The network effect in accounts payable. Why connected beats best-in-class

Why AP tools have plateaued

Best-in-class AP software has done remarkable work at the task level. Invoice capture that used to take minutes per document now takes seconds. Approval routing that used to live in email and spreadsheets now runs on rules. Sync to Xero, Sage, NetSuite, and the rest of the accounting stack is, for the leading tools, close to seamless.

The ceiling is everything that comes before and after the document. Is this supplier real? Are these the correct bank details, today, not 18 months ago when they were first captured? Is this invoice consistent with how this supplier has billed every other buyer in the last quarter? Will this payment actually reach the right account? None of those questions are answered by anything that lives inside one company's four walls. They are answered by data that, in today's market, no single buyer has.

That is the plateau. AP automation has solved the part of the problem that sits on the buyer's desk. It has not touched the part that sits in the relationship between the buyer and the supplier, because the relationship is not inside the software.

What changes when supplier data is shared, not re-collected

The interesting shift is what happens when supplier identity, bank details, and payment behaviour stop being collected by every buyer in isolation and start being maintained once, by the supplier, and shared across every business they trade with.

Three things change immediately.

Onboarding stops being a project. Today, taking on a new supplier means collecting documents, validating them, keying them into the vendor master, and absorbing the risk that any of that data is wrong or out of date. When the supplier already maintains a verified profile, the buyer connects to it rather than rebuilding it. Days of dwell time on onboarding compress into minutes.

Fraud signals get better, not noisier. Authorised push payment fraud relies on isolation. A bank detail change that arrives by email looks the same to a buyer whether it is legitimate or not, because the buyer has no second source. In a network, that change is attested by the supplier, logged with attribution, and visible to every other buyer in real time. Cross-buyer signals (a supplier's bank country changing twice in a month, a registration address that does not match the account it is paying into) become detectable. Single point of failure becomes multi-point of verification.

Payment certainty becomes a property of the relationship, not the transaction. Once a network sees how a supplier behaves and how a buyer pays, both sides can price risk properly. The supplier knows which buyers pay on time. The buyer knows which suppliers tend to invoice cleanly. That information has been absent from B2B trade for as long as B2B trade has existed, and its absence is one reason the category still feels twenty years behind consumer payments.

The economics of a network

The economics of a network look nothing like the economics of a tool.

In a tool, every new customer is a new instance of the same problem. The vendor sells a workflow, the buyer fits their suppliers into it, and the value plateau is reached on day one. The hundredth customer experiences roughly what the first customer experienced. Onboarding their suppliers is still onboarding their suppliers.

In a network, every new buyer reduces the cost-to-verify for every other buyer, because some share of their suppliers are already on the network. Every new supplier increases the value of the network to every buyer, because the universe of pre-verified counterparties grows. Trust signals compound. The hundredth customer experiences something genuinely different from the first, because by the time they arrive, the network has already verified a large share of the suppliers they will ever need to pay.

This is why category positioning matters. An AP tool's defensibility is its UX and its integrations. A network's defensibility is its data and the relationships it represents. Those are different categories of asset, with very different ceilings.

What this looks like on Accounting Links

In practical terms, this is the architecture Accounting Links is built on. Suppliers maintain a verified profile, including company registration, banking, and authorised contacts, once. Buyers connect to those profiles rather than re-collecting the same information. Bank detail changes are attested by the supplier and propagated across every connected buyer in real time, with an audit trail. Payment behaviour, on both sides, becomes part of the shared record.

For a finance leader, that means three things on day one. New suppliers go from a multi-day onboarding cycle to a connected session that takes minutes. The supplier master stops drifting out of date the moment it is captured. And the team gains the kind of payment behaviour signal that traditional AP tools, by design, cannot produce.

For a fuller view of how the connections actually run, see How It Works and the Supplier Network product page.

The test for buyers

If you are evaluating AP software in 2026, there is one question that separates a tool from a network. Ask the vendor: how does your data improve when another one of your customers joins?

For a closed AP tool, the honest answer is that it does not. A new customer means more revenue for the vendor and another instance of the same workflow. Your supplier master, your fraud signals, your payment behaviour data, all unchanged.

For a network, the answer is concrete. Some share of your suppliers are already verified. New cross-buyer signals improve fraud detection across the base. Payment behaviour data gets denser and more predictive. The product you bought today is a more useful product six months from now, not because the vendor shipped features, but because the network grew.

Best-in-class describes the tool. Connected describes the architecture. For finance leaders thinking past this budget cycle, the second matters more than the first.

‍