The quagmire of B2B payments
Every company must pay its suppliers and collect from its customers. These B2B payments are a massive portion of the global economy, accounting for $125 trillion annually. It is one of the largest opportunities for innovation in any market, as the processes used are archaic, highly manual and paper-based, requiring significant time and resources to manage and causing many delays and inefficiencies.
There is a lack of standardization and transparency, and the processes are strewn with errors, delays and difficulties in tracking and reconciling payments.
This morass of inefficiency ties up significant amounts of working capital. Goldman Sachs estimates that businesses spend nearly $3 trillion each year on manual, paper-based payment processing, a huge burden in terms of time and money. The bank sees a fresh $1 trillion revenue opportunity in payments & software from solving these problems, in addition to the scope for freeing up working capital, driving substantial cost savings and boosting the global economy.
What are the worst inefficiencies in B2B payments?
The biggest delay and friction in B2B payments is not the speed with which money moves – it is getting the “okay to pay” signal from the Buyer (i.e. approving the invoice).
In addition to payment terms, money doesn’t flow until the buyer has accepted the goods or services, and so the seller is at the mercy of the buyer’s approval processes and must wait and chase. Approval typically takes 10 to 20 days but can be much longer.
So regardless of payment method or terms, there is a minimum multi-week delay in all B2B payments where the buyer is a corporate, simply because of the approval process.
With this perspective, the focus on being able to move money for a B2B payment at sub-second speed is misplaced when it takes weeks just to get the “ok to pay” signal from the Buyer. This is the biggest friction in B2B payments. And it is a perfect use case for AI.
Couldn’t credit cards help – why have they so far failed to penetrate B2B payments meaningfully?
Credit cards have delivered terrific benefits and efficiencies in the B2C market, removing the frictions described above that persist in B2B.
In B2C, payments move between buyers and sellers almost immediately. So, sellers are not subject to waiting and chasing.
Buyers benefit from a 30-day statement period from their bank, after which they can make a single payment for the aggregated payments on their credit card bill. Everything can be digital, with rich data for each transaction. Sellers pay a fee to enable this credit card capability, which is substantially less than the cost they would incur for handling cash.
Why hasn’t this happened in B2B?
Of the $10 trillion of payables of the Fortune 1000 companies, only $200 billion, or 2%, benefits from cards’ capabilities, which can now be deployed with virtual cards rather than plastic cards.
Given the clear benefits for all concerned, why is the penetration of cards in B2B so low?
The answer is that the perceived value proposition of cards is simply not sufficiently attractive for the suppliers, who must pay the fee. Part of the reason for this is the misperception that receiving a check or ACH is free.
But the key reason is that unlike in B2C, the seller does not receive payment acceleration. If the buyer uses a card, the seller still has to wait weeks or months for payment. And so the sellers rightly ask, “Why would I accept a fee when I still have to wait and chase to get paid? I am still dependent on the Buyer’s painful approval processes”.
How does AI change the paradigm?
Just as ChatGPT has brought the power of large language models into common usage, Previse’s “SmartPay” platform brings the power of deep neural networks and reinforcement learning to B2B payments.
Each invoice is analyzed in real-time by powerful AI so that the vast majority can be authorized for prepayment, funded with bank credit. Sellers can be paid automatically within 24 hours of receipt of the invoice — a consumer-grade experience enabled by Previse’s SmartPay platform, which makes B2B card payments operate just like B2C card payments.
At the core of this process is the SmartPay decision engine, which automates the “OK to pay” signal.
The decision engine comprises a suite of algorithms that use machine learning to recognise patterns in spending, approval, dilution and payments in real-time. Each sub-algo is specifically trained on a particular aspect of the problem (supplier scoring, outlier and fraud detection, application of set-offs) using Previse’s spend database.
The algorithms are highly accurate, enabling an extremely high acceptance rate of payments while almost completely eliminating overpayments. Previse’s algorithmic approach to B2B payments is patented in the US.
How do cards become a working capital tool?
The automated “pay-on-day X” functionality is far more attractive to suppliers because they no longer have to wait and chase. As a result, vastly more suppliers want to be paid with a card, and the 2% penetration of cards in B2B is moving to 10% and beyond. Just like in B2C, the Buyer benefits from a card statement period provided by their funding bank, and with meaningful payments going on card, it can now be used as a powerful tool to optimize working capital.
After a decade of historically low interest rates, followed by an unprecedented 500 bps rise in rates in 18 months, optimizing working capital is now a strategic priority for many companies and finance professionals. Adopting AI-powered virtual cards is a powerful tool in the arsenal of forward-thinking Treasurers.
There is no longer any need for a B2B payment to be contingent on the buyer accepting the goods or services. AI has removed the biggest friction in B2B payments.
Buyers benefit from the statement period their bank provides, along with rebates, richer data, easier reconciliations and operational efficiencies.
Sellers benefit from greatly accelerated payments, a digital consumer-grade experience, richer data, easier reconciliations, and operational efficiency.
AI makes B2B payments data-driven, friction-free, and fair for all.