> ## Documentation Index
> Fetch the complete documentation index at: https://help.get-ryze.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Crawl errors

> How to diagnose crawl errors in Search Console — 404s, soft 404s, server errors, redirect problems, and blocked pages — and which ones actually matter.

Crawl errors appear in **Search Console → Pages**. Most stores have some, and most of them do not matter. Knowing which do is the useful skill.

## The error types

| error                        | meaning                             | matters?                                |
| ---------------------------- | ----------------------------------- | --------------------------------------- |
| **404 / Not found**          | Page does not exist                 | Only if it has links or traffic         |
| **Soft 404**                 | Returns 200 but has no real content | Yes — wastes crawl, signals low quality |
| **Server error (5xx)**       | Site failed to respond              | Yes — fix urgently                      |
| **Redirect error**           | Chain, loop, or broken destination  | Yes                                     |
| **Blocked by robots.txt**    | Crawler was disallowed              | Only if unintentional                   |
| **Excluded by noindex**      | Directive present                   | Only if unintentional                   |
| **Discovered — not indexed** | Google found it, chose not to index | Often, at scale                         |
| **Crawled — not indexed**    | Google read it, chose not to index  | Often — usually thin content            |

## Which to fix first

**Server errors.** A 5xx means the site was unreachable. Persistent 5xx damages crawl rate and rankings. Check hosting and app-induced load.

**Soft 404s.** Empty product pages, empty categories, pages that removed their content but still return 200. Fix by populating, removing, or [redirecting](/shopify/redirects). These accumulate silently on stores that empty pages instead of deleting them.

**Redirect chains and loops.** Collapse `A → B → C` to `A → C`. Loops break crawling entirely.

**404s with inbound links or traffic.** These are lost value. Redirect them to the closest genuine equivalent.

**"Crawled — currently not indexed" at scale.** Google read the pages and declined. Nearly always thin or duplicate content — see [programmatic SEO](/concepts/programmatic-seo).

## Which to ignore

**404s for pages that should not exist and have no links.** A 404 is the correct response. You do not need to redirect every deleted page.

**Blocked URLs you intentionally blocked.** Filter and sort parameters showing as blocked is the system working.

**Small numbers of "Discovered — not indexed"** on a large catalogue. Some of this is normal.

Do not chase zero errors. A large store with zero reported issues usually means something is not being crawled at all.

## Diagnosing at scale

If discovered-but-not-indexed counts are very large relative to your product count, the likely cause is a crawl trap — usually [faceted navigation](/shopify/faceted-navigation) generating unbounded URL combinations. Google is spending its budget on filter permutations instead of products.

Check crawl stats: if parameter URLs dominate, that is the problem. See [crawl budget](/concepts/crawl-budget).

## What Ryze does

Crawl errors, redirect chains, soft 404s, and broken pagination are detected in the [technical audit](/ecommerce/technical-audit), prioritised by impact, and fixed with approval.

## Related

* [Crawl budget](/concepts/crawl-budget)
* [Orphan pages and canonicals](/troubleshooting/orphan-pages-canonicals)