Sitemaps still matter, for a boring reason

Link-following has blind spots

Crawlers find pages by following links. Anything reachable only through a search box, deep pagination, or a filter combination may never be reached at all.

A sitemap is the direct answer: here is the list, no traversal needed.

It feeds your llms.txt too

A sitemap is the natural input for generating an llms.txt — the URL set, grouped by path prefix, is most of the structure you need.

Sites without a sitemap tend to end up with a thin llms.txt, because there is nothing to build the page list from.

Generating one

In Next.js App Router this is a file, not a build step:

// app/sitemap.ts
import type { MetadataRoute } from "next";

export default function sitemap(): MetadataRoute.Sitemap {
  const now = new Date();
  return [
    { url: "https://acme.com", lastModified: now, priority: 1 },
    { url: "https://acme.com/pricing", lastModified: now, priority: 0.9 },
  ];
}