A site audit gives you a count of broken URLs, and the count looks like a to-do list. It is not. Twenty broken URLs might mean twenty deleted pages, one broken navigation link repeated twenty times, or an old article address that still has a perfectly good replacement. Each situation needs a different fix.
We recently found this distinction while auditing NetPageTwo. The crawl reported broken URLs, but some affected destinations had existing articles under different dated addresses. Publishing replacement articles would have created more work and potentially duplicate content. The first job was to map the old address to the right existing resource.
Do not work from a screenshot of the audit summary. Export at least four columns: broken destination, source page, HTTP status, and link text. The source page tells you how widespread the issue is. If the same broken URL appears in the header, footer, and twenty blog posts, fix the shared template first. If it appears in one article, the repair is probably editorial.
Confirm the status directly. A 404 means the resource was not found. A 410 means it is intentionally gone. A timeout, blocked bot, or 5xx error may be temporary or infrastructure-related. A link checker can also report a false positive when a website blocks automated requests. Open the destination and inspect its response before changing anything.
Search your own WordPress posts and pages by the old slug, title, and topic. Look for a published resource that answers the same question. The closest topical match is not necessarily the correct match: a page about how schema works is not a substitute for a removed schema testing tutorial. Read the content and ask whether a person following the original link would reasonably feel that they reached the expected answer.
If the article exists under a different date-based permalink, update the internal link to the canonical current address. If the old URL has backlinks or may still be used externally, a permanent 301 redirect to the equivalent page may be appropriate. If no equivalent exists, remove the broken link or replace it with an honest explanation, rather than redirecting everybody to the homepage.
| Situation | Preferred action | Reason |
|---|---|---|
| Same article exists at a new URL | Update internal links and consider a 301 from the old URL | Preserves the intended destination |
| Article merged into a broader, genuinely equivalent guide | 301 to the relevant section or guide | Consolidates useful information |
| Article intentionally removed with no replacement | Remove links; allow an appropriate 404 or 410 | Avoids misleading redirects |
| External reference is broken | Find the official current source or remove the claim | Keeps evidence verifiable |
| Important buyer question has no useful page | Create a new page based on demonstrated need | Fills a real content gap |
Sending every removed URL to the homepage is convenient for the website owner and usually unhelpful for the visitor. Google may treat irrelevant redirects as soft 404s. Redirect chains also make debugging harder. Prefer one direct redirect from the old address to the best equivalent page, and update your own internal links so visitors do not need the redirect at all.
Be especially careful with dated WordPress permalinks. If an article moved from a June address to an October address, verify the canonical URL and publication history before changing the site’s permalink structure. A global permalink change can create a much larger problem than the one you started with.
After updating a link, load the source page and click it. Check that the destination loads, its canonical tag points to the intended URL, and the final content matches the link text. If you added a redirect, test the HTTP response and ensure it ends at the right page without looping. Re-crawl the affected pages after the changes are live.
Use Google Search Console to watch indexing issues, but remember that a crawler’s broken-link report and Search Console’s indexing report are different datasets. Neither should be treated as a complete inventory of every link on your site.
Start with broken URLs that appear in navigation, pricing, service pages, or pages with meaningful traffic. Then handle internal links in articles, followed by broken external citations. For each unique destination, record the replacement URL or reason for removal, the pages updated, and whether a redirect was created. This gives you a small, auditable change log rather than a vague claim that “technical SEO was fixed.”
Once the list is clean, review your internal links by topic. Your AI search introduction, service pages, and relevant Field Notes should connect naturally where they answer successive buyer questions. Links are there to help people continue their research. They are not an excuse to attach every keyword to every page.