What BNPL Actually Costs You: A Merchant’s Comparison of SA’s Buy Now, Pay Later Providers

Most South African eCommerce operators can tell you which buy now, pay later providers sit on their checkout. Very few can tell you what each one costs them per transaction, or whether the provider they chose fits what they actually sell. BNPL merchant fees in South Africa are seldom published, rarely compared, and almost never tested against basket size, which is how merchants end up paying premium rates on a poor fit.

That gap matters more than it used to. BNPL doubled its share of total SA payment volume from 3% in 2024 to 6% in 2025, and it is now the single most-requested new payment method, with 34% of consumers asking merchants to add it. The demand is real. So are the costs, and they sit meaningfully above what you pay to process a card.

This article does what most BNPL content avoids: it puts actual, published merchant costs side by side, works out what those costs mean in rands on a typical basket, and gives you a defensible way to decide which provider fits your business.

A note on sourcing first. Several providers quote merchant rates privately rather than publishing them. Where a rate is published by a payment gateway, it is cited. Where it is not, that is stated plainly rather than guessed at. Vendor-reported performance claims are flagged as such.

Table of Contents

1. The three business models hiding under one label
2. What each provider actually costs a merchant
3. What that costs you in rands on a real basket
4. The incrementality question that decides everything
5. Which provider fits which type of business
6. The costs that are not in the fee
7. The regulatory picture worth watching
8. How to decide

The three business models hiding under one label

Before comparing costs, it is worth understanding that “buy now, pay later” in South Africa covers products that work in fundamentally different ways. Comparing their fees without understanding this leads to bad decisions.

Credit-style BNPL. The customer takes the goods immediately and pays in instalments afterwards. The provider settles you upfront and carries the collection risk. PayJustNow, Payflex, MoreTyme and Happy Pay all work this way. This is what most people mean by BNPL.

Card-linked instalments. Float pioneered this locally. It does not issue new credit. It sits on top of the customer’s existing Visa or Mastercard credit limit and splits the purchase into interest-free monthly instalments. The customer gets the goods now and keeps their card rewards.

Digital lay-by. LayUp’s model, and the outlier. The customer pays off the purchase in interest-free instalments but only receives the product once it is fully paid. No credit, no credit check, and effectively zero default risk.

Revolving credit. Mobicred is a fourth thing entirely: an actual credit facility backed by RCS and regulated under the National Credit Act. The customer applies, passes an affordability check, and pays interest on their balance. It is not interest-free BNPL and should not be marketed to your customers as though it is.

The model determines who the product serves, which determines whether the fee is worth paying for your specific customer base.

What each provider actually costs a merchant

Here is the picture, drawn from published gateway pricing where it exists.

A few things stand out immediately.

Payflex through Peach Payments runs at 4.85% plus R4.00 per transaction for Growth accounts, with Enterprise accounts negotiating a custom rate typically between 4.0% and 4.85% plus R4.00 based on transaction volume. That volume-based negotiation is worth knowing about. If you are doing meaningful BNPL volume and still on a standard rate, you are leaving margin on the table.

MoreTyme through Payfast is charged at 5.5% plus R2.00, while Mobicred sits at 3.5% and standard Visa and Mastercard processing at 3.2% plus R2.00.

PayJustNow’s published POS pricing through Peach Payments is 5.35% plus R4.80 plus R1.50 per transaction with a R199 monthly fee. Online rates are quoted per merchant, but this gives you a defensible reference point for negotiation.

The Mobicred number is the one most merchants miss. At 3.5%, it is barely above card processing. That is because the economics are different: the customer pays interest on their revolving balance, so the provider is not funding an interest-free period out of your merchant fee. If you sell considered, higher-value goods and your customers are comfortable with a credit facility, Mobicred is by some distance the cheapest way to offer instalments.

What that costs you in rands on a real basket

Percentages are abstract. Rands are not.

The average BNPL basket size in South Africa is around R1,568, meaningfully higher than card transactions. Run each rate against that basket:

So on a typical BNPL order, you are paying somewhere between R28 and R38 more than you would if that same customer had paid by card.

