Black Friday failures don’t always look like failures at first. A retailer can keep its website online, process every payment and watch the sales dashboard climb while the operation behind it is already starting to slip.
The problems are usually much less dramatic than a site going down. Someone has to recreate a waybill, an address needs fixing before a courier will accept it, a collection doesn’t happen when expected or tracking stops updating. During an ordinary week, staff can often work around those problems without thinking too much about them. Add several times the usual number of orders and the workarounds start competing with the work.
South African retailers have more of that volume to deal with every year. World Wide Worx’s Online Retail in South Africa 2025 research estimated that local online retail grew 35% to R96 billion in 2024 and projected sales above R130 billion in 2025, putting e-commerce close to 10% of national retail sales.
Black Friday puts unusual pressure on that growing market. Peach Payments processed more than 2.69 million transactions worth R1.86 billion between Black Friday and Cyber Monday in 2025, with the value processed up 93% from the previous year. Its transaction value across November grew even faster, at 109%, as retailers spread promotions across more of the month.
Sahil Affriya, founder and CEO of South African logistics platform Shiprazor, says he tends to encounter the same operational problems when the November rush is already underway. “I find myself having the same conversations with business owners in late November, always too late to make a difference,” he says. By then, a few strong trading days can be followed by a pause in collections and a backlog that has to be cleared while new orders are still arriving.
Start with the minutes after payment
Affriya looks first at what happens between a successful checkout and the creation of a courier waybill. “Your first possible bottleneck sits just past checkout in the time between an order landing and you being able to create a waybill,” he says. If that takes hours, or somebody is still typing customer details into a courier portal, the business already knows where additional volume is going to hurt.
Retailers can miss that weakness because staff are good at compensating for it. An employee notices an order stuck on pending and fixes it, spots a duplicate shipment before it leaves or corrects an address without anyone recording that the system needed help. Five awkward orders scattered across a normal week are irritating rather than disastrous. The same percentage of awkward orders during a large promotion gives that employee a queue.
Affriya recommends recording failed connections, duplicate shipments, pending orders and any other cases where a person has to intervene before the parcel can move. That’s particularly relevant because retailers already spend so much effort getting people to complete the purchase in the first place. South African online merchants contend with very high cart-abandonment rates, so getting a customer through payment only to introduce avoidable delays immediately afterwards wastes some of that work.
A retailer preparing for Black Friday on 27 November can test this long before the big promotions start. Take an ordinary busy day and measure how long orders sit between payment and waybill creation, then count how many needed someone to step in. The cases people usually fix without mentioning them are probably the ones worth paying attention to.
Customers notice when the information disappears
The next problem often surfaces as a customer-service message. Affriya puts it simply: “Late parcels don’t automatically lose customers but silence can.” A delivery arriving later than expected and a delivery that appears to have vanished create very different experiences, even when both eventually arrive.
That difference has become harder for retailers to ignore because South African shoppers have grown accustomed to far more visibility around delivery. Checkers has previously been confident enough in Sixty60 to offer a delivery guarantee, while Amazon Prime has brought same-day delivery to parts of the country.
An order placed after Amazon Prime’s advertised midday cutoff on a public holiday still arrived a little over three hours later. One order can’t tell us how Amazon performs across every route and every day, but experiences like that inevitably affect what customers consider possible when they buy elsewhere.
An independent retailer doesn’t have Amazon’s logistics network and shouldn’t need one to provide a decent delivery experience. If it promises three-day delivery, though, it should know what has happened to a parcel during those three days and be able to give the customer something better than a promise to phone the courier.
Affriya suggests paying attention to failed delivery attempts, non-delivery reports, return-to-origin rates, repeat attempts and the volume of customers asking where their orders are. Those numbers can expose trouble that an average delivery time hides. A retailer may still hit its delivery target most of the time while a relatively small group of problem parcels generates a disproportionate amount of support work.
A courier’s national promise can hide local weak spots
Courier performance also looks different once a retailer stops treating the advertised transit time as the whole story. “Most merchants set their courier rules once and never look at them again,” Affriya says. He argues that retailers should test those promises against what actually happens on the routes they use most often.
That can produce a very different picture. A courier may be consistently good between major metros while struggling on routes that happen to matter more to a particular retailer. Missed collections matter too, especially around weekends and public holidays, because Friday’s unfinished work can still be sitting there when Monday’s orders start arriving.
Affriya also argues for having more than one courier available, saying that “relying on a single courier means your only backup plan is hope”. Shiprazor has a commercial interest in that position because it sells a platform that gives merchants access to multiple couriers from one dashboard, so using several delivery providers shouldn’t be treated as a cure on its own.
A retailer can have accounts with several couriers and still have a bad fallback plan. If shifting an order from one provider to another means logging into another portal, rebuilding the shipment and copying information across manually, the backup introduces more work at exactly the moment the business is already behind.
The manual work is easier to see when you count it
Some of the most revealing preparation has nothing to do with buying new software. Affriya suggests retailers “question every manual process: courier allocation, waybill generation, address correction, customer notifications, status updates”. The issue isn’t that every manual task is automatically bad. A small retailer may quite reasonably decide that automating something used twice a week isn’t worth the cost.
What matters is knowing which jobs rise directly with the number of orders. If someone fixes ten addresses, checks five courier exceptions and manually sends dozens of status updates during an ordinary week, that workload doesn’t stay fixed when sales multiply. It follows the orders.
A September or October promotion gives retailers a chance to watch this happen at a manageable scale. Staff can record which orders need rescuing, where information has to be copied by hand, which courier problems recur and what customers begin asking when deliveries slow down. They should keep watching after the promotion ends too, because a busy sales day can look successful while the fulfilment work it created is still sitting in the warehouse two days later.
That overlap is the scenario Affriya keeps seeing: several strong trading days, a collection pause and then old orders waiting to move while new ones continue arriving. Black Friday makes the consequences worse, but it doesn’t create the underlying process.
Black Friday has also stretched well beyond a single Friday. Peach Payments’ 2025 figures show how much spending now happens across November, which means an inefficient fulfilment process has longer to accumulate unfinished work before the traditional Black Friday weekend even arrives.
Most of the warning signs are already sitting inside ordinary orders: the label someone has to recreate, the address that needs fixing, the courier somebody has to chase and the customer who has to ask where their parcel is. Those jobs are manageable now because there aren’t hundreds of them arriving together. In November, they may be.