Sites

Your owned-domain index. Sites is where you watch what Lumear has crawled, kick off re-crawls, and triage pages that AI keeps citing back to you.

What “Sites” actually means

Every domain you add to a brand becomes a row on the Sites page. Each domain has a list of pages we’ve crawled, with their last-seen HTTP status, page-type classification (homepage / blog / product / pricing / docs / etc.), word count, and crawl timestamp.

The page count and the per-page page-type breakdown are what feed the recommendation engine. A thin or stale index will produce thin recommendations.

Filtering pages

Search — substring match against URL and title.
Domain — narrow to one brand domain (only relevant when a brand has multiple).
Page type — filter to blog posts, product pages, comparisons, etc. Useful for “show me every comparison page on the site”.
Section — filter by the first URL path segment (e.g. /blog,/products, /investors). Best for “everything under the Resources tree”.

Crawl scopes

Each domain has three crawl modes; pick based on how recently the site changed:

Topical (default) — Firecrawl /map + a filtered batch scrape. Lumear picks the pages most likely to match the brand’s prompt sets. Fast and cheap; recommended for re-crawls.
Cited only — re-fetch only the URLs AI assistants have actually cited in recent runs. Tiny scope, near-real-time freshness. Good for “a single page changed and I want it re-indexed.”
Full audit — crawl every page Firecrawl can find up to the cap. Expensive; reserve for first crawls and post-redesign refreshes.

First-time crawl is automatic

When you add a domain to a brand, Lumear immediately fires a topical scope crawl in the background. You don’t need to click “Crawl” manually unless you want a different scope or a re-run.

Per-page status

Each row carries an HTTP status badge. 2xx means the page was indexed successfully; 3xx means it redirected (we follow); 4xxmeans broken (Lumear ignores the URL); 5xx usually means the origin was flaky and we’ll retry on the next crawl.

Using Sites to debug recommendations

If your Recommendations feel off-target, the Sites page is usually the first place to look:

Filter by section to see how thin a key area is. A “build a /comparison/ page” recommendation makes more sense if your section count there is zero.
Look at the page-type distribution. If you have 200 blog posts and three product pages, your AEO content gap may be product-side, not editorial.
Spot stale crawls — pages last crawled 30+ days ago may have drifted from what AI actually sees today.