Now put that against margin. On a 25% contribution margin, that R1,568 basket generates R392 in gross contribution. An extra R38 in fees is roughly 9.7% of your contribution on that order, gone.

Stated that way, BNPL sounds expensive. But that framing is incomplete, and it is where most merchants stop thinking.

The incrementality question that decides everything

The extra fee only matters if the customer would have bought anyway.

If someone was always going to complete that R1,568 purchase and simply chose the BNPL button over the card button, you have paid R38 for nothing. That is cannibalisation. If someone bought because the instalment option made the purchase feel affordable, that order did not exist before, and its full contribution is incremental.

So the real question is: what proportion of your BNPL orders need to be genuinely incremental for the programme to pay for itself?

The arithmetic is more forgiving than most operators assume.

On a R1,568 basket at 25% contribution margin, each genuinely incremental BNPL order nets you R392 in contribution minus R90 in fees, so roughly R302. Each cannibalised order costs you R38 in unnecessary fee. Break even when:

incremental orders × R302 = cannibalised orders × R38

That resolves to roughly 11%. If just over one in ten of your BNPL orders is genuinely incremental, the whole programme has paid for itself. Everything above that is profit.

That threshold moves with your margin:

The strategic read is straightforward. If you run healthy margins, BNPL is very likely paying for itself even with substantial cannibalisation. If you run thin margins on low-value goods, the bar is genuinely high and you need to measure rather than assume.

The practical version of this: pull your BNPL average order value and your card average order value side by side, monthly. If BNPL volume is growing but total orders are flat and BNPL AOV is not meaningfully higher than card AOV, you are most likely funding a payment preference rather than buying incremental revenue.

Which provider fits which type of business

Fee is only half the decision. Fit is the other half, and fit is determined by basket size, category, and who your customer is.

Payflex: lower-value, impulse-led categories

The customer pays 25% upfront and the remaining 75% in three instalments over six weeks, interest-free. It suits smaller, more impulsive baskets in fashion, beauty and accessories, where a six-week, four-payment structure feels light.

Best fit: AOV between roughly R500 and R2,000. Fashion, beauty, accessories, supplements, homeware. High purchase frequency, lower consideration.

Cost read: At 4.85% plus R4.00, it is mid-range. The fixed R4.00 component hurts proportionally more on small baskets, so below about R500 the effective rate climbs meaningfully.

PayJustNow: broad reach and brand recognition

The most widely recognised BNPL brand locally. Core product is Pay in 3, with a Pay in 12 option for larger purchases that does carry interest for the customer. It gives merchants access to a consumer base of over four million. The Shoprite and Checkers rollout in early 2026 has grown its reach further.

Best fit: General retail with AOV between roughly R800 and R4,000. Businesses where checkout hesitation from an unfamiliar payment logo is a real risk. Mass-market positioning.

Cost read: The most expensive of the published options. You are paying a premium for consumer recognition and network reach. Whether that premium is worth it depends on how much of your traffic is cold versus returning.

Float: big-ticket, credit-card customers

The one most merchants overlook, and arguably the most interesting. Float splits any purchase into up to 24 interest-free monthly instalments using the credit the customer already has on their existing Visa or Mastercard. The average order value on Float is around R10,000, almost ten times the average reported by regular BNPL. The top three categories are electronics at 29%, furniture and home decor at 22%, and appliances and gadgets at 19%.

South Africa has around 5.5 million credit cardholders carrying roughly R200 billion in outstanding balances, which is the pool Float draws from.

Float reports a 134% average uplift in average order value and 20% to 30% higher conversion rates across its merchant base. Those are vendor-reported figures and should be treated as directional rather than guaranteed, but the mechanism behind them is credible: removing the ceiling on what a checkout can process for a customer with existing credit.

Best fit: AOV above R5,000. Electronics, furniture, appliances, sports equipment, automotive, healthcare. Anywhere a single payment is the main reason for cart abandonment.

