The brand-system audit we run before any visual work.

A brand system audit is the first thing we do on any project, and we do it before anyone opens a design file. Not a mood board, not a logo exploration, not a homepage concept. An audit. Because visual work built on an unexamined brand is decoration, and decoration falls apart the moment it meets a second touchpoint.

The point of the audit is to see the whole system at once: what the brand says, how it sounds, how it looks, and how all of that is actually built and shipped. Most brands are not broken in one place. They are slightly wrong in ten places, and the ten small wrongs add up to a brand that feels amateur without anyone being able to say why.

Below is the checklist we run, dimension by dimension. For each one, I have described what good looks like and the red flags that tell you the foundation needs work before any pixels get pushed.

Positioning and message architecture

Everything downstream inherits from positioning, so it is audited first. If a brand cannot say who it is for, what it does, and why it is different in three plain sentences, no type scale will save it.

We look for a message architecture: a ranked structure of what the brand says. A primary promise, a small set of supporting pillars, and proof beneath each pillar. Ranked, not a flat list. When everything is equally important, nothing is, and the design ends up giving equal weight to a tagline and a footnote.

• ▸Good: one primary promise, three to five supporting pillars, evidence under each, and a clear audience it is written for.
• ▸Red flag: five taglines and no hierarchy, or messaging that would fit any competitor if you swapped the logo.
• ▸Red flag: the team cannot agree on the one-sentence version, which means the disagreement will resurface as design revisions later.

Verbal identity: voice, naming, and tone

Verbal identity is audited before visual identity because words carry the brand into far more places than the logo ever will. Support replies, error messages, button labels, and email subject lines are all brand surfaces, and they are usually written by people who have never seen the guidelines.

We check that voice is defined as a set of decisions, not adjectives. "Confident" is not a voice. "We make claims and back them with numbers, we never hedge, and we do not use exclamation marks" is a voice, because someone can follow it.

• ▸Good: a documented voice with do and do-not examples, a tone that flexes by context, and consistent naming for products and features.
• ▸Red flag: the same feature called three different names across the site, the app, and the invoice.
• ▸Red flag: microcopy that swings from stiff legalese to breezy slang between two adjacent screens.

Visual identity: logo, type, color, spacing, grid

Now the visual layer, and we audit it as a system, not as a collection of nice assets. The question is never whether the logo looks good. It is whether the whole kit holds together under real conditions and scales without a designer in the room.

Logo system

A single logo file is not a system. We look for the full set: primary, secondary, and a mark that survives at sixteen pixels. Clear-space rules, minimum sizes, and behavior on light, dark, and busy backgrounds. If there is only one lockup, it will get stretched, recolored, and placed on a photo it was never meant to sit on.

Type scale, color, spacing, and grid

These four are the load-bearing structure of any interface, and they are where most brands quietly fail. We want a modular type scale with defined roles, not a folder of random sizes. We want color defined as tokens with documented roles and states, not a swatch image someone eyedrops from. And we want a spacing scale and grid that the whole team actually uses.

• ▸Good: a type scale of six to nine steps with named roles, line-heights, and weights that pair to a real hierarchy.
• ▸Good: color as tokens (surface, text, accent, and states) with contrast checked, not a raw palette of twenty hex values.
• ▸Good: a spacing scale on a consistent base, such as a four or eight pixel rhythm, applied across every layout.
• ▸Red flag: fourteen font sizes across the site, none of which are the same by accident.
• ▸Red flag: greys that multiply into a dozen near-identical values because nobody owns the token list.

Asset and component inventory

You cannot fix what you have not counted. We inventory every recurring element: buttons, forms, cards, navigation, icons, and the ten variants of each that crept in over two years. The goal is to see the sprawl in one place.

The usual finding is that a brand has seven button styles doing the work of two. The audit surfaces every variant, and the redesign collapses them back to a defensible set. A component that exists in one honest version is a component the whole team can trust.

Accessibility and contrast

Accessibility is audited as part of the brand, not bolted on at the end. A palette that fails contrast is a broken palette, however pretty the swatches look in isolation. This is a standard, not an opinion.

• ▸Good: text meets WCAG AA contrast against every surface it lands on, including hover and disabled states.
• ▸Good: focus states are visible, hit areas are large enough, and meaning is never carried by color alone.
• ▸Red flag: light grey body text on white because it looked elegant in the concept and unreadable in production.
• ▸Red flag: an accent color that passes on one background and fails on the three others it actually appears on.

Consistency across touchpoints

A brand is not the guidelines document. It is what a real person meets across the website, the product, the app store listing, the invoice, the sales deck, and the cold email. We audit those surfaces side by side, because that is where the gap between intention and reality shows up.

A brand is not what you designed. It is the average of every place a customer actually meets it, including the ones you forgot you owned.

Elena Roa, Design Director at Aivento

The red flag here is a beautiful homepage attached to a product UI that looks like a different company built it. That gap tells you the system lives in a PDF, not in the tools the team ships from.

The tech and handoff layer

This is the dimension most brand audits skip, and skipping it is why so many rebrands decay within a year. A brand only holds if it is built into the code, not just the guidelines. We audit whether the design decisions exist as design tokens and a real component library that engineering uses.

1. 01Confirm design tokens are the single source of truth for color, type, and spacing, shared by design and code.
2. 02Check that the component library in the design tool maps one-to-one to the components in the codebase.
3. 03Verify that changing a token updates every instance, so the brand can evolve without a manual sweep.
4. 04Look for documentation that a new engineer can follow without asking a designer what the primary button is.

When this layer is healthy, consistency is nearly automatic, because the correct choice is the easiest one to make. When it is missing, every developer reinvents the brand slightly, and the ten small wrongs return within two sprints.

Run the audit before you design

You can run this checklist yourself. Work through the eight dimensions, mark each as solid, shaky, or missing, and you will know exactly where the foundation needs work before a single screen gets drawn. The audit is not glamorous, but it is the difference between a redesign that lasts and a repaint that peels.

If you would rather have a senior team run the audit and rebuild the system properly, that is the brand and web design work Aivento does.