Website Migration SEO: Keep Your Rankings When You Move

An enquiry came in recently from a business with a healthy site and years of accumulated rankings, wanting to move platforms. They’d had quotes already. Every quote talked about the build — the design, the features, the launch date. Not one of them mentioned what would happen to the rankings, and that silence was the thing keeping the owner awake. She didn’t need convincing the new platform was better. She needed to know the move wouldn’t cost her the ground she’d built.

That’s the right thing to worry about, and it gets worried about far too late. So here’s the whole picture: why migrations lose rankings, what actually protects them, and the checklist I run on every move I do.

Why rankings drop after a website migration

Google doesn’t rank your website. It ranks your URLs — individually. Every page that brings you traffic has earned its position over years: it’s been crawled hundreds of times, linked to, clicked on, and slowly trusted. That trust is attached to the address, not to your business name.

Move platforms and nearly every address changes. WordPress puts products at /product/; Shopify puts them at /products/. Wix generates its own URL patterns; so do Magento and EKM. The content might come across perfectly, and as far as Google is concerned every page it trusted has vanished, replaced by a site full of strangers.

It’s moving house without telling anyone — not the bank, not the doctor, nobody but the milkman. The post doesn’t follow you. It piles up at a house you no longer live in, and eventually the senders stop writing.

A wobble, not a loss

Will a migration affect your rankings? Yes. Anyone who tells you Google won’t notice a platform move is selling something. Google recrawls the whole site, reconciles old addresses against new ones, and re-evaluates as it goes. Expect two to six weeks of turbulence — positions shifting, impressions dipping and recovering.

Managed properly, that’s the whole story: a wobble, then business as usual, and often better than usual, because the new site is usually faster and page speed is a ranking factor. I’ve watched sites come out of a migration stronger than they went in — see the migration timeline guide for what that recovery actually looks like week by week.

The difference between those two outcomes isn’t luck, and it isn’t the platform. It’s whether somebody built the redirect map.

What the impact actually looks like

“Two to six weeks of turbulence” is the honest summary, and it’s also uselessly vague if you’re the one signing off the move. So here is the shape it takes in the migrations I’ve run — my own observations rather than published research, because published averages fold together migrations done well and migrations done badly, and those are different events with different outcomes.

The first thing that moves is impressions, not positions. Google recrawls the new addresses over days, and while it does, pages drop out of the index temporarily and reappear. Search Console shows impressions falling before it shows rankings falling, which is why impressions-by-page is the number to watch daily rather than a rank tracker. A tracker checks a handful of terms once a day from one location, and it will happily tell you everything is fine while a third of your pages sit uncrawled.

Positions follow, and they wobble rather than fall in a straight line. A page sitting at four might read six, then eleven, then five, then four again across a fortnight. That is reconciliation, not damage, and watching it hour by hour will only cost you sleep. What matters is the direction of travel at week four. Rankings still drifting downwards at week six or eight are not recovering — they are telling you something is broken, and it is nearly always one of three things: a missed redirect, a blocked crawl path, or metadata that got rewritten during the move.

Brand queries come back first, because Google is already confident about those. The long tail comes back last, and that is where the quiet losses hide. A page that used to pick up a dozen small queries can stop picking them up without anything visible happening to the terms anyone is tracking, which is why the baseline needs to be a full Search Console export rather than a list of your twenty favourite keywords.

301 redirects: the map where migrations are won

A 301 redirect is a permanent forwarding instruction. It tells Google: this page has moved, here’s the new address, transfer everything it earned. It is the single mechanism by which your SEO survives a migration, and it has to cover everything — every product, every category, every page, every blog post. Mapped individually, old URL to new URL, in a spreadsheet you can hand to anyone and defend line by line.

Not wildcards and hope. Wildcard rules catch the pattern and miss the exceptions, and the exceptions are always the pages that matter — the product that ranks page one for its money keyword, the guide that brings half the organic traffic. One missed redirect on a page like that costs more than the migration fee.

On Shopify the mechanics are straightforward: a built-in URL Redirects tool that takes a CSV import, which is what any store of real size needs — I’ve built redirect maps that run to hundreds of rows from a single WordPress store, and a 24,000-customer EKM migration where the export was so non-standard the map had to be built from scratch. The tool is simple. The discipline is the product.

The redirects get treated as the boring bit, which is exactly why they get skipped. The build gets the meetings; the redirect map gets whoever’s left on a Friday afternoon.

The site migration checklist — and why it starts before the build

Every migration I do runs the same sequence, and the order is the point. The first deliverable in a migration isn’t the design. It’s the baseline.

