What ten thousand orders in a day teaches you

A store we run had a day when it took something north of ten thousand orders between breakfast and bedtime. At the peak it was several thousand in a single hour, which works out to an order roughly every second for sixty minutes.

We knew the date months out and had run a day like it before. It still found things to break, and none of them were the things you would guess.

The part that does not break

Checkout. This surprises people. Shopify is battle-hardened by a decade of Black Friday weekends, and they publish their own numbers after each one if you want to see what the platform shrugs off. The site staying up is the one thing you can stop worrying about. Everything that goes wrong happens after the money has been taken.

The parts that do

• Stock. We run days like this as pre-orders: forecast the first run, then top up as the real orders land and the team produces against them. It works, and it is still a tricky juggle. Every big launch teaches us something about it that the last one did not.
• Your size curve. Whatever ratio of smalls to larges you normally see does not hold when everyone arriving is the same age from the same place. A size that is never a problem became the one we were chasing all day.
• The inbox, one day later. Orders go up twentyfold on the day. Enquiries go up twentyfold the day after, when every confirmation email has landed and ten thousand people have a question. Your support team had better be ready, and well lubricated with coffee.
• The floor. It becomes chaos, and there is no version of the day where it does not. The job for everyone not putting ink on cotton is to do everything possible in the background so the people who are can keep going: ordering blanks, clearing the queue, keeping the freight moving.

What actually saves you

Not committing to the whole size curve up front. Forecast the first run, produce against real orders from there, and being wrong costs you a slower week rather than a warehouse full of one size.

The rest is preparation. Decide before the day what you will pause when it starts to hurt, and in what order. Not what you will scale up. What you will switch off. The abandoned-cart email. The low-stock alert. The review request. Decide it while you are calm, because nobody makes a good call about a queue while the queue is on fire.

The orders arrive in a day. They leave over a fortnight. The fortnight is the job.

The fortnight

The big day is the story people want to hear. The two weeks after it are a print queue, a pick line, a freight booking and somebody reading every message that comes in. Nobody photographs that part.

Ten thousand orders is a number. Ten thousand parcels that turned up when you said they would is a business.