How South Africans shop in 2026: key insights from our annual consumer report

Every year, Stitch speaks with thousands of South African consumers to understand how their shopping and payment behaviours are shifting. This year's report, How South Africans Shop in 2026, examines the full purchase journey of the everyday South African: where they shop, how they decide what to buy, what they expect at checkout and where businesses might be losing them.

To unpack the findings, Jessica Manthey, Head of Sales at Stitch, and Dewald Müller, Director of Client Success, hosted a webinar walking through the data with a practical lens for merchants. Here are the key takeaways.

The methodology behind the numbers

The 2026 report surveyed around 3,000 South African consumers, supplemented by 27 in-depth interviews to verify and contextualise the quantitative findings, and looked at our own transaction data from Q1. It’s important to note that the sample focuses on digitally active consumers, or those who indicated they transact online. 

The payments landscape has reset

For the first time across all four years of this research, there is no standout dominant payment method. Card and cash, long the default, are facing real competition across every income group and demographic.

93.3% of consumers tried a new digital payment method in the past year, and most continued using it. That figure alone shows how quickly the market is moving.

Apple Pay, Google Pay and Samsung Pay together saw 57.5% adoption in the past year. Buy Now Pay Later was adopted by 38.9%. Capitec Pay and PayShap, under the broader Pay by bank umbrella, saw 38.6% adoption.

In Q1 2026, Stitch data showed Apple Pay becoming the single most-used payment method by volume on the Stitch platform specifically for e-commerce businesses, overtaking card. 

Dewald Müller put this into context: "A normal card transaction can take 15 to 20 seconds to complete. Digital wallets get this done in under three seconds. You can expect digital wallets to convert at 92 to 95%, whereas card can be materially lower than that depending on the merchant setup."

The speed difference maps directly to revenue. When e-commerce merchants launched Apple Pay through Stitch, more than 25% of customers chose it immediately, and transaction success rates reached nearly 96%, compared to around 80% for card. Within two months, merchants saw 30% or more of their card transactions shift to Apple Pay.

For merchants still offering card only, this is the gap. Jessica Manthey was direct: "If you are a merchant that has constantly relied on the premise that ‘if I offer card, I'm fine,’ that is no longer the case."

One in four card payments fails before it even reaches the bank

The conversion picture goes deeper than method choice: Stitch platform data from Q1 2026 shows that nearly one in four online card payment attempts does not result in a successful transaction. The biggest reason is not insufficient funds, but friction.

The most common failure points are invalid card details (around 27% of non-successful attempts), transactions exceeding withdrawal limits (around 15%) and 3D Secure being abandoned mid-flow (around 5%). Many of these failures happen before the user even gets to the bank's authorisation screen.

Dewald framed the cost this way: "A failed transaction is not an event that the user goes back to a payment page from. A failed transaction is a data point. That should be the start of the journey for that specific user, not a reset of the payment journey."

The practical implication is that the failure reason should drive the next action automatically. A user abandoning 3DS needs a different prompt vs a user who has hit a withdrawal limit. Merchants that surface this granularity and build intelligent retry logic around it are recovering revenue that others write off.

This becomes more complex in recurring collections contexts, where timing matters as much as method. Stitch transaction data shows more than 80% more transactions happen on the 1st of the month than on the 22nd, with peak volume on the 25th – right around payday. Scheduling collection retries earlier in the first week rather than on the 1st itself can meaningfully reduce failure rates.

Variable Recurring Payments (VRP) opens up new use cases for bank payments

Pay by bank has historically solved for once-off transactions. The rise of PayShap Request has made it faster and more uniform across banks, but it still left two use cases unserved: subscription-style recurring payments and variable-amount transactions like on-demand grocery delivery.

Variable Recurring Payments (VRP), first introduced in South Africa by by Capitec Pay, addresses both. Consumers can now authorise a merchant to collect variable amounts from their bank account within predefined limits, without needing to re-authenticate each time. This replaces what previously required a stored card, and it is already live with at least one major South African retailer.

Jessica noted: "What VRP is unlocking is an entirely new rail on an entirely new use case, and I think it is just going to aggressively compete with card."

BNPL is mainstream, but the customer split is more nuanced than most merchants assume

71% of credit-active consumers now use Buy Now Pay Later at some frequency. One in five use it regularly, and the South African BNPL market is projected to reach $1.11 billion in 2026.

The misconception worth addressing is who is actually driving that adoption. The strongest engagement comes from households earning between R25,000 and R50,000 a month, not from lower-income consumers as might be expected, and not from the very wealthy. Jessica described three distinct buyer types observed in the research.

The first is the deal seeker: someone who could afford to pay in full but prefers not to. The interest-free element is the primary draw, cited by nearly half of respondents. 

The second is the liquidity manager: someone with funds but a preference to preserve cashflow for large or life-event purchases such as electronics or baby items. 

The third is the no-funds buyer. This is someone with a genuine affordability gap, but this group is also often excluded by BNPL credit checks, which means the 71% figure reflects adoption among approved, credit-active consumers more than the total addressable market.

