The setup
A manufacturer came to us for a rebuild: a new website and new systems underneath it.
Ask any owner who's been through a rebuild what scares them and it usually isn't the design. It's this: the old site, ugly as it may be, has spent years earning its place in Google's results. Customers find it. It brings in work. A careless rebuild can wipe that out in a weekend, and most owners only find out a month later, when the phone gets quiet and somebody finally looks at the traffic.
So before we talked about what the new site would look like, we talked about what it was not allowed to lose.
What we set out to do
The goal was simple to say and strict to enforce: rebuild the site and the systems, and carry the search presence over with it. Not approximately. Page by page.
We work diagnostic first, which in this case meant counting before building. If we couldn't tell the client exactly how many pages Google had indexed before the rebuild, we'd have no honest way to tell them what survived it. "Trust us, it's fine" is how traffic disappears.
What we did
First, we inventoried the starting point. Google had 237 pages from the old site in its index. Every one of those pages was a small asset: a product page, a resource, a landing page somebody had found through search at some point. Two hundred thirty-seven was the number the rebuild had to answer for.
Then we mapped the old site to the new one. Every indexed page got a destination. If a page still deserved to exist, it was carried over. If the content was being consolidated, the old address was pointed at the new home so the value moved with it. Nothing was left to chance, and nothing was quietly deleted because it was inconvenient to rebuild.
Then we built the new site and systems on that map, launched, and counted again. The count is the whole point. A rebuild postmortem that doesn't end in a number is just a story.
One more thing worth naming, because owners misunderstand it: "carried over" means the substance survived, not that the old site was frozen in amber. The pages kept their content and their standing; where an address changed, the old one was pointed at the new one so anyone, or any search engine, following the old trail still arrived. And this was never just a website job. The rebuild covered the site and the systems underneath it, which is exactly when presence usually gets lost, because everyone is watching the new machinery and nobody is watching the index.
That's also why the check has to happen promptly after launch, not at the next quarterly review. If a page does go missing, you want to know while fixing it is a small job, not after the traffic reports come in and the trail has gone cold.
What we measured
235 of 237.
Two hundred thirty-five of the old site's indexed pages were carried over intact into the new one. That's every page but two, preserved through a full rebuild of the site and the systems under it.
We'll be honest about the limits of that number. It measures indexation: the pages Google knew about before are still there and still known after. It's the foundation search traffic stands on, and it's the part rebuilds usually destroy.
What didn't make it
Two pages didn't carry over.
We're publishing the 235 instead of rounding it up to "everything" because the two matter. A rebuild process that can't tell you which pages it lost can't tell you anything. The discipline isn't perfection. It's knowing, page by page, what you have.
What an owner should take from this
First, your search presence is an asset, even though it never appears on a balance sheet. Years of indexed pages are years of customers being able to find you. Treat a rebuild like you'd treat moving a store: the regulars have to be able to find the new location on day one.
Second, if you're hiring anyone to rebuild your site, ask one question before you sign: "How will you prove to me, in numbers, that my search presence survived?" If the answer is a shrug or a promise, that's your answer. The proof is a count before and a count after, and it costs almost nothing to do if it's planned from the start. It's nearly impossible to reconstruct afterward.
Third, sequence matters. The inventory has to happen before the rebuild, not after the drop. By the time traffic falls, the map of what you had is already gone.
This is why we build the foundation before we build anything clever on top of it. A manufacturer doesn't need a prettier way to lose two hundred pages of search equity. They need the new system, and the traffic that feeds it, in the same launch. That's what 235 of 237 means.