PayShap Request to Pay is not trying to replace every payment method in the country. It offers something more practical and, for SMEs, more useful: it removes the awkward middle step between agreeing to pay and proving that the money has landed. For low-value collections, this is a real win. The South African Reserve Bank’s latest payment oversight report records 218,991 Request to Pay transactions worth R74.2 million, which puts the average request at about R339 and shows where the service is already finding a home.
That number tells you almost everything about the product’s current shape. This is not a corporate treasury tool or a theoretical bank feature waiting for a use case. It is being used in the messy part of business life, where a tutor wants a deposit, a courier wants payment before dispatch, a mechanic wants the balance after the job is done, and an informal retailer wants a fast way to collect without babysitting an EFT reference that may arrive with the wrong name attached to it.
The quiet appeal is control
Traditional collections still lean on a brittle routine. A business sends bank details. The customer types them in, usually on a phone, often while distracted. A reference gets entered badly or not at all. Then someone on the business side checks a statement, opens an inbox, looks for a proof of payment screenshot, and tries to match a deposit to an invoice or an order number.
PayShap Request to Pay removes most of that manual plumbing.
Instead of asking the customer to reconstruct the payment from scratch, the business sends a structured request into the customer’s banking app or online channel. The amount is already there. The reference is already there. The approval happens inside the banking flow, so the customer is not free-styling an account number or copying and pasting details from a WhatsApp chat. For small businesses, this is more important than flashy payment branding ever could be.
The practical gain is simple: fewer wrong references, fewer delayed reconciliations, fewer messages asking whether the transfer has gone through. Businesses spend less time matching a bank line item to a sale that already happened.
The Reserve Bank’s figures support the idea that this is being used where it should be used. An average request of about R339 points to ordinary collections, not chunky once-off transfers. This is the sweet spot for small jobs, deposits, service fees and the kind of balances that create unnecessary admin when paid through a standard EFT.
What the test shows
The real test is not whether PayShap exists. It is whether it behaves cleanly across banks when a business is trying to get paid for a R350 transaction.
Across participating banks such as Absa, FNB, Nedbank, Standard Bank, Capitec, Investec, TymeBank and African Bank, the core flow is broadly the same. The business creates a request from its banking app or online banking. It identifies the customer through a ShapID, often a cell number or ID number, or through bank account details. It sets the amount and reference. The customer gets the request inside their own banking channel, checks the details and authorises it with a PIN, fingerprint or face scan.
Settlement is the strongest part of the experience. The money lands instantly, around the clock, every day of the year. This is already better than ordinary EFT, which still leaves room for settlement lag, banking cut-off frustrations and the ritual of waiting for proof before work can continue.
Fees are more mixed, but still useful for small values. The research pack points to fixed charges in the R5 to R10 range for the sender, which is usually the business. FNB’s charge for a PayShap payment is R6.50. Capitec charges R5. Those numbers are not tiny in absolute terms, but on a R350 payment they are far easier to reason about than a percentage-based fee that starts looking silly as soon as the transaction value drops.
The limits also fit the use case. PayShap as a broader rail can go much higher, with an overall transaction ceiling of R100,000, but Request to Pay itself is aimed lower, generally in the R3,000 to R5,000 band per transaction depending on the bank, with daily limits varying by provider. That is enough for deposits, small invoices and service work. It is not built for every business payment, nor should it pretend to be.
The comparison with other options is blunt
For a R350 payment, the alternatives each have a different weakness.
Card links are familiar and easy for customers who already live online, but they usually come with gateway fees that are more expensive than a flat bank-based request when the ticket size is small. They are useful when the customer expects a card checkout experience, but they are not naturally efficient for a plumber, a tutor or a trader collecting a modest balance after a job.
Instant EFT closes part of the gap by speeding up bank-to-bank payments, but it still depends on the customer leaving one platform, entering details, authenticating the transfer and hoping the reference survives intact. It is better than ordinary EFT, but it still leaves room for user error and still needs some form of confirmation handling on the business side.
Ordinary EFT is the slowest and the messiest of the three for small collections. It is cheap in a narrow sense, but the administrative cost of chasing payments, checking references and reconciling deposits can dwarf the bank fee itself. Many SMEs do not lose money on the transfer. They lose time, and time is the thing they are least able to waste.
PayShap Request to Pay is strongest when the business wants a clean, bank-to-bank request with an instant result and minimal back-and-forth. It is weaker when the customer does not understand what the request is, or when the bank interface makes the request too easy to miss or too awkward to approve. That gap is the biggest drag on adoption. A tool can be elegant on paper and still fail if customers do not recognise the prompt quickly enough.
Where it already fits
The service suits businesses that collect small amounts regularly and need a tighter link between invoice and payment.
That includes:
- Delivery businesses collecting payment on dispatch or on completion
- Tradespeople asking for deposits or final balances
- Tutors and coaches collecting session fees
- Informal retailers chasing small outstanding amounts
- Service providers who need a payment record that maps cleanly to a job or invoice
For these users, the value is not just speed. It is the removal of friction at the exact point where friction usually causes problems. A structured request is less likely to be misread than a text message with bank details. It is less likely to be paid with the wrong reference than a manual transfer. It is easier to match in the books because the payment trail starts with the business, not with a customer trying to remember what to type.
The limits are still real
PayShap Request to Pay is not a universal answer, and pretending otherwise would be lazy.
Customer awareness is still uneven. If someone does not know what the notification means, the request loses momentum immediately. This is not a product flaw in the narrow sense, but it is a practical constraint for any business that serves a broad customer base.
Bank interfaces are the other bottleneck. Some apps make the request obvious and quick to approve. Others bury the experience inside layers that make the customer stop and think twice. Small businesses do not care which institution has the neatest product deck. They care whether the customer can pay without calling back to ask how it works.
Then there are the hard limits of the rail itself. Request to Pay is designed for low-value collections, which is exactly where it shines. It is not the first choice for large invoices, and it is not a substitute for every checkout or invoicing workflow. Businesses that need card acceptance, broader consumer familiarity or higher-value flexibility will still keep other rails in the mix.
The verdict for SMEs
PayShap Request to Pay is already useful, and the Reserve Bank’s latest numbers suggest it is finding the right job. The average transaction value sits in the low hundreds, where admin pain is high and payment mistakes are common. This is the territory where a structured request can save real time.
For South African SMEs, the service works best as a collections tool, not a universal payment strategy. It reduces reference errors, shortens the gap between approval and settlement, and makes reconciliation cleaner than ordinary EFT. On a R350 payment, the economics are straightforward enough. Flat fees around R5 to R10 are easier to swallow than a percentage fee, especially when the alternative is chasing proof of payment and fixing a bad reference after the fact.
The catch is adoption. If customers do not recognise the request quickly, or if a bank’s interface turns the approval into a small obstacle course, the advantage shrinks. That leaves PayShap in a sensible but imperfect position. For low-value collections, it is one of the cleaner tools available. For everything else, businesses will still need a mix of rails and a clear sense of when each one earns its keep.
