Brand
The Brand Guidelines Nobody Uses
Companies pay for an identity, get a beautiful PDF, and six months later every channel looks like a different company. Why, and what to demand.
A pattern worth recognising, because it is expensive and it is common.
A company commissions a brand refresh. Six weeks later a studio presents an identity — a new mark, a palette, a typeface, a 60-page PDF. Everyone is pleased. The invoice is paid.
Six months on, the sales deck, the website and the LinkedIn posts look like three different companies.
The identity was not the problem. The absence of a usable system was.
The real test of an identity
Not whether it looks good in the presentation. The test is:
Can your sales lead build a deck at 11pm on a Tuesday and have it look right?
If the answer requires them to open a PDF, find the correct hex value, remember the type hierarchy, locate the logo file in the right format, and apply the spacing rules from memory — the answer is no. They will use last quarter's deck, change the numbers, and the identity will decay one file at a time.
This is not a discipline failure on their part. It is a delivery failure on the studio's.
What a usable delivery contains
Templates, not just rules. A presentation template where the master slides are already correct. A document template. Social formats at the right dimensions with type and safe areas set. The default should be right so that being correct is the path of least effort.
Editable source files in open formats. If the only editable artwork lives in a file format your team cannot open, you have rented your identity rather than bought it.
Every logo variant that will actually be needed. Primary, horizontal, stacked, monochrome, reversed for dark backgrounds, a mark-only version for small sizes, favicon and app icon sizes. Each in both vector and raster. The variant that is missing is the one someone will improvise badly.
A colour palette with accessible contrast documented. Not just hex values — which pairs pass WCAG contrast, and which do not. A brand blue that fails on white is a brand blue that will produce unreadable pages for the next five years.
A type system with real hierarchy. Sizes, weights, line heights and when to use each. "Headings in Outfit, body in Plus Jakarta Sans" is a font pairing, not a system.
A working design system for web and product, if you have either — components, spacing scale, states. Not a screenshot of components: the actual tokens the front end will build against. A brand that exists only in Figma and never survives into code has been half delivered.
The multi-market problem
If you operate in more than one language, an identity that was only tested in one is not finished.
- A wordmark spaced for English can break on German compound nouns. "Nutzungsbedingungen" does not fit where "Terms" did.
- A right-to-left layout for Arabic or Hebrew is not a mirror image. The mark may need a variant, and the type system needs a script that actually has the weights you specified.
- CJK typography has different optical sizing and line-height requirements. A leading value tuned for Latin text produces cramped Japanese.
Test the identity in every script you sell in, during the design process rather than after. Retrofitting is more expensive and the result is visibly a retrofit.
What to ask for before you sign
A short list that separates a deliverable from a document:
- Which editable source files do we receive, in what formats?
- Which templates are included? Presentation, document, social, advertising, print.
- Is a web design system included, as tokens and components?
- Which scripts and languages has the type system been tested in?
- Is contrast accessibility documented for the palette?
- Is there a handover session with the people who will use it?
- Do we own it outright, with no ongoing licence?
The last one matters more than it sounds. Some studios license the identity rather than transferring it, and the terms surface when you want to change agencies.
The handover is part of the work
Every identity programme should end with a working session — not a presentation, a session — where the marketing and sales teams open the templates and make something. Whatever confuses them in that hour is what will be wrong in every asset for the next two years, and it is fixable that afternoon.
The guidelines document is the reference. The deliverable is a team that can use the system without asking.
When not to rebrand
Worth saying plainly, because agencies rarely do: if your mark has recognition, replacing it destroys value that took years to build.
The honest question is what is actually broken. Usually it is not the logo. It is that:
- there is no system around it, so nothing looks consistent
- the palette was never tested for accessibility
- there are no templates, so everyone improvises
- the assets exist in one format and one language
All four are fixable without touching the mark. That is a smaller, cheaper engagement, and any studio that pushes you toward a full rebrand without first asking what is broken is selling the larger project rather than the right one.
An evolution that keeps the equity and fixes the system is very often the correct answer. It is also the less impressive slide in a pitch, which is why it gets proposed less than it should.
