You are about to rebuild your website, or you already did and something feels off. Calls dropped. Pages that used to show up in Google do not. Either way, you need one document: the full list of addresses your site has ever used.
Here is how to build it this week.
Why do old web addresses still matter after a redesign?
Because other people kept them. Google has them indexed. Customers bookmarked them. A supplier linked to one from a directory in 2016. A review site points at another.
When an address stops working, all of that goes with it. The visitor sees an error. Search engines see a page that used to earn attention and now earns nothing.
A New Leaf Painting Contractors, a Jacksonville company that has been painting homes and businesses since 2010, learned this the hard way. Their site was moved to a new platform and the move went badly. Hundreds of old addresses were redirected to the homepage, and pages that search engines still listed were gone.
The fix started with a list.
Start with the addresses you can see today
Begin with the easy part. Pull every address on the live site.
- Open your XML sitemap if you have one. It is usually at your domain followed by /sitemap.xml.
- Run a crawl of the site with any standard crawler and export the results.
- Export the page list from your content management system.
Put them in a spreadsheet, one address per row. Add columns for the page title, the status, and a blank one called "decision." You will fill it in later.
Do not stop here. This list only shows what exists today. The trouble lives in what used to exist.
Dig back through the older platforms your site lived on
Most business sites have lived in more than one place. A site builder, then a WordPress install, then something a nephew set up. Each platform had its own habits for naming pages.
Ask yourself and your team:
- What platforms has the site been on?
- Did blog posts, service pages, or town pages ever have a different pattern, like dates in the address or a category folder?
- Is there an old backup, export, or staging copy sitting on a hard drive or with a former developer?
If you can find an old database export or backup, open it. It will contain every page that existed at that point. The Internet Archive's Wayback Machine is also worth a look. Enter your domain and browse old snapshots for pages you forgot.
This is the step most people skip. It is also where the old, valuable addresses hide.
Collect what search engines still list
Search engines remember longer than your CMS does. Use them.
- In Google Search Console, export the pages report and the performance report. Look at every address that has received impressions, not just the ones that are live.
- Search your domain with the site: operator and scroll through what Google shows. Odd addresses stand out fast.
- Check the pages report for entries marked as not found or redirected. Those are addresses Google still knows about that your site no longer serves properly.
Add every new address to the spreadsheet. Mark where each one came from. You are not judging yet. You are collecting.
Add the addresses your own records remember
Two more sources are worth ten minutes each.
Analytics. Pull a long date range of landing pages. Any address that brought a visitor in is an address that matters, even if it is gone now.
Server logs. If your host keeps them, they show what browsers and crawlers actually requested, including addresses nobody remembers publishing.
Then merge everything and remove duplicates. Be careful with small differences: a trailing slash, uppercase letters, or www versus no www can make the same page look like two. Pick one pattern and note the variants.
What you now have is the real inventory. For a site with years of blog posts and local pages, it can run to hundreds of rows. That is normal.
Decide per address: replacement page or clean gone
Now fill in the "decision" column. Every address gets one of two answers.
A replacement page. The old address had a real counterpart on the new site. Same topic, same service, same town. Point it there.
A clean gone. Some pages should never have been published. Thin pages, near-identical duplicates, and leftovers with nothing to say. These should return a plain "gone" message. That is honest, and search engines handle it well.
A New Leaf Painting had both kinds. Their old addresses were traced back through older platforms, and each one was sent either to the page that replaced it or to a clean gone for the thin, near-identical pages.
When you are unsure, ask one question: would a person who landed here be glad to arrive at the replacement? If yes, redirect. If you would have to squint to call it a match, it is gone.
Make each address answer in a single hop
A redirect should take a visitor from the old address straight to the final one. One step.
Chains happen when an address from the first platform points to one from the second, which points to one from the third, which finally lands on the current page. Every extra step is a chance to break, and it slows the visitor down.
To avoid chains, map every old address directly to the current destination, not to the previous destination. Your spreadsheet should have a column for the final address, and that column should never contain an address that itself redirects.
This is what the A New Leaf Painting team did. Every address answered in a single hop, to its replacement or to a clean gone. You can read how the whole project came together in the A New Leaf Painting case study.
Why redirecting everything to the homepage backfires
It is tempting. One rule, done in five minutes, no spreadsheet needed. Send everything to the homepage.
Please do not.
- A visitor who clicked a link about interior trim painting lands on a generic homepage. They leave.
- Search engines see an old, specific page that now points at something unrelated. Many treat that as a missing page anyway, so you keep the confusion and lose the value.
- You cannot tell which addresses are broken, because everything appears to work.
That last one is the sneaky part. The A New Leaf Painting move sent hundreds of old addresses to the homepage, and the pages search engines still listed were gone. A redirect that lands somewhere unrelated is not a fix. It is a quiet failure.
A clean gone is better than a lazy redirect. At least it is true.
Test the list and keep it for next time
Before launch, and again right after, test every row.
- Run the full list of old addresses through a bulk status checker or a crawler in list mode.
- Confirm each redirect returns a single hop and lands on the right page.
- Confirm each "gone" address says so, and does not quietly bounce somewhere else.
- Spot check twenty or so by hand, especially the ones that earned the most traffic.
After launch, watch Search Console for new not-found reports over the following weeks. Anything that appears is an address you missed. Add it to the list.
Then save the spreadsheet somewhere safe. The next redesign, platform change, or agency handoff will need it, and you will be the only person who knows it exists.
Your next step
Open a blank spreadsheet today and export your sitemap into it. Then pull the pages report from Google Search Console and add anything that is not already there. That is the first twenty minutes, and it will show you how big the job really is.
If the list turns out to be longer than you hoped and you want help tracing it, our SEO and analytics work covers exactly this, and you can reach us through get started.