1.4 Your starting point

The Role · Section 1.4

An honest inventory, so you spend your time on the gap rather than on what you already have.


What you already have

production code        the hard half, and the one that takes years
systems thinking       architecture, state, data flow
debugging              including the browser rendering path
shipping               the discipline of finishing things
technical constraints  you know what is expensive before you try it

That last one is a real asset in design conversations. When a designer proposes something, you can say what it costs — and offer the nearest cheap alternative. That is a contribution, not an objection.


What is missing

the eye          seeing that something is wrong before being told
vocabulary       saying why, in terms that transfer
visual decisions  choosing between three acceptable options
feel             judging motion, timing, responsiveness
design fluency   reading a Figma file the way you read a codebase

The first is the foundation. Most engineers can tell when an interface is bad; far fewer can tell when it is slightly wrong, and fewer still can name the cause. That specific conversion — from "looks off" to "the gap inside the group matches the gap between groups" — is what chapters 2–5 build.


The honest difficulty

Design judgement is harder to acquire than it looks, for a reason that catches engineers out: there is no compiler.

You are used to feedback that is immediate and unambiguous. The code runs or it does not. Design has none of that. You will make a choice, it will be slightly wrong, and nothing will tell you. The feedback loop is slow, subjective, and requires you to develop your own error detection.

This is why the method in chapter 17 matters more than any list of principles. Reading about hierarchy teaches you the word; rebuilding interfaces teaches you to see it.


The shortest useful first step

Paco Coursey, design engineer at Linear, previously Vercel:

Copy and re-implement work you admire until you can proudly create for yourself.

Concretely: pick one interface you think is beautiful. Rebuild it in code, as closely as you can, without looking at its source.

You will fail in specific ways, and the failures are the lesson. The spacing will be slightly wrong. The type will be too heavy. The animation will be too slow. Each gap between your version and theirs is a decision you did not know was being made.

Do this five times and you will have learned more than any article can give you — this one included.


The mental shift

One more, from the same source: start prototyping in code instead of only in Figma.

For you the shift is different. You are not moving to code — you are already there. Your shift is to stop waiting for a mockup before you start. When you catch yourself thinking "I'll build it once design sends something," that is the habit to break.

The role begins when you can start from the problem.

On this page