The quickest way to lose a decade of search visibility is to rebuild a site with a tidier URL structure. The new structure is usually better. It is almost never better enough to be worth the loss.
When a company has been publishing for years, its URLs have accumulated value that doesn't show up in analytics. A buyer researching your category six months ago bookmarked your comparison page. A procurement team's internal wiki links to your pricing breakdown. A journalist's draft references your methodology post. None of these show up as referring domains until someone clicks, and by then the redirect chain has already cost you some of that traffic.
What actually breaks
Every inbound link from every article, directory and partner page that ever mentioned you now points at a 404. Redirects help, and they are not free: each one is a hop, and chains of them get truncated.
Search engines follow redirects, but they discount them. A 301 passes most of the original page's authority, but "most" is not "all." When you redirect thousands of URLs, those small losses compound. A page that ranked fourth might slip to seventh. A featured snippet you held for two years might go to a competitor whose URL never moved.
The technical cost is subtler. Browsers cache redirects aggressively. If you fix a redirect later because the first destination was wrong, visitors who loaded the old version will keep hitting the wrong page until their cache expires. Server logs fill with redirect chains that nobody intended: old URL to new URL to newer URL after a second migration. Each hop adds latency. Three hops on a slow mobile connection can mean the difference between a page that loads and one that times out.
Why teams want new URLs anyway
The desire is rational. Your original URLs were probably chosen in haste, or they reflect a product structure you've since outgrown. A SaaS company that started with /features now has three product lines and wants /product-name/features to keep them distinct. An agency that used /blog/post-title now publishes case studies, guides, and research, and wants the URL to signal content type.
These are real problems. The mistake is thinking the URL is where to solve them. A visitor scanning a search result cares whether your page answers their question, not whether your information architecture is elegant. The URL structure is for you. The content and metadata are for them.
What we do instead
We keep the URLs and change what is underneath. It is less satisfying and it costs the client nothing.
The pages get new templates, new navigation, new calls to action, and often new content. The technology stack can change completely. What doesn't change is the address. A page that ranked well at /old-path keeps ranking well at /old-path, now served by better code and better design.
When a URL genuinely must change—because the page topic has shifted beyond recognition, or a product was discontinued—we redirect it and accept the cost. But we treat it as a cost, not a convenience. The default is to preserve what already works.
This approach matters most for the pages you are not actively watching. Your homepage and your top three landing pages get monitored weekly. The guide you published in 2019 that still brings in two qualified leads a month does not. It keeps working because its URL never broke.
What actually breaks
Every inbound link from every article, directory and partner page that ever mentioned you now points at a 404. Redirects help, and they are not free: each one is a hop, and chains of them get truncated.
We keep the URLs and change what is underneath. It is less satisfying and it costs the client nothing.
Leave a Reply