When you reserve an Align, your device does not come off a shelf. It gets built as part of a numbered run, and the first one is twenty units. This note is about why, and about what it costs us — because batch production is usually written about as though it were only an advantage.

What a batch actually is here

A batch is a fixed number of devices built, checked and shipped together. Batch 01 is twenty units. When those twenty are spoken for, the run closes and the next one opens.

Nothing is manufactured before it is reserved. An order is placed, someone calls to confirm it, and only then does that unit enter the run.

The honest reason: capital

The first reason is the least romantic one, and it would be dishonest to lead with anything else. We are a small team funding production ourselves. Building a hundred units speculatively means putting the money for a hundred units into components and assembly before knowing whether a hundred people want them.

Twenty is a number we can fund, build carefully, and deliver without needing outside money or a warehouse. If nobody wanted them, we would have twenty devices and a lesson, rather than a hundred devices and a problem.

The better reason: every unit is checked by a person

At twenty units, each device can be tested by hand before it ships. Not sampled — each one. Somebody powers it on, works through the interface, checks the display and the LED and the tone, confirms the button behaves, and confirms it holds a reference.

That is possible at twenty. It is not possible at two hundred without a different process, different tooling and different people. Batch size and quality process are the same decision wearing two hats, and choosing a small batch is choosing to keep a human in the loop.

It also means a fault found in one unit can be checked against the other nineteen before any of them leave, because they are physically together and none has shipped.

Changes between runs, not during them

Everything in a batch is built to the same revision. If we learn something during Batch 01 — a component that is hard to source, an assembly step that is fiddly, a firmware behaviour worth changing — it goes into Batch 02.

This keeps support tractable. When someone contacts us about a device, knowing which batch it came from tells us exactly what is inside it. A continuous production line with rolling changes means every unit is potentially slightly different from every other, and a year later nobody can reconstruct what a particular customer actually has.

Why the counter on the site is real

The batch status on the website reads from the order system. It is not a number we set by hand to create urgency, and it does not reset when it gets uncomfortably high.

This matters to us more than it probably matters to anyone reading. Fake scarcity is common enough in direct-to-consumer selling that a real counter is indistinguishable from a fake one at a glance — which is an argument for not having one at all rather than for having a dishonest one. When Batch 01 fills, the page says the batch is closed, because it is.

What it costs

Being straight about the downsides:

  • You wait. A device is built after you order it, not before. Ordering from stock is faster and we cannot match that.
  • We are more expensive per unit. Component pricing improves substantially with volume. Twenty units buys none of that. Some of that cost is in the price.
  • A batch can close before you reach it. Then you wait for the next run. This is genuinely inconvenient and we do not have a way to avoid it that does not involve building speculatively.
  • It does not scale. Hand-checking every unit stops working at some size. If demand grows past that point, this process has to change, and we would rather say so now than pretend the current approach is permanent.

Cash on delivery, and why it fits

We take payment in cash when the courier arrives. No card details are collected anywhere on the site.

Partly this reflects how people buy in Egypt, where cash on delivery is normal and expected rather than a fallback. Partly it follows from the batch model: if we are asking someone to wait for a device to be built, asking them to pay up front as well is a lot of trust to request from a company they have not bought from before. Paying when the box is in their hand moves that risk to us, which is where it belongs.

It costs us: cash on delivery has higher failure rates than prepayment, and every order needs a phone call to confirm before anything is built. At twenty units that call is affordable, and it catches address problems and misunderstandings before they become a courier's problem.

Where this goes

We expect this to change. Batch sizes will grow, hand-checking every unit will become sampling with better test tooling, and some of the calls will become automated confirmations.

What we would like to keep is the part that is actually load-bearing: nothing is built before somebody wants it, the number on the site is the real number, and someone can still tell you what is inside the device you own.