Cost read: Not published, and configurable. Float can be fully configured on settlement model, number of instalments offered, and channel. That configurability means the fee is genuinely negotiable against how aggressive an instalment offer you want to make. Ask for the rate at 6, 12 and 24 months separately, because they will differ.

MoreTyme: TymeBank reach, mass market

TymeBank’s product, now moving to the GoTyme Bank app. Available via Payfast on Aggregator accounts, with merchants paid instantly and in full. A meaningful footprint among customers the traditional banks have historically underserved.

Best fit: Mass-market general retail, groceries, essentials, mobile and electronics at accessible price points. Businesses targeting customers outside the traditional credit-card demographic.

Cost read: At 5.5% plus R2.00 it is the most expensive published percentage rate. Justifiable if TymeBank’s audience is genuinely additive to your customer base. Hard to justify if it simply gives your existing customers another button to press.

Happy Pay: acquisition-linked, zero deposit

Commercially the most differentiated model in the market. Happy Pay is a zero-deposit BNPL provider integrated with Peach Payments, using an affordability-based approval system rather than relying on credit scores.

The important part for merchants is how it makes money. There are two revenue streams: transaction fees paid by merchants when a sale is completed, and advertising revenue where merchants pay on a cost-per-acquisition basis when the platform drives a purchase.

This means your Happy Pay cost is not purely a payment processing line item. Part of it functions as customer acquisition spend. If you evaluate it against your gateway fee alone, you will conclude it is expensive. If you evaluate the CPA component against your blended acquisition cost from paid media, the comparison may look very different. The company reports growing average basket sizes by 190% for merchants on its platform, though as a vendor figure that warrants your own verification.

Best fit: Merchants already on Stitch or Peach who want a genuine acquisition channel rather than just a payment option. Businesses where zero deposit meaningfully lowers the barrier.

Cost read: Ask for the transaction fee and the CPA rate separately, and budget them to different lines. Comparing the blended number to a pure payment rate is comparing two different things.

Mobicred: cheapest instalments, different customer

Mobicred charges a commission of 3.5% per transaction, and this is the only cost to the merchant. The customer pays the cost of the credit. Available through PayU, PayGate, Payfast and Peach Payments.

Best fit: Higher-value considered purchases where your customer is credit-active and comfortable with a revolving facility. AOV above R2,000.

Cost read: By far the cheapest way to offer instalments. The trade-off is a smaller addressable audience, since it requires a credit application and affordability check, and you cannot market it as interest-free because it is not.

LayUp: no credit, zero risk, delayed fulfilment

Interest-free instalments with no credit and no credit check, where the customer receives the goods only once they have paid in full. Because nobody is extending credit, there is effectively zero default risk, and it reaches customers with no credit card at all, including cash payers via a network of physical payment points.

LayUp is PASA approved as a Third Party Payment Provider through Bankserv, and offers three different settlement structures. LayUp reports completion rates increasing by up to 60% and average order values increasing by up to 25%, again vendor-reported.

Best fit: Big-ticket items, pre-orders, made-to-order goods, event tickets, and any category where delayed fulfilment is acceptable or already the norm. Also the only workable option for customers with no card and no credit access.

Cost read: Not published. The critical commercial variable is not the fee but the settlement structure. Because you receive funds as instalments are collected rather than upfront, LayUp affects your working capital position differently to every other option here. Model the cash flow impact before the fee.

What is the cheapest BNPL option for South African merchants?

Mobicred is the cheapest instalment option for South African merchants at a published rate of 3.5% per transaction, only marginally above standard card processing, because the customer pays the interest on their revolving credit balance rather than the merchant funding an interest-free period. Among interest-free BNPL providers, Payflex is generally the most affordable published option at 4.85% plus R4.00 per transaction for standard accounts, with volume-based negotiation available for higher-turnover merchants.

The costs that are not in the fee

Three operational costs sit outside the headline rate and are routinely missed at the evaluation stage.

Return handling complexity. Refunding a BNPL order means unwinding a payment plan, not reversing a single transaction. Some providers handle this cleanly with instant wallet refunds. Others create a reconciliation exercise for your finance team. Ask specifically how refunds and partial refunds are processed before you sign.

