Most companies already have a way to work with their suppliers. The problem is the way. A purchase order goes out by email, the vendor confirms in a WhatsApp message, then a PDF invoice arrives at a different address. Someone in your office retypes the quantities into a spreadsheet. When a delivery is short, the only history is a thread nobody can find.
A vendor portal (also called a supplier portal) is a signed-in place for the companies that sell to you. Each vendor logs in and sees only their orders, their catalogue, their invoices and the status of what they are owed. Your team gets one queue instead of a pile of attachments.
It is not a client portal, and it is not a B2B ordering portal for your own customers. Those are useful tools, and I have written about both. This post is the other direction: inbound supply.
Who this kind of portal is for
The pattern fits when you buy repeatedly from other businesses, and the volume has outgrown inboxes:
- Distributors and wholesalers who purchase from many manufacturers or importers
- Retailers or restaurant groups that reorder from a fixed set of suppliers
- Manufacturers who buy components or packaging on recurring terms
- Facility and maintenance companies that order parts from regular vendors
- Any team that is retyping vendor invoices or chasing confirmations every week
If you only have two or three suppliers and a handful of orders a month, a shared mailbox and a careful spreadsheet may still be enough. A portal starts to earn its place when the same questions keep coming back: did they confirm, has it shipped, and is this invoice already in the system.
How it differs from a client portal or a trade ordering portal
A client portal is for the people who pay you. They log in to see files, invoices you issued, and the status of the work.
A B2B ordering portal for distributors and wholesalers is for your trade customers. They log in to see their prices and place orders with you.
A vendor portal turns that around. You are the buyer. You publish the purchase order. The vendor confirms, ships and invoices you. The records are purchase orders, inbound deliveries and payables, not sales orders and receivables. If you tried to stretch a customer portal or an online store to cover this, you would spend the project fighting the language of the screens.
Some companies eventually need both a customer side and a vendor side. They can share item records and one admin desk, but they should still feel like two products, with two logins and two permission models.
What the first version should include
I plan a first version around the weekly work, not around a full procurement suite.
A real login, per vendor company
Each vendor company gets an account, and one or more users under that account. Your staff get a different kind of login, with roles that match the job: someone who raises purchase orders, someone who records what arrived, someone who handles invoices, and someone who can do all of it.
A vendor should never see another vendor's prices, orders or invoices. That sounds obvious, and it is the requirement that has to be designed in from the first screen, not added later.
Keep sign-in boring and solid: saved sessions, a password reset, and a way to disable a user who has left the vendor's team.
A vendor profile you can actually use
Name, contacts, the items or categories they supply, payment and delivery notes you have agreed, and a place for the documents you already collect, such as a contract or a certificate. I do not invent a legal checklist; we put in the documents your process already requires.
This profile is also where you record whether the vendor is approved to receive new orders, or on hold.
The items they supply
A vendor-facing list of the products or materials they are expected to fulfil: your item code, their item code if it differs, units, and the usual lead time. Prices can be fixed, dated, or specific to a contract. If different vendors can supply the same item, the item belongs to you, and the vendor link holds their code and their price.
Do not start with a public catalogue. Most of this information is commercial. It stays behind the login.
Purchase orders
This is the core. Your team creates a purchase order, or issues one that originated in your accounting system. The vendor sees it, confirms it or flags a problem (a wrong quantity, an item they cannot fulfil, a later date), and both sides look at the same document from then on.
A good purchase-order screen shows status clearly: draft, sent, confirmed, partly received, closed, cancelled. Line-level status matters when one item arrives and another does not.
Deliveries and receiving
The vendor should be able to say what has shipped, or at least see that you have recorded a receipt. Your team needs a simple way to mark what arrived, what was short and what was rejected, against the original lines.
You do not need a full warehouse system in version one. You need the fact of receipt tied to the purchase order, so invoice matching has something to stand on.
Invoices and payment status
Vendors want to know whether you have their invoice and whether it has been paid. Your finance team wants invoices that already point at a purchase order, not a PDF with a different quantity and no reference.
In the first version I usually include: the vendor submits or attaches an invoice against a purchase order, your team accepts or queries it, and the vendor can see a simple payment status (received, queried, scheduled, paid). The portal is not a bank. It records status. When a payment method comes up in the wider project, the ones I work with are bank transfer, Whish or OMT.
A narrow message thread
A short thread on the purchase order or the invoice is enough. It beats a WhatsApp group because it stays attached to the record. It does not need to become a complete helpdesk in version one.
What can wait
A lot of procurement software is sold as a complete suite. Most of it can wait until the core above is in daily use:
- Formal tendering, auctions or multi-round requests for quotes
- Vendor scorecards and automatic ranking
- Deep electronic-document connections, unless a large vendor already requires one
- Forecasting and automatic replenishment
- Complex approval chains with many branches (a simple approve or reject on the purchase order is usually enough at the start)
- A vendor-facing analytics dashboard
If your process truly cannot raise a purchase order without a three-step approval, we put a small approval step in version one. We do not build the rest of a procurement platform around that one need.
How it should connect to the systems you already have
The portal should not become a second set of books. The usual connections are:
- Items and vendors, already stored in an accounting or ERP system, imported and kept in sync so codes do not drift.
- Purchase orders raised in that system and published to the portal, or raised in the portal and sent back.
- Receipts and invoices returned to finance so payables are not typed in twice.
I design the API and the data model so those connections are possible from the start, even if the first release is a careful import plus a weekly export. Forcing a live two-way sync on day one is a common way to stall the project. We go live with a flow the team can trust, then tighten the sync.
How I build it
This is custom CRM and portal development. A vendor portal is one of the portal types I scope from the workflow: a login, the records each vendor is allowed to see, and nothing else.
The usual shape is a web app the vendor and your staff use in a browser, built in Next.js or with a .NET Core API, a Postgres database, and an admin desk for your team. It is deployed with Docker on a Linux server, behind a CI/CD pipeline, with backups and monitoring. If your buyers or receivers spend their day on a warehouse floor, we can add a mobile app later on the same backend. We do not start there unless the job clearly needs a phone.
NeonTask is not a vendor portal, and I will not pretend it is. It is the public example of a signed-in operational system I designed and built: a field-service platform for facility-maintenance companies, with a web app for dispatchers and admins covering intake, queues, a list and map, the schedule, receivables and reports. The same approach applies here: watch the real workflow, build the core screens around it, and give each person only the part their job needs.
Enchanté CRM is a custom CRM I designed and built in 2025. There is no public feature breakdown for that project, so I will not invent one.
What to prepare before we start
The best briefing material is what you already use:
- A recent week of purchase orders, including messy ones
- The spreadsheet or system you use for vendors and items
- A typical invoice and a typical query
- Who confirms orders, who receives goods, and who handles invoices, on your side and on the vendor side
- Whether Arabic, English or both are needed on the vendor screens
- Any system the portal has to connect to
If you want existing vendors and open purchase orders in the new system on day one, we plan the import as part of the first version, including the messy records, not only the clean ones. The same discipline I use when moving off a spreadsheet applies: clean what you can, keep a record of what you changed, and never go live with two competing lists.
You own the code after the final payment, and you can take the vendor records with you. That is one of the reasons to build this rather than rent a generic procurement tool that will not match the day.
Getting started
Email hello@eliechaaban.com, or use the contact section, with a short description of how you buy today and what is breaking down. We can go through a real week of orders on a call and agree a first version your vendors will actually log into.