Enterprise platform migrations
Replatforms are the highest-stakes work in SEO. Done badly they cost millions. Done properly, they are a chance to fix years of accumulated problems in one go. I led the SEO for Chemist Warehouse's 18-month replatform through its 2025 launch. Organic revenue grew 38% in the year after.
Why migrations go wrong
Almost every migration horror story traces back to the same small set of causes. The URL mapping gets treated as a spreadsheet chore and handed to whoever has time. Redirects get validated late, or only sampled. Nobody crawls staging properly because staging is behind auth and the crawler was never set up for it. And rendering changes ship silently, because to the platform team they are not changes at all, just how the new stack works.
None of these are exotic problems. They are boring, procedural problems at a scale where boring and procedural is exactly what fails. A mapping error rate of one percent is invisible on a five hundred page site and a catastrophe on five hundred thousand.
The pattern I have seen over and over: migrations do not fail on the hard decisions. They fail on the tedious ones that nobody owned.
How I approach one
Before anything moves, a baseline: what currently ranks, what earns, what crawls, and where the traps are in the existing architecture. Then the URL mapping, which at enterprise scale is a data engineering task, not a spreadsheet. Rules first, exceptions second, validation over the whole set rather than a sample. Staging gets crawled and rendered the way Google will crawl and render it, not the way the demo walkthrough does.
Launch week is monitoring: log files, index coverage, redirect behaviour in the wild, and a short feedback loop with the developers so issues get fixed in hours rather than sprints. Then the part that gets skipped most often, a genuine post-launch phase. The first 30 days tell you whether the mapping held. The first 90 tell you what the migration actually did to performance.
What I have learned
- A migration is a stakeholder project wearing a technical costume. Half the job is making risk legible to people who do not read crawl logs.
- The SEO cannot own the whole project, and should not try. The job is to know exactly which details are load-bearing and to hold the line on those.
- Treat the replatform as an opportunity, not just a defence. The Chemist Warehouse result came from fixing architecture during the move, not from merely surviving it.
- Post-launch attention is where good migrations are separated from lucky ones.
