Forzabuilt — Full-Stack B2B Platform Migration

Forzabuilt is a B2B industrial adhesives manufacturer serving Florida and the Southeast construction market. When they rebranded and outgrew WordPress, they needed a full platform transition — not a reskin — without the indexing collapse that kills most B2B migrations.

The problem

The legacy site carried three risks at once:

Performance debt — media-heavy product pages with poor Core Web Vitals on a catalog buyers actually search by specification.
URL sprawl — years of WordPress paths that procurement teams had bookmarked, linked, and indexed.
Wrong information architecture — products organized for marketing, not how adhesive buyers search (application, substrate, compliance).

A failed migration would mean lost organic equity right when the rebrand needed inbound to work.

My role

I owned the platform end to end: Astro/React frontend on Vercel, serverless proxy routes to a Heroku Express/PostgreSQL API, product/catalog admin ops, contact/newsletter lead pipeline with Salesforce sync, migration-safe routing, and internal AI tools for the team.

What I built

Split-stack platform architecture

Vercel/Astro public site — product catalog, blog, industry pages, contact/newsletter forms, and serverless API proxies.
Heroku Express/PostgreSQL backend — authoritative product catalog, blog CMS, contact submissions, newsletter subscribers, images/documents, backups, and audit logs.
Lead pipeline — validation, honeypot, IP-hash rate limiting, Postgres persistence, async Salesforce OAuth lead creation with retry and sync-status tracking.

Migration without the indexing cliff

B2B catalog migrations fail when redirect coverage is an afterthought. I treated URL continuity as a first-class deliverable:

Canonical route structure replacing legacy WordPress patterns across 200+ product pages.
Layered redirect rules — middleware plus edge configuration — preserving equity on high-traffic legacy paths.
Pre-launch crawl/index validation so we could ship the rebrand and platform change in one transition.

B2B-native catalog architecture

Industrial buyers do not browse hero banners. I restructured IA around procurement intent:

Category and product pages organized by application, substrate, and specification.
Product and FAQ schema across the catalog for rich results.
Downloadable spec paths and comparison-friendly layouts for technical buyers.

Production-grade performance

On a media-heavy catalog, I sustained 99% Lighthouse desktop through:

Selective React hydration (islands) instead of shipping a full SPA.
Image optimization and edge delivery on Vercel.
Code splitting so category pages stay fast under real catalog weight.

Internal AI tools

Meeting Master — voice brain-dump → Claude structured extraction → Microsoft Graph calendar match → Neon Postgres → email reminders.
Manager Hub — Claude feedback generator from manager notes with DiSC/Crystal tone adaptation and role-based access.

Outcomes

Metric	Result
Lighthouse desktop	99% sustained
Technical SEO health	99% post-remediation
Qualified inbound	28 qualified inbound submissions (26 unique contacts) in first four months
Baseline	Effectively zero qualified web inbound pre-rebuild

Leadership credited the rebuild with the best single website contact in company history — the kind of inbound that validates both the technical migration and the procurement-focused IA.

What this demonstrates

Full-stack platform ownership: frontend, backend API, database, CRM integration, and production ops.
Shipping a catalog migration on 200+ SKUs without the visibility loss common in B2B rebuilds.
Layered redirect strategy when URL patterns change materially.
Practical AI tooling for internal workflows — not public-site chat hype.