What a Brand Identity Includes, and What You Can Postpone

An identity is a set of rules, not a delivery of files

The confusion that makes identity projects expensive is treating the outcome as assets. Assets are the by-product. The outcome is a set of decisions that removes the need to ask a designer what to do next, which is what stops a brand drifting the moment more than one person touches it.

The practical test is whether someone outside the project can produce a correct page, post or label using only what was delivered. If they have to guess a size, a spacing or which of two blues is the real one, the identity is incomplete regardless of how many files it contains.

It is equally useful to know what an identity is not. It is not a website, and it does not decide your page structure or your copy. It is not a marketing strategy, and it will not tell you which channel to spend on. It is not a product design system either: a component library describes interface behaviour for engineers, while an identity describes brand behaviour for everyone. Buying one while expecting the other is the most common disappointment in this category.

The four things that make an identity usable

The mark with its clear space and its smallest honest size. A mark that has not been tested at the size it will actually appear — a favicon, a stitched label, a boarding-pass footer — is an untested mark.

A type pair with real sizes, not just names. Heading and body, the sizes they take on a phone and on a desktop, and what happens to a long heading. A type choice without sizes is a preference, not a rule.

A colour set with values for every medium you use. Screen values and print values, and a stated rule for which colour carries meaning and which is decoration. One flat colour chosen deliberately is cheaper and more reliable than a gradient that drifts between suppliers.

One page per surface you publish on, showing the rules applied. Three surfaces is usually enough to prove the system works and to catch the contradictions that only appear in use.

What arrives as files is ordinary and worth agreeing in advance: vector masters for the mark, a small set of exported sizes, the type files or a licence reference, colour values per medium, and the applied pages as a single document. Editable sources matter more than the count of formats, because the next surface is always one you did not list.

What you can safely postpone

Illustration style, photographic direction, motion, packaging systems, merchandise and sub-brand architecture are all real parts of a mature identity and none of them blocks a launch. Commission each when a surface needs it, because a style decided before its first real use is usually redecided after it.

Postponing is only cheap if the first four decisions are firm. If the type sizes or the colour rule are still open, every later surface reopens them, which is the actual source of the rework people blame on scope creep.

When a company needs an identity at all has a practical answer: when more than one person produces public material, or when the same thing has to look the same twice. A single founder posting from one phone does not need a rulebook yet. The moment a contractor, a printer and a marketplace listing are all producing brand material, the absence of rules costs more than the rules would have.

How this differs from a logo project

A logo project answers one question: what is the mark. An identity answers a different one: what are the rules for using it alongside type, colour and layout. Buying the second when you needed the first is the most common overspend in this category, and buying the first when you needed the second is why brands with a good mark still look inconsistent.

If the mark does not exist yet, start there and keep the brief tight. The scope of a mark brief is a separate question with its own answer.

The sequence matters more than the labels. Name, then mark, then rules. Each step constrains the next, and running them in parallel means the later step is done twice. If a name is still being argued about, an identity built around it is rebuilt rather than adjusted.

What guidelines have to specify to be worth paying for

Useful guidelines are written as constraints a non-designer can follow: minimum sizes, clear space expressed in a unit that scales, the behaviour of a long heading, what to do when the background is a photograph, and which combinations are forbidden. Pages that only show the mark looking good on a neutral field do not survive contact with a real layout.

The three VITON13 concept projects below were each built against a deliberately hard surface, which is the part that exposes whether a rule is real.

Scope and revisions belong in the same document as the rules. Agree how many review rounds the price covers, what counts as a new surface rather than a revision, and who signs off. Most overruns in this category are not design disagreements; they are an unstated definition of done.

How the work is produced changes the shape of the process, not the deliverable. An AI-assisted pass explores more variants early and compresses the drafting stage, which suits a scope that is already decided. A human-led pass spends that time on judgement instead: which constraint to break, which surface will embarrass you later. The published ladder prices both, and the honest reason to choose the second is an undecided scope, not a prettier file.

Concept proof: three marks tested on difficult surfaces

Movement — fitness equipment. CONCEPT, designed by VITON13. The hardest surface for a mark: a dark photograph with motion blur behind it. Two solid wedges, no thin strokes to disappear.

FlyFly — aviation. CONCEPT, designed by VITON13. Placed over the brightest and the darkest part of one frame at once. The italic keeps its counters open at boarding-pass size.

PUMP DRINK — sports drinks. CONCEPT, designed by VITON13. One flat colour on purpose. A bottle is printed, not rendered: one ink is cheaper per unit and cannot drift between suppliers.

These are concept projects, not client engagements. They are included because each one records a specific decision and the constraint that forced it, which is the kind of detail a guideline page has to carry.

A portfolio that shows only the finished surface teaches nothing about what to ask for. Each note above names the constraint that forced the decision, which is what a buyer can actually reuse when briefing their own project.

How to accept the work: Mark supplied with clear space and a stated smallest…

Accept on behaviour, not on appearance. Ask for the mark at its smallest stated size, a long heading in the real type at phone width, the colour values for both screen and print, and one surface you did not commission rebuilt using only the delivered rules. If all four can be produced without a question, the identity is finished for this stage.

Agree the postponed list in writing at the same time. A scope that names what is deliberately not included is what keeps the next invoice predictable.

A brief that saves money is short and factual: what the business sells, who decides, the exact legal name and any trading name, the three surfaces that matter first, the languages in play, and anything already fixed such as an existing typeface licence or a printer limitation. Send constraints rather than references, because a reference says what you liked while a constraint says what must survive.