A broken link is a promise that fails on click — small individually, corrosive in aggregate. Visitors read 404s as neglect; Google reads them as decay; and the site accumulates them silently, because nothing announces a link’s death.
Where they come from
Your own restructures (pages renamed without redirects), the web’s rot (that supplier you linked in 2022 changed URLs), deleted media, and typos that never got clicked by anyone who’d report them. Every aging site has them — the question is only how many.
Finding them, free
Search Console’s indexing report lists 404s Google hit; free crawlers (Broken Link Checker plugins, online scanners) walk every link on demand. Monthly for busy sites, quarterly for calm ones — findings drop to near zero once the backlog is cleared.
The fix hierarchy
Internal links: correct them at the source. Incoming dead URLs: 301 to the closest living page. External rot: relink or unlink. And the safety net — a useful 404 page with search and paths onward, because some breakage always slips through. We clear link backlogs inside maintenance — the least glamorous fix with the most quiet gratitude.
Frequently Asked Questions
Indirectly but really — wasted crawl, leaked authority and decay signals accumulate with the 404s.
Renames without redirects, the web's natural rot, deleted media and unnoticed typos — every aging site collects them.
Search Console's 404 report plus a free crawler — monthly or quarterly depending on how often the site changes.
Correct at the source — the link itself; redirects are for incoming URLs you can't edit.
Your design, a search box and paths onward — the safety net for what slips through.
No — the web rots naturally; the signal is trend and volume, not perfection.
English 
Czech