Process

How we build.

One loop, repeated: sketch, prototype, test with people, decide. Here is what each step means in practice.

The loop

Four steps, then again.

  1. 01Sketch Write down the problem and draw the smallest thing that could help.
  2. 02Prototype Build the cheapest version that actually works, on a desk, by hand.
  3. 03Test with people Put it in front of people who did not make it, and watch.
  4. 04Decide Choose what deserves to be built next, and what to drop.

Every pass through the loop ends in a decision: keep going, change direction, or stop. Stopping is a valid result, and it gets written up like any other.

Choosing a prototype

What earns a prototype.

  1. 01A real problem. Someone we can name has it, and it shows up more than once.
  2. 02Small enough to test in weeks. With parts we can afford and skills we have or can learn quickly.
  3. 03Useful without the internet. The core has to work offline. Anything online is a bonus.
  4. 04One sentence. If we cannot say why it is useful in one sentence, it is not ready to build.
  5. 05Both of us agree. Two founders, equal say. If we disagree, we test the cheapest version and let it decide.

Testing

How we test with people.

  1. 01People who did not build it. Friends, classmates, family. Not us.
  2. 02We watch, we do not explain. If we have to explain it, that is a finding.
  3. 03We note what happens first. Where they look, what they press, where they get stuck, what they ignore.
  4. 04We ask what they would stop using. Not whether they like it. People are kind; behaviour is honest.
  5. 05We write it down. Results go in the journal, including the ones that make us look bad.

Scope

What we will and will not build.

Will build

  1. 01Small physical objects with one clear job. Starting with a desk companion: a face, a focus timer, a few buttons.
  2. 02Software that supports them. Only where it makes the object more useful.
  3. 03Digital tools we would use ourselves. Later, and only one thing at a time.

Will not build

  1. 01Features because competitors have them. A feature has to earn its place.
  2. 02Anything that needs constant internet to be useful.
  3. 03Emergency or safety features we cannot stand behind. We will not promise detection or rescue. Those claims need systems we do not have.
  4. 04Things designed to demand attention. A desk object should help you focus, not compete for it.

Failures

When something does not work.

It goes in the journal: what we expected, what actually happened, and what we will change. We do not delete failed experiments or rewrite them to look like plans.

Concept images are always labelled as concepts.

Read the journal

What we do not do

The short list.

  1. 01No client work. Superlab only makes its own products.
  2. 02No fake progress. No invented numbers, testimonials, press logos or deadlines.
  3. 03No data we do not need. An email address to write to you. Nothing else unless a product genuinely requires it.
  4. 04No pre-orders before a working prototype. And when there are pre-orders, deposits are refundable and dates are estimates.
  5. 05No dark patterns. No countdown timers, fake scarcity or guilt-trip buttons.