Platform Migration
A migration is a data problem, a redirect problem and a trust problem at once. The technical part is usually the easiest of the three — the hard part is that the business keeps trading while you do it.
How I use it
Two catalogues that disagree, and no obvious winner
On the Wee Rock Art rebuild, the Square catalogue turned out to be a 2019 snapshot — sync worked at launch and then quietly died, stranding 129 products added over the four years since. Auditing both before touching either is what established WooCommerce as the source of truth rather than assuming the newer system was.
Seeding a live production catalogue without breaking trading
Wrote idempotent, re-runnable importers that created 132 missing items with images, built ten collection categories and corrected eighteen stale 2019 prices upward. Verified by live read-back: 150 of 150 present, zero price mismatches, zero errors.
Legacy records that look like junk but are not
The same catalogue held 121 orphaned items that would normally be archived on sight. They were the owner’s point-of-sale library for in-person markets, so I excluded them at the storefront layer instead of touching her live system. Migrations go wrong when you optimise the data and break the business using it.
Doing the same migration twice
Built Switch Cart, a WooCommerce-to-Payload migration tool, rather than repeating the same export-and-reconcile work by hand.
What I use
- WooCommerce
- WordPress
- Square Catalog API
- Astro
- Python
Where you can see it
Related skills
Want the detail behind any of this?
I can walk through any of it in as much depth as you want.