48.5% of respondents said they expect BNPL to be available both online and in-store, showing a growing preference for a cohesive online and in-person experience. Merchants need to take note that updating their online payment offerings is certainly useful, but they can’t ignore their in-store payment methods either.

Consumer preference for brand websites remains strong, but international platforms are closing the gap

74.6% of consumers prefer to shop directly on brand websites. That figure is encouraging for South African merchants. But 48.5% have already shopped on international low-cost platforms like Temu and Shein, up from a negligible proportion just two years ago. They are not switching; they are adding these platforms to their repertoire.

Jessica's advice: "As a South African brand, you can win on three things that international platforms cannot replicate: local trust, delivery speed and payment experience. Being local and on the ground and choosing the right provider is an edge that they'll never really be able to compete on."

Fraud is more visible at the merchant level than most realise

Security is the second-most-cited factor in consumer payment decisions, consistently across all four years of Stitch's research. But Dewald raised a less-discussed issue: most South African merchants significantly underestimate their exposure to fraud because they lack the tooling to see it.

A merchant may see ten separate payer accounts making transactions. Without shared device or card data surfaced at the merchant level, those accounts look unrelated. With the right platform data, clusters of accounts sharing a single card or device become visible, and the patterns that indicate syndicate behaviour can be identified before they scale.

"The cost goes beyond fraudulent transactions," Dewald noted. "The operational overhead is immense. And bad actor networks can skew data so materially that a business ends up making decisions using the wrong insights."

AI is already in the purchase journey

31 to 34% of South Africans are active users of ChatGPT, and the report found consumers using it for product research, price comparisons and purchase decisions. According to a Shopify study from May 2026, AI-referred shoppers convert at nearly 50% higher rates than those arriving through organic search.

The forward-looking question is how payment infrastructure keeps pace. As Jessica described it, agentic commerce means an AI agent receives a goal from a consumer, for example "find me the best price on this product delivered by Friday under R500," and then discovers options, evaluates trade-offs and executes the purchase autonomously. The agent cannot complete a 3DS verification in a banking app. It cannot enter an OTP. Traditional checkout flows are not designed for it.

Merchants that will be visible to AI-mediated commerce are those with structured product data, clean catalogues and payment infrastructure that can accept programmatic transactions via API. Those that rely on manual-only checkout flows will be, as Jessica put it, invisible.

Stitch has published a four-part blog series on what merchants need to do to prepare. For a deeper read on this, visit stitch.money/blog/what-is-agentic-commerce.

Three things merchants should take away

Jessica and Dewald closed the webinar with three priorities for any merchant building payment strategy in 2026.

First, diversify the payment mix. The goal is not to add one more method for marginal gain, but to build a payment ecosystem that can intelligently route based on user behaviour and failure data. 

Second, fix what is already broken. The revenue lost to failed card transactions and abandoned 3DS flows is the lowest-hanging fruit available, and most of it is recoverable. 

Third, understand what you do not know yet about fraud and AI. Both are already affecting merchant revenue, and neither will slow down.

The full 2026 Consumer Payments Report is available to download and the full webinar can be viewed here.

‍

FAQs

What is the most popular payment method in South Africa in 2026?

There is no single dominant method for the first time in Stitch's four years of research. Apple Pay became the most-used e-commerce method by volume on the Stitch platform in Q1 2026, overtaking card. Capitec Pay accounts for around 40% of total payment value across Stitch transactions. 93.3% of South African consumers tried a new payment method in the past 12 months, which reflects how quickly the landscape is fragmenting.

Why do so many card payments fail online in South Africa?

Stitch platform data from Q1 2026 shows nearly one in four online card payment attempts does not result in a completed transaction. The leading causes are invalid card details, transactions exceeding bank withdrawal limits and 3D Secure being abandoned mid-flow. Most of these failures happen before the user reaches the bank's authorisation screen, and each represents recoverable revenue if merchants build intelligent retry and routing logic.

Who is actually using Buy Now Pay Later in South Africa?

The strongest BNPL adoption is among middle-income households earning R25,000 to R50,000 per month. Lower-income consumers often don't qualify due to credit checks, and higher-income consumers tend to use it for cashflow management rather than affordability. 71% of credit-active consumers use Buy Now Pay Later at some frequency, and nearly half of all survey respondents expect it to be available both online and in-store.

What is Variable Recurring Payments (VRP) and why does it matter for merchants?

‍ Variable Recurring Payments (VRP) allow consumers to authorise a merchant to collect variable amounts from their bank account within agreed limits, without re-authenticating each time. This unlocks bank-to-bank payments for recurring subscriptions and variable use cases like on-demand grocery delivery, which previously required a stored card. Stitch supports Capitec Pay VRP and already has the product live with South African merchants.

How should South African merchants prepare for agentic commerce?

AI agents executing purchases on behalf of consumers cannot complete OTPs or navigate traditional checkout flows. Merchants need structured product data that agents can read programmatically, and payment infrastructure that accepts transactions via API without requiring human interaction at each step. Stitch has published a series on this topic here.