48V Products ecommerce rebuild

By · Updated

Short answer

48vproducts.com left its old WordPress store for the shared storefront in this monorepo (`apps/web` and `apps/cms`). Catalog, checkout, and fulfillment run in that stack today. The old host has 301-redirected to the new store since late October 2026, and sales held steady or grew after cutover. Post-cutover PageSpeed lab scores are 100 on desktop and 75 on mobile.

Key facts

Store
[48vproducts.com](https://48vproducts.com/) — the only nameable ecommerce client in our case studies
Before
WordPress commerce on a separate subdomain
After
Next.js storefront (`apps/web`) and Payload CMS + Medusa commerce (`apps/cms`) in this monorepo
Commerce
Live Stripe checkout, order email, and product catalog synced from CMS to commerce
Business result
Sales held steady or grew after launch (no public revenue figures)
Cutover
Late October 2026 — path-preserving 301s from the old WordPress host (old `/product/<slug>/` URLs map to `/products/<slug>/`)
PageSpeed lab score after cutover
100 desktop, 75 mobile (Lighthouse 12 performance score, home page, measured 2026-11-07)
48vproducts.com storefront home — hero, navigation, and product lineup after rebuild off WordPress
Live 48vproducts.com (November 2026): Medusa checkout and catalog in this monorepo, rebuilt from its old WordPress store.

The starting point

48vproducts.com sold sodium-ion and related power products on WordPress, hosted on a separate subdomain. Product copy, pricing, and merchandising lived in the WordPress admin; plugins and theme overrides carried years of incremental changes. That stack was acceptable for early sales, but inventory, structured data for Merchant Center, and a modern checkout path were harder to evolve safely inside WordPress alone.

What we rebuilt

The store moved into this monorepo:

Editors manage catalog and editorial pages in CMS; commerce events (order placed, fulfillment created) flow through the same shared commerce platform as sibling storefronts in the monorepo. The work matches our website migration and legacy modernization playbooks: inventory URLs, redirect discipline, and parity checks before traffic switches—not a theme refresh on the old host.

What had to keep working

An ecommerce cutover that loses checkout, tax, or feed integrity is a failed launch. This rebuild preserved:

Tracking and ads configuration were re-tested on staging; details mirror what we document on methodology—human sign-off before DNS, automated checks in CI where this repo provides them.

Results we can share today

How this fits the wider rebuild program

Together with eight large healthcare WordPress rebuilds (case study), this store is the ninth production site PunkSolid cites as rebuilt off WordPress into code the client owns. The ecommerce path proves the same code ownership outcome as multi-thousand-page content sites—only the checkout and feed constraints differ.

Sources

FAQ

Why are there no before-and-after speed numbers?

Nobody recorded PageSpeed scores on the WordPress store before cutover, so we publish only post-cutover lab scores and say how we measured them. Mobile LCP is the open item: hero imagery loads in about 4.9 seconds on a throttled mobile profile, and image work is next on the store's backlog.

Can shoppers still buy on the same domain?

Yes. The public store is [48vproducts.com](https://48vproducts.com/). DNS and redirects moved traffic off the old WordPress host; product URLs and checkout were rebuilt in the monorepo rather than patched inside WordPress.

Who owns the code now?

The storefront and CMS live in the client’s monorepo with full repository access at handoff—same model as our other migrations. See [code ownership](/services/code-ownership/) for how independence works after delivery.

Replatforming a WooCommerce or WordPress store?

Tell us what checkout, feeds, and fulfillment integrations must survive cutover—we map them before DNS moves.

Start a project

Related

Full code ownership at handoff

Every PunkSolid client receives full access to finished code—maintain it in Cursor, hand it to another developer, or stay independent if we part ways.

Service

Healthcare sites rebuilt off WordPress

Eight healthcare provider sites of 1,000–6,000 pages each, rebuilt in Cursor with their tracking stack. Mobile PageSpeed went from 65–75 to 92–98.

Case study
Start a project