Settlement timing and working capital. Most credit-style providers settle you within 24 hours to a few days. LayUp, by design, settles as instalments are collected. If you are in a cash-tight growth phase, that difference is more material than a percentage point of fee.

Reconciliation overhead. Every additional provider adds a settlement file, a fee structure and a reconciliation process. Two providers is manageable. Six is a part-time job for someone on your team, and that cost never appears on a rate card.

This is the strongest practical argument against stacking providers. Two well-chosen options that match your basket profile will outperform six logos in a row at checkout, both commercially and operationally.

The regulatory picture worth watching

South African BNPL currently sits in a regulatory grey area. Providers often argue they are not credit providers because they charge no interest and operate within a short payment cycle, and therefore claim exemption from National Credit Act obligations. The Intergovernmental Fintech Working Group has described BNPL as falling into a regulatory void, with the National Credit Regulator taking limited action and the Financial Sector Conduct Authority yet to issue clear guidance.

That is changing internationally. The UK’s Financial Conduct Authority will regulate BNPL under new legislation taking effect in 2026, requiring affordability checks and FCA authorisation. Australia has brought BNPL under its National Consumer Credit Protection Act, and in the US the Consumer Financial Protection Bureau has classified BNPL loans accessed via digital accounts as credit cards.

For a merchant this is not an immediate compliance issue, because the regulatory obligation sits with the provider. It is a supplier risk issue. If local regulation tightens, providers with weaker capitalisation or looser approval processes will face the most disruption. Choosing well-backed providers with transparent affordability processes is a reasonable hedge, and worth weighting in your decision alongside the fee.

How to decide

A workable process, in order:

Start with your average order value. Below R2,000, look at Payflex or PayJustNow. Between R2,000 and R5,000, PayJustNow or Mobicred. Above R5,000, Float first, with LayUp as the no-credit alternative.

Check your margin against the incrementality table. If you are above 25% contribution margin, the bar is low enough that you can move without extensive analysis. Below 15%, measure before committing.

Ask for the rate in writing, and negotiate it. Volume-based pricing exists at most providers. If you are doing real volume on a standard rate, you are overpaying by default rather than by decision.

Pick two, not six. One recognised credit-style option matched to your basket, plus one that covers a customer segment the first one misses.

Then move the instalment amount to the product page. This is the part most operators get wrong. An instalment option displayed only at checkout is discovered after the affordability decision has already been made. Displayed near the price, it reframes the number the customer is evaluating. Same money, different mental arithmetic, and the arithmetic happens on the product page.

That placement decision is conversion rate optimisation work, not payments work, and it is usually worth more than the difference between any two providers on this list.

Then measure. Compare BNPL average order value against card average order value monthly. Growing BNPL volume on its own tells you nothing. Growing BNPL volume alongside a higher BNPL AOV and a stable or growing total order count tells you the fee is buying something real.

Which BNPL provider is best for a South African online store?

The best BNPL provider depends primarily on average order value. For baskets under R2,000, Payflex suits impulse-led categories like fashion and beauty, while PayJustNow offers the broadest consumer recognition. For big-ticket items above R5,000, Float’s card-linked model carries a significantly higher average order value and no cap on transaction size. Mobicred is the cheapest instalment option at 3.5% but requires a credit application. LayUp is the only viable option for customers without credit access, though goods are released only after full payment. Most stores are best served by two well-matched providers rather than a crowded checkout.

Not sure whether your BNPL setup is earning its fee?

We work with South African eCommerce brands on checkout and conversion strategy, including whether your payment mix is buying incremental revenue or quietly funding a preference. Book a free 30-minute call and we will look at your numbers with you.

Book your free discovery call

Written by the Saleleni team. Saleleni is a Johannesburg-based digital and AI consultancy helping mid-market eCommerce brands convert more, sell smarter, and prepare for the autonomous commerce era.

Merchant rates cited are those published by payment gateways at the time of writing and are exclusive of VAT. Rates are subject to change and to negotiation based on volume. Verify current pricing directly with your provider before making a commercial decision.