I've watched a brand burn through seven figures of organic traffic in eleven days. Not a penalty. Not a competitor. A redesign that launched without a single redirect map, pushed live on a Thursday afternoon, right before a holiday weekend. The new site looked fantastic. It just didn't rank for anything anymore.
That's the whole risk in one story. A redesign touches your URLs, your templates, your internal links, and your checkout flow all at once, and every one of those layers carries traffic you've already paid for in time and content. So the real work isn't picking a new theme. It's building a pre-launch audit that protects what's already working. Smart brands bring in a seo agency for ecommerce before the wireframes get approved, not after the launch traffic chart starts sloping down, because the decisions that matter get made months before go-live.
Here's the checklist I'd run, in the order I'd run it.
Start with your highest-performing pages, not your homepage
Most redesign briefs open with the homepage. That's backwards. Your homepage probably converts fine and ranks for your brand name, which means it can survive almost any redesign. The pages carrying real weight are the category pages and the long-tail product pages that pull in non-branded search traffic. Pull your last twelve months of organic landing page data in Google Analytics or Search Console and sort by sessions. The top 50 entries on that list are your protected assets. For each one, write down three things: the exact URL, the primary keyword it ranks for, and the revenue it produced. That spreadsheet becomes the accountability document for the entire project.
When a designer suggests merging two category pages because they look similar, you check the sheet. When a developer wants to change the URL structure to something cleaner, you check the sheet. I'd rather see a brand delay a launch by three weeks than watch a top-20 landing page 404 on day one. One more thing on this step: check which of those pages rank for question-style queries. More than a few will, and those are the pages you least want to break.
Build the redirect map before anyone touches a URL
This is the step brands skip, and it's the one that does the most damage. Every URL that changes needs a 301 redirect pointing to the closest equivalent page on the new site, not to the homepage. Redirecting forty product pages to your homepage tells search engines those pages are gone, and the rankings go with them.
Google's own guidance on site moves and redirects is the place to start, and it's worth reading twice. According to Google Search Console documentation, a 301 passes signals from the old URL to the new one when the content is genuinely equivalent, and it's the only reliable way to preserve rankings through a migration. A few practical rules worth holding to:
Map old URL to new URL one to one wherever the content still exists.
Retire pages that no longer have an equivalent, and point them to the nearest parent category.
Keep the old redirects live indefinitely. There's no expiration date that helps you.
Recheck the map after launch, because developers add temporary URLs during the build.
I'd also stop anyone from stacking redirect chains. One hop, every time.
Navigation changes quietly break your internal linking
Think about how much traffic your navigation and footer send to category pages. Now imagine a designer flattening a five-level menu into three because it tests better in a prototype. It often does test better, and it often strips internal links from pages that depended on them.
Internal links are how search engines understand which pages on your site matter. The World Wide Web Consortium publishes the technical standards that underpin how links work on the web, and the W3C has documented hyperlink fundamentals since the earliest days of the web. The practical takeaway for a redesign is simpler than the spec: if a page had twenty internal links pointing to it before, it should have something close to twenty after. Before launch, crawl both versions of the site and diff the internal link counts per page. Any page that loses more than a third of its inbound internal links deserves a manual fix, usually by adding contextual links from category copy or related-product modules.
Test mobile usability on a real budget phone
Your team tests on the newest iPhone. Your customers don't all own one. Pull up the staging site on an older Android device with a slow connection and try to complete a purchase. Time each step from product page to confirmation screen. That's the test that matters, and it rarely happens because nobody on the project owns an old phone.
Watch specifically for sticky add-to-cart bars that cover product options, filter panels that trap the scroll position, and forms that require pinch-zooming. On mobile, every extra tap before checkout costs you conversions you'll never see in the data as a lost sale, only as a slightly worse conversion rate you'll blame on ad quality.
Rewrite product content, don't just migrate it
This is the step that separates a redesign from a rebuild. If you're moving 2,000 product pages into new templates, that's your one chance to fix thin descriptions, duplicate manufacturer copy, and missing specifications without doing it twice.
Tackle it by revenue tier. Your top 100 products get fully rewritten descriptions with specs, sizing, materials, and answers to the questions your support inbox keeps receiving. The next 400 get a lighter pass. Everything else gets migrated as-is and flagged for later. That tiered approach is the difference between a content project that finishes and one that stalls at product 60.
Set your analytics baseline a month before launch
You can't measure what you never recorded. Pull a clean baseline thirty days out: organic sessions by landing page, conversion rate by device, average order value, and checkout abandonment by step. Save it somewhere outside your analytics platform, because tracking codes break during migrations and you'll want a reference point that doesn't depend on the new setup working on day one.
Then verify your tracking before launch, not after. Test that events fire on the staging site, that cross-domain tracking survives if you use a separate checkout subdomain, and that your conversion goals still map to real URLs. Jakob Nielsen's research on usability at the Nielsen Norman Group, available at Nielsen Norman Group, has consistently shown that users abandon flows when friction appears at unexpected points, and broken or misattributed tracking creates exactly that kind of blind spot for the team trying to fix it.
The launch week sequence I'd use
Keep the old site accessible on a staging environment for at least two weeks after launch. Submit the new sitemap the day you go live. Watch Search Console's coverage report daily for the first week, and treat any spike in 404s as a five-alarm fire, not a backlog item. Announce the redesign to your email list a few days in so returning visitors don't assume you've closed.
And schedule the launch for a Tuesday morning. Never a Friday. Never before a major sale event. That single scheduling decision has saved more revenue than most technical fixes I've seen.
A redesign is one of the few moments when you can improve everything at once, which is exactly why it's also the easiest way to lose ground you spent years gaining. Run the checklist, protect the pages that pay your bills, and give the new site a real chance to outperform the old one. What's the first thing on your list, and who on your team owns it?
Copyright © 2026 California Business Journal. All Rights Reserved.
For California Business Journal Disclaimers, go to https://calbizjournal.com/terms-conditions/.
