#This is where dealers lose traffic without noticing
A new site goes live. It looks better. Everyone is pleased. Three months later leads are down and nobody can say why, because the drop was gradual and the new site is obviously nicer than the old one.
What happened is that a few hundred web addresses stopped existing, and the search rankings attached to them went with them.
This is one of the few areas in marketing where the risk is almost entirely preventable and almost entirely ignored. The prevention is unglamorous and takes an afternoon.
#Write down what you have before touching anything
Before launch, get a complete list of every page currently on your site, and which of them get traffic or bring in leads.
Most stores are surprised here. There are usually old model pages, service pages and campaign pages that quietly bring in business and are not in anyone plan for the new site.
Nobody can protect pages nobody knew existed. This one step prevents the majority of migration damage, and it costs you an export and half an hour of reading.
#Every old address needs a specific new home
When a page moves, the old address must point at the new equivalent. Do that and the value mostly carries over. Skip it and you start from scratch on that page.
Match like to like. The old brake service page goes to the new brake service page, not to a general service page, not to the homepage.
This is tedious and it is the whole job. If your new site vendor cannot produce a list showing where each old address will point, they have not done this part, and that is worth catching before launch instead of after.
#Do not redirect everything to the homepage
It is the most common shortcut and it is damaging. Search engines treat a mass redirect to the homepage as the pages being gone, so whatever those pages had earned is lost.
It is also bad for the customer. Someone searching for your transmission service lands on your homepage and has to start over. Most will not.
If a page truly has no equivalent, send it to the closest relevant section, not the front door.
#Inventory and tracking break quietly
Two things reliably break on launch and neither announces itself.
Inventory: confirm vehicles are appearing, that their pages work, and that the feed is running. A site that looks perfect but has not pulled inventory since Tuesday is not obvious from the outside.
Tracking: confirm your analytics and lead notifications are installed on the new site. Plenty of stores discover weeks later that nobody has been receiving form submissions, which is the expensive version of this mistake.
#Treat launch day as a beginning, not a finish
The work is not done when the site is live. That is when checking starts.
Same day: submit the new site to Google, confirm the old addresses are redirecting properly, test forms and service booking on an actual phone, and place a real test lead to confirm it reaches a person.
First month: watch which pages Google reports as missing and fix them as they come up. Most migration damage is repairable if caught in the first few weeks, and much harder to unwind after a few months.
#What normal looks like afterward
Expect a dip. A modest drop for a few weeks while search engines re-learn the site is normal and not cause for panic.
What is not normal is a sharp sustained fall, or specific important pages disappearing from search entirely. That means addresses were not mapped properly, and it is fixable if someone looks.
Set the expectation with your vendor before launch, in writing: what size dip is acceptable, how long recovery should take, and who investigates if it does not. Having agreed that in advance is the difference between a diagnosis and an argument.
TABLE
The migration checklist, in the order that prevents the loss
Steps one and two happen before the new site exists. Almost every recoverable migration failure is a store that skipped them.
| When | Step | Skipped means |
|---|---|---|
| Before | Export every indexed URL, ranking and page-level traffic | You cannot prove what was lost, or get it back |
| Before | Inventory every tracking tag and conversion | Weeks of blind reporting after launch |
| Before | Map every old URL to one specific new URL | Redirect-to-homepage, and the pages are treated as gone |
| Launch day | Crawl the new site and diff it against the export | Missing pages found by a customer, not by you |
| Launch day | Verify the inventory feed and every tag fires | Silent breakage nobody notices for a month |
| Week 2 | Re-crawl, check indexation and redirect chains | Chains and loops that bleed authority quietly |
| Week 6 | Compare traffic and rankings against the export | The point at which recovery is still cheap |
Carbide migration engagements, reviewed 2026-09-02.
Migrations are the one project where the damage stays invisible until it is permanent, which is why they are revisited in the dealer website SEO insights.