Most finance teams treat supplier onboarding as an admin problem. It is actually a trust problem, and the way the industry solves it has not meaningfully changed in twenty years. Automation has made the existing model faster, not better. The replacement is not a better form. It is a different architecture, where verification becomes shared, continuous, and portable across the businesses suppliers trade with.
Every time a business takes on a new supplier, someone collects the same set of documents, types the same details into another system, and verifies the same bank account that ten other buyers have already verified. None of that work carries forward. The supplier fills out the same forms for every buyer. The buyer takes on the same fraud risk every time. The verification expires the moment it is done.
That is not onboarding. That is duplication, dressed up as due diligence.
Why the current model fails
Supplier onboarding sits inside each company's accounts payable system as a closed workflow. The buyer collects bank details, company registration, VAT numbers, and proof of identity. Someone in finance reviews them, ticks a box, and the supplier is added to the master vendor file. From that moment on, the trust decision is frozen in time.
This causes three problems that show up in the data.
First, fraud. Authorised push payment fraud against UK businesses ran into hundreds of millions of pounds in recent years, and a meaningful share of that traces back to compromised supplier records, social engineering at onboarding, or bank details quietly changed mid-relationship. Once a supplier is in the system, most AP processes assume they remain who they said they were. They often do not.
Second, friction. Suppliers, particularly smaller ones, can spend weeks bouncing between buyers' onboarding portals, each demanding the same evidence in slightly different formats. The cost of being a supplier has gone up, and the businesses paying that cost are often the ones least equipped to absorb it.
Third, blindness. The buyer has no view of how that supplier behaves elsewhere. Are they paid on time by other businesses? Have their bank details changed in the last 90 days? Have they been flagged at another company for suspicious behaviour? In a closed onboarding model, none of that information is available, because none of it is shared.
What automation alone does not solve
A lot of the response to this has been to bolt automation onto the existing model. OCR pulls data off invoices. RPA fills in vendor master files. AI verifies a passport image. These are real improvements at the task level.
They do not change the underlying architecture. A faster way to do the wrong thing is still the wrong thing. As long as every buyer maintains a private, static record of every supplier, the duplication, the fraud surface, and the blindness all remain.
The replacement is a network
Supplier onboarding stops being broken when verification becomes shared, continuous, and portable. That is what a network changes.
In a connected accounts payable network, a supplier is verified once, and that verified identity, including company registration, banking details, and authorised contacts, becomes available to any buyer they trade with, with the supplier's permission. When the supplier's bank details change, every connected buyer sees the change in real time, attested by the supplier themselves. When a buyer pays an invoice, that payment behaviour becomes part of the supplier's profile, visible across the network.
This is the difference between an isolated record and a living relationship. The supplier maintains their own verified profile, the way a business maintains its Companies House record. Buyers connect to that profile rather than rebuilding it from scratch. Trust accrues to the supplier, not to each buyer's spreadsheet.
What changes in practice
Three things become possible that were not possible before.
Onboarding collapses from weeks to minutes. A supplier already on the network is not onboarded again. The buyer connects to the existing verified profile, agrees commercial terms, and is ready to transact. The hours finance teams spent rekeying are returned to the business.
Fraud risk drops. Bank detail changes that previously slipped through email-based social engineering become network events with attribution and an audit trail. Mismatches between business registration country and bank account country are flagged automatically. Suspicious activity at one buyer is visible to others, the same way card fraud signals are shared across the payments industry.
Payment behaviour becomes legible. Buyers can see whether a supplier is reliable. Suppliers can see how a buyer pays. Both can make better decisions, and the market gets the kind of payment behaviour data that has been missing from B2B trade for as long as B2B trade has existed.
A different kind of category
The companies still describing themselves as accounts payable software are competing for share of a shrinking problem. Closed-system AP works at the level of one company processing its own invoices. That category is mature, crowded, and increasingly commoditised.
The shift underway is from accounts payable as an internal workflow to accounts payable as connected infrastructure. The unit of value moves from the document to the relationship. The thing being automated is not the invoice. It is the trust between two businesses that sits underneath the invoice.
Supplier onboarding does not get fixed by a better form. It gets replaced by a network where the form was never the answer in the first place.
.jpg)
.jpg)