Design Intent

Every other rule rules something out. This one decides the rest.

The rules under this sentence are all subtraction. A margin rule says how close to the edge nothing may go. A clear space rule says how near the mark nothing may sit. A type rule says how many sizes may appear and how far apart two of them have to be. Each one removes options, and what is left over is still enormous.

Almost everything made for you lands in what is left over. Nothing in your brand says whether a card gets a hairline border or a soft shadow. Both are legal. Your colours permit both, your spacing permits both, and no rule you have ruled mentions either one. Something still has to pick, and if nothing you wrote picks, whoever or whatever is building picks instead.

That is the job this field does, and it is the only thing in your brand that speaks to the territory the rules do not cover.

Where the sentence actually gets used

It is not read the way a rule is read. A rule is checked: the margin either is five per cent of the short side or it is not. This sentence is consulted, and only at a particular moment.

The moment is a fork where both branches are allowed. Something is being built, it arrives at a choice, and every rule you have stated is silent on it. Border or shadow. A soft corner or a sharp one on a thing your brand has never classified. Two greys that both pass. A headline at the top of the page or against its centre axis. At that fork your sentence is what decides, and it decides the same way every time, which is the whole of the point. Consistency across work you never reviewed is not something rules can buy you, because rules only cover what somebody thought to rule.

So the test of a design intent is not whether it is true about your brand. It is whether it picks.

What a good one looks like

A good one has two halves that pull against each other, and the order they come in matters.

Take "structured, but approachable". Structured on its own would put everything on a grid and give it a hard edge. Approachable on its own would round every corner and soften every contrast. Neither half is usable alone, because a sentence that only pulls one way runs off the end of itself: applied to enough decisions it produces a brand that is nothing but its one adjective. The pair gives a direction and a limit in the same breath. Lean structured, and stop before it stops being approachable.

The order is working too. Put approachable first and you have described a different brand, one that softens by default and reaches for structure only where it has to. Two sentences, the same two words, and they settle the same fork differently.

Two halves is a shape rather than a requirement. What is required is that the sentence can lose an argument. Read yours, then read a fork it ought to settle, and watch whether it picks a side.

The two that decide anything are the two another brand might refuse to sign. That is not a coincidence, and it is the fastest check there is. If every brand in your category could put its name to your sentence, it is narrowing nothing, and a sentence that narrows nothing cannot settle a fork.

The three that look like an intent and are not

A mission statement. It is about what your organisation does for people, it is usually true and often good, and it says nothing about how a thing should look, so it cannot resolve anything visual. This is the most common answer typed into this field, and it is the one that feels most like a real answer while it is being typed.

A pile of adjectives. Bold, modern, clean, human. Four words nobody would argue with, which is precisely the problem: an adjective no one would disclaim excludes nothing. Cut the list down to the two that are in tension and you often have your sentence already, sitting inside something you have had for years.

A rule you have already ruled. "We use generous margins" is your margin rule written a second time and less precisely. The intent has to speak to what the rules do not reach, so anything it says that a rule already says is spent.

Getting one out of a book that never states it

Most brand books do not contain this sentence. They contain its consequences, which is better than nothing, because consequences can be worked backwards.

Start with the pages that argue. Almost every book has a spread of pairs, this way and not that way, six or eight of them, each wrong one with a small cross against it. Nobody produced those pairs from nothing. Somebody had a reason, applied it eight times, wrote down the eight outcomes and threw the reason away. Read all eight and ask what single sentence would have produced every one of them. It usually falls out in a couple of minutes, and you are not inventing a sentence, you are recovering the one that was already operating.

Then look at the personality page, if there is one, and cross out every adjective a competitor would also claim. What survives is rarely more than two words, and they are usually in tension, because the ones in tension are the ones that were hardest to agree on and the safe ones are what got added afterwards.

If neither page exists, use the rejections. Work your brand has turned down is a sharper record of intent than work it approved, because approved work is consistent with a great many sentences and a rejection narrows it to a few. Think of the last thing that came back and what was wrong with it. The answer will have a specific case stuck to it, something like the photography being too styled. Take the case off and you have a sentence about everything.

If you have nothing written down

Then you are deciding rather than transcribing, and there is a version of this that takes five minutes.

Think of the last time two options were both fine and you picked one. Not a case where one of them was wrong, a case where either would have passed and you had a preference anyway. Write down why you picked. Then strip out everything specific to that pair, until what is left would still be true about a completely different pair. That is a design intent, and it is a better one than anything you would reach by thinking about your brand in the abstract, because it came from a decision you actually made.

You can change it later. It is one sentence in a field, it is recorded with your name and the date against it, and a brand that sharpens it in six months has lost nothing by stating a rougher one first. A sentence that is roughly right and stated beats a better one you are still working on, because the one you are still working on is deciding nothing.

What happens if you leave it blank

Nothing breaks, and that is the difficulty with it.

Not stated is a real state rather than an error. Every rule you have ruled goes on working exactly as it did. The card raises it once you have begun to rule things, and it is raised at the lowest of the three priorities, below anything that would make something confidently wrong. If you have ruled nothing at all the card stays quiet about it, because your next move is the same either way.

What you lose is not visible in any single piece of work. Everything you asked for still gets made. At every fork the rules leave open something picks, and the pick is always plausible, because an implausible pick is easy to spot and nobody makes it. Nobody looks at a reasonable border and asks what decided it. The cost turns up across a run of work rather than inside one item of it: forty things that are each defensible and are not recognisably one brand, because nothing was making them agree. That is slow to notice, and it is slow to undo.

Where the sentence travels, and when it does not

Your sentence goes out with your named roles. It leads that file, above them, so that anything building with your brand has read it before it reaches the first choice it has to make. The placement is the point: a sentence met after the list has already chosen is decoration.

Which means a brand that has stated an intent and named no roles is not sending the sentence anywhere yet. It is stored, it is yours, and it starts travelling the moment there are roles for it to lead.

There is one other case and it is rare. Where a file has almost no room left, the roles, the rules and the sentence are dropped together rather than sent in halves, and you are told that they were. Roles with their reason removed would be a worse file than no roles at all.

The other rules

Everything under this sentence is one of three rules and a set of components. Grid and margins is the one to read first, because it sets out the idea the other two are built on: a rule is stored as a relationship and never as a measurement. Clear space is that idea applied to your mark and type hierarchy to your sizes. Brand components covers the card underneath, where the parts a screen is built from get described.