The baseline earns its place twice. It’s protection during the move, and it’s proof afterwards — the before-and-after that separates “the migration cost us traffic” from “the migration cost us nothing and the new site converts better.” Without it, whoever did the move wears every dip that follows, deserved or not. I take the baseline before I quote, because I want to know the job is winnable before either of us commits to it. If you’re still weighing up whether to move at all, that’s a different decision — and sometimes the honest answer is to stay put.

Website redesign SEO: a redesign is a migration wearing a nicer name

The trap that catches businesses who never planned to migrate: a redesign that changes URLs is a migration. Same platform, same domain, new structure — pages renamed, categories merged, old blog posts culled. Every one of those changes breaks an address Google trusted, and a “refresh” that never called itself a migration gets none of a migration’s discipline. No baseline, no redirect map, no monitoring. Just a slow, unexplained decline that gets blamed on the algorithm.

If your agency is proposing a redesign, ask one question: does any URL change? If the answer is yes — or a shrug — the checklist above applies in full.

How many hours should you budget for migration SEO?

It’s a fair question and a hard one to find answered, so here’s the straight version: the SEO half of a migration is not the expensive half, and it does not scale the way the build does.

The baseline takes a morning: a full crawl of the old site, a Search Console export, and a rankings snapshot. The tooling does most of the work, and the value isn’t in the hours — it’s in having taken the measurement before anything changed.

The redirect map is the part that scales, and it scales with URL count rather than with how complicated the site is. A fifty-page brochure site is an afternoon. A few hundred products is a day, and most of that day goes on the exceptions: the URLs that never followed the pattern, the pages renamed years ago that still carry links, the categories that got merged and forgotten. A 24,000-customer store whose export came out non-standard took days, because the map had to be rebuilt from the raw data rather than generated from it.

Then monitoring, which is an hour a day through cutover week and an hour a week for the month after. Not because Search Console takes an hour to read, but because reading it properly means comparing it against the baseline and chasing whatever looks wrong.

Across a typical small-business migration that’s two to four days of SEO work, spread over about six weeks. If a quote presents migration SEO as a separate line item running into five figures, ask what’s in it.

If the migration has already gone wrong

If you’re reading this after the fact — traffic down, nobody quite sure why, and someone telling you it’s an algorithm update — the diagnosis is usually quick, and it’s worth doing before anyone rebuilds anything.

Start with the old URLs rather than the new site. Take the Search Console export from before the move if you have one, or the Wayback Machine if you don’t, and request a sample of the old addresses. Every one that returns a 404 instead of forwarding to its replacement is a page whose earned authority is currently going nowhere. That single check explains the majority of post-migration declines, and the fix is the redirect map that should have existed in the first place.

If the redirects turn out to be sound, look at what changed on the pages themselves. Titles rewritten wholesale during the move are a common culprit, because the migration and a copy refresh got bundled into one event and now nobody can tell which change caused the drop. That’s the argument for moving first and improving afterwards: one change at a time, so the trail stays readable. I keep every change to a client’s site in version control for exactly this reason — when a page moved from position five to position forty-nine on a site I looked after, tracing it took five minutes, because there was one commit, one date, and one line of difference to look at.

Then check you’re crawlable at all. A staging build that shipped with a noindex header still in place, or a Disallow: / left in robots.txt from the pre-launch lockdown, is a startlingly common way to go invisible — and it presents exactly like an SEO problem to anyone looking at the traffic graph instead of the headers.

Recovery runs slower than the loss did. Redirects put in three months late do still work, because authority transfers whenever Google next crawls the old address. But the pages have to be recrawled, re-evaluated and re-earned, and you’ve spent the interim not ranking for terms you used to own. Which is the entire reason the baseline comes first.

What this means if you’re about to move

The platform decision gets all the attention, and it matters — but whichever way you go, the SEO risk is the same shape, and it’s managed the same way. Baseline, redirect map, metadata, sitemap, monitoring. None of it is glamorous. All of it is the difference between a move and a loss.

If you’re planning a migration and the quotes you’re holding don’t mention redirects, that’s the tell. The migration cost guide covers what a proper move costs and what’s included — the redirect map isn’t an add-on, it’s the job. And if you want a straight answer on whether your rankings would survive a move, get in touch — the baseline takes me a morning, and it’s the most useful document you’ll read before you sign anything.

Related: the Shopify migration cost guide covers what a proper move costs, the migration timeline shows the two weeks day by day, WordPress to Shopify and Wix to Shopify walk the two most common source platforms, and should I fix or rebuild my website is the decision that comes before any of it.