Black Friday falls on 27 November 2026. That is 17 weeks from today. For a South African distribution centre, the useful planning window is the next eight of them, because anything that touches slotting, labour planning or dispatch design needs time to bed in before volumes climb.

The last peak set the benchmark. Capitec customers spent R28.8 billion over Black Friday 2025, up 14% year on year, with online spending up 43%. Online transactions over the Black Friday to Cyber Monday period grew 30.4%, lifting online from 7.9% to 10.3% of total holiday sales. Total November 2025 retail sales came in at R117.9 billion. E-commerce is now close to 10% of South African retail and still growing faster than the store base.

Why the DC becomes the constraint

Between November and January, order volumes through a distribution centre can double. Three things happen at the same time, and they compound.

Order profile changes. Peak orders are smaller, more frequent and weighted toward a narrower set of promotional SKUs. A slotting plan built on the March to September profile puts the wrong stock in the wrong place for exactly the eight weeks that matter most.

Labour gets thinner and greener. Permanent staff take annual leave in December, and the gap gets filled with temps who are new to the building. Productivity per picker falls at the moment throughput needs to rise, and error rates climb with it.

Returns eat the floor. In apparel, footwear and electronics accessories, returns can occupy up to 20% of warehouse capacity during peak. Space budgeted for outbound stock ends up holding inbound returns.

What can still be fixed in 17 weeks

Robotics and racking changes are off the table this cycle. Lead times and capex approvals will not clear before November. The changes that can land in time are the ones that alter how existing people and systems execute the work.

Re-slot against the peak profile, not the annual one. Pull last year's November and December order lines, rank SKUs on peak velocity and order affinity, and move the promotional lines to the shortest travel positions. This is a data exercise followed by a physical move, and it needs to happen in September so the floor is stable by late October.

Fix the dimension file before you rely on it. Cartonization and slotting logic both depend on accurate SKU dimensions and weights. If the dim file is stale, peak is when the errors show up as split shipments and wasted cube. Rebuilding dimensions across 8,000 to 30,000 SKUs takes four to six weeks, which puts the start date in August.

Design the peak pick strategy now. Batch sizes, zone allocation and cart configuration that work at normal volume often break at double. Model the change against your own data before the temps arrive, so supervisors are training people on a method that has already been tested.

Decide the channel mix before you need it. Pickup points and courier capacity get booked out. Knowing which orders go to owned fleet, which to courier and which to collection, and what each costs per drop, is a September decision. Carrier capacity secured late costs more and delivers worse.

Give returns a dedicated lane. Plan the space and the labour for returns in advance rather than absorbing them into the outbound area in week two of December.

A workable timeline

August: rebuild the SKU dimension file and pull last peak's order data.

September: re-slot against the peak profile, model and agree the pick strategy, lock carrier and channel capacity.

October: run the new method at normal volume, train supervisors and temps on it, hold the layout stable.

November: freeze changes. Peak is for executing, not for testing.

Where OptiChain sits

Most of what determines peak performance is decided in the eight weeks before anyone feels busy. The operations that hold their cost per order through December are the ones that treated August and September as the peak project.

OptiChain Solutions works with South African retailers, distributors and 3PLs on peak readiness: modelling the peak SKU profile against current slotting, sizing the picking and cube opportunity, and building the labour and channel plan against real demand curves. If you want to know where your building loses time at double volume, the data will tell you now, while there is still time to act on it.

Sources