Shopify and Vercel are rebuilding Hydrogen as an open, runtime-agnostic foundation. Hydrogen now dep…
Customer Portal Frontend Over CRM and Billing: Self-Service Without a Rewrite
A customer portal frontend over CRM and billing means the login and self-service surface runs as its own frontend layer in front of your CRM, billing, and core systems. The account record, the invoice, and the contract status stay exactly where they are. Only the surface gets rebuilt, without touching CRM or billing.
The problem: customer portals age faster than the backend underneath them
Most customer portals grew over years: a CRM for account data, a billing system for invoices and subscriptions, maybe a ticketing tool for support on top. Each system brings its own UI, and eventually a customer clicks through three different interfaces just to change a contract.
The frontend is usually the oldest link in that chain. It was wired to whatever CRM was in place at the time and rarely touched since, because a rewrite means new login logic, new invoice views, new roles and permissions, all at once. The risk of a full rebuild keeps many teams from touching the portal at all, even once the interface has fallen years behind the rest of the brand.
What a customer portal frontend actually is
A customer portal frontend is the presentation layer customers see and use: login, account overview, invoice history, contract changes, support tickets. It is deliberately separated from the systems holding the data: CRM for customer data, billing for invoicing, an OMS or ticketing tool for process.
That separation is not an academic distinction. It decides whether you ship a rebranding request in days or wait for the next CRM release because the UI lives inside the CRM's own templates.
Architecture: a frontend over CRM and billing, not a replacement for either
The pragmatic path is not "rip out CRM, bring in something new," it is a composable frontend layer that talks to the existing CRM and billing APIs. At Laioutr, Orchestr, our GraphQL layer, does that work, normalizing more than 50 backend systems today (see Composable Headless Frontend). Account data comes from the CRM, invoices and payment status from billing, both land in one unified data model that the portal components consume.
For login, that means the existing session or auth token from CRM or billing gets passed through, not reinvented. The frontend checks it, renders it, redirects on it, but authorization stays exactly where it already lives. That is the difference between a frontend project and an identity migration project, and you did not want to touch the latter anyway.
Self-service without a rewrite: what changes and what doesn't
What changes: the surface gets faster, more consistent, and easier to maintain. Layout, forms, and microcopy in the portal can be adjusted through the platform's content management, without marketing or support opening a developer ticket every time (see Content Management). A new self-service form for address changes becomes a component configuration, not a sprint.
What doesn't change: your data ownership. CRM stays the single source of truth for customer data, billing stays it for invoices and payment runs. Compliance teams don't need to sign off on a new database, because none gets created. The frontend reads and writes through the existing APIs.
Blueprint: four steps to the frontend layer
- Inventory. Which systems feed the portal today: CRM, billing, ticketing, maybe a separate login system? That's the map for the API connections.
- Data model. Normalize account, invoice, and contract data into one frontend schema, regardless of how the CRM names them internally.
- Auth handoff. Define how the session or token moves from the existing system into the new frontend, without building your own identity logic.
- Staged rollout. Start with one area, invoices for example, and migrate area by area while the old portal keeps running for the rest.
This blueprint works for B2B self-service portals as well as classic B2C account areas. See our patterns on B2B self-service frontends and on punchout and tiered pricing in B2B portals, both starting points for the same frontend-over-backend approach.
When it pays off, and when it doesn't
This approach pays off when CRM and billing are functionally fine but the surface is holding growth back, or when you want to bring several systems together under one login experience. It doesn't pay off when the real problem sits inside the CRM itself: outdated data structure, no usable API. Laioutr is built for the frontend layer, not as a CRM or billing replacement. If your CRM genuinely needs replacing, that's a different project, and it should get solved first.
For B2B-heavy self-service scenarios with punchout, tiered pricing, or quick order, it's worth a look at our B2B Growth Kit, which ships exactly these patterns as production-ready components.
FAQ
Do we need to replace our CRM or billing system? No. The approach deliberately keeps CRM and billing as the backend and only rebuilds the frontend layer.
How does login work if CRM and billing have separate logins? The frontend passes through the existing session or token instead of building its own identity solution. Where two systems have separate logins, the rollout decides which one leads.
How fast can a first area go live? That depends on the API state of CRM and billing. A realistic first self-service area, invoices for example, ships well ahead of a full portal rewrite, because no backend switch sits in the timeline.
More on the customer portal frontend
The full solution overview for self-service portals over CRM and billing lives on our Portals and Self-Service solution page.