Nobody adds a broken link to their own website. Every broken internal link on your site was a working link on the day it was written.
That is the thing to understand about link rot: it is not an event, it is an accumulation. Links do not break when you touch them. They break when something else moves, and the link is left pointing at where that thing used to be.
The ordinary ways links break
Almost every broken internal link on a small business site comes from a handful of routine causes:
Renamed pages. You tighten a page title, the CMS regenerates the URL slug to match, and every link to the old slug now points at nothing. Some platforms create a redirect automatically when this happens. Many do not, or only do it for some content types.
Deleted content. A retired service, a discontinued product, an old staff bio. The page is removed thoughtfully, but the links to it, from blog posts, footers, and other service pages, are not part of anyone's deletion checklist.
Restructures. Moving blog posts under a new section, reorganizing services, changing how categories work. One restructure can quietly invalidate dozens of internal links at once, especially links written by hand inside body text where the CMS cannot see them.
Platform migrations. New platforms generate URLs differently. Trailing slashes appear or disappear, dates leave the blog URL format, capitalization rules change. The migration looks complete because the pages all exist. The old links to them do not know that.
Linked files, not just pages. PDFs of menus, price lists, and intake forms break the same way pages do, and they are even less visible, because file links rarely appear in any CMS content listing.
Each cause is mundane. That is why the problem compounds: a five-year-old site has lived through enough renames, deletions, and reorganizations that some percentage of its internal links are stale, and nothing in day-to-day operation surfaces them.
What broken links actually cost
The cost is real but easy to misjudge in both directions.
For visitors, a broken internal link is a dead end you built. A visitor reading your blog clicks through to the service the post describes and gets a 404. Some will hunt for the right page. Many will not. The visit was almost converted, and the link spent it.
For search engines, the picture is more nuanced, and worth stating honestly. A few broken links will not tank your rankings, and anyone who tells you otherwise is selling urgency. But internal links are how crawlers discover your pages and how authority flows between them. Every link pointing at a 404 is discovery effort spent on nothing, and a page that has lost its inbound internal links becomes an orphan that crawlers may revisit less often. On large sites this is a crawl efficiency issue. On small sites it is mostly a signal-quality issue: a site full of dead ends reads as unmaintained.
For trust, the effect is quiet but cumulative. Visitors do not file bug reports. They just register, somewhere below the level of conscious thought, that the site feels neglected, and neglected sites make people hesitant about handing over a phone number or a credit card.
Why you have not noticed
Broken internal links hide well, for structural reasons.
You navigate your own site through the same few paths every time, and those paths are the well-maintained ones. The broken links live in the long tail: old blog posts, deep service pages, the footer of a template nobody has opened in a year.
There is also no error on your side. The linking page loads fine. The breakage only exists at the moment of the click, and the person clicking is a visitor, not you. The site fails politely, one visitor at a time, with no log entry anyone reads.
How to find them without making it a project
You do not need to make link checking a hobby. A practical routine has three parts:
Run a crawler occasionally. Free desktop and web-based link checkers will crawl your site, follow every internal link, and list the ones returning errors. On a typical small business site this takes minutes. Doing it once will probably surprise you; doing it quarterly keeps the pile from rebuilding.
Watch your 404s. Your analytics or your hosting logs can tell you which missing URLs people actually request. This is the triage view: a broken link nobody clicks is a low priority, while a broken link receiving weekly traffic is costing you visitors right now and deserves a redirect today.
Fix at the source when you can, redirect when you cannot. The clean fix is editing the link to point at the right page. When the old URL is referenced from places you cannot edit, external sites, old emails, printed materials, a redirect from the old URL to the best current equivalent catches everyone. Prefer one direct redirect over stacking redirects on redirects, which creates its own slow-motion problem.
And when you delete a page deliberately, take thirty seconds to search your own site for links to it. Deletion checklists prevent most future rot at the moment it would otherwise begin.
Where monitoring fits, honestly
Continuous monitoring and a full link audit are different instruments, and it is worth being precise about which does what.
A full crawl-based audit examines every link on every page. It is thorough, and it is a point-in-time snapshot that starts going stale the day after you run it. Ongoing monitoring, the kind of always-on checking we describe on our start here page, watches key pages and key failure modes continuously, but checking a handful of important URLs is not the same claim as auditing every link on the site, and a monitoring product should say so plainly.
The combination is what works: an occasional full crawl to clear the backlog, plus continuous watching of the pages where a dead end costs you most. If you are weighing what different tools actually check, our comparison page lays out the differences without the marketing gloss.
Link rot is not a crisis. It is entropy. Sites accumulate broken links the way houses accumulate drafts, and the owners who stay ahead of it are simply the ones who check on a schedule instead of waiting for a complaint that, statistically, will never be filed.