Console and the top linked pages from a backlink tool.

2. 2

   Add the hidden families

   List ?p= IDs, attachment pages, date and author archives, feeds and paginated archives, which sitemaps usually leave out.

3. 3

   Tag each URL with a decision

   Keep, redirect to a named destination, or retire. Note traffic and links beside each one so the decisions are defensible.

4. 4

   Write the routes to match

   Build Next.js route folders that reproduce the kept patterns exactly, including the trailing slash setting and any custom category base.

5. 5

   Generate canonicals and sitemaps from the same data

   One source of truth for every address means the sitemap, canonical tags and internal links cannot disagree.

6. 6

   Replay the list against staging

   Request every old URL on the preview build and confirm each returns 200 or a single permanent redirect to the right page.

Theme choice changes the inventory more than people expect. A Divi site often carries project and portfolio post types with their own bases, a Newspaper site tends to have deep category trees and heavy pagination, and a WoodMart store adds product, category and filter URLs that the WooCommerce guide maps separately.

Who guards the URL rules after launch

On WordPress, anyone with admin access can open Settings, Permalinks, pick a new structure and change every post address on the site in one click. It happens more often than you would think, usually during a well-meant tidy-up. In a Next.js build the route structure lives in code, reviewed and tested before deploy, so a change of that size is a decision rather than an accident.

Your team still controls what matters day to day. In Athena CMS a page or post slug is a plain field with a real preview, and every version is kept. Atlas flags orphan and thin pages, so archives that stop earning their place get noticed. If you would rather keep editing in WordPress and only replace the front end, the same URL rules apply, as described in keeping WordPress as the CMS behind Next.js. Either way, we keep your URLs and redirect every old one.