Build a color system that can scale from the start

A brand starts with one product line and a tidy palette. Two years later it launches a second line and assigns it a new color. In year three it adds a third line, and yet another color.

By year five, look at the whole portfolio: five dominant colors, none of which can talk to the others.

You're explaining to your team why the orange of line A and the orange of line C are different, even though they look almost the same.

This is not an aesthetic problem. It is an architecture problem. The palette breaks down because it has no tiers, and no rule distinguishes what is fixed from what is allowed to change.

Why brand color falls apart

The common approach: every time a new line launches, pick a color that looks "right for that line." Nobody asks how this color relates to the existing palette.

After a few rounds, the palette is no longer a system. It is a collection of separate decisions, each reasonable at the time, but with nobody looking at the whole picture.

The distinctiveness of a brand color is measured by Fame and Uniqueness (according to researcher Jenifer Romaniuk). Fame is the degree to which customers associate that color with the brand.

Uniqueness is how distinct the color is within the industry. When the palette breaks down, both of these metrics drop: nobody remembers which color is "the brand's color" anymore.

The gap between "having a palette" and "a palette that works in operation" usually comes down to this: the guidelines only describe the current colors, not the rules for adding colors later.

Tiering: the foundation of a system that doesn't break

A scalable color system needs at least two clearly separated tiers.

Tier one: foundation colors. These are the brand identity colors of the parent brand. They usually include one to two primary colors and one to two neutral supporting colors.

This tier does not change when you add product lines. Every line must use this tier's colors at some point, even if only on packaging or communication materials.

Tier two: product-line colors. Each line has its own primary color drawn from this tier. Tier two is allowed to change and expand.

Adding a new line means adding one color to tier two, without touching tier one.

Distinctive assets only build up through consistent, repeated use. Constant change burns what you've accumulated.

Jenifer Romaniuk, Building Distinctive Brand Assets

The problem for most brands whose palettes are falling apart is that they put product-line colors in the place of foundation colors. When a new line is added, the foundation color gets overridden or diluted.

The process of building color tiers before a new line launches

The first step is taking stock of what you already have. List every color in use across all touchpoints: packaging, web, social media, sales materials.

Specify which color appears where, and how often.

The second step is classification. A color that always appears regardless of the line is a candidate for tier one. A color tied to only one specific line belongs to tier two.

The third step is to check the distance. Are the tier-two colors different enough from each other when placed side by side?

In color design, “different enough” can be measured by the difference in hue angle on the color wheel.

You don't need an absolute number, but you do need a clear rule. For example: two product-line colors must not sit closer than 30 degrees apart on the HSL color wheel.

The fourth step is setting the rules for expansion. When you launch a new line, what conditions must a new color meet? Does it hold up on white and on a tier-one background color?

Does that color still work at small sizes (a 200-pixel thumbnail, an app icon)?

On a physical shelf, color is the first signal shoppers recognize, before they read the name. When the palette breaks down, that signal stops working.

Setting up a color dictionary: the foundation for not repeating mistakes

A tiering rule only survives if it's written into a document people can actually use, not one made for display.

In digital system design (design systems), the concept of the design token (a unit of design value) solves this problem: instead of recording colors in each individual file, all colors are stored in one place as variables.

When something needs to change, change it once and everywhere updates. The W3C Design Tokens Community Group standard has been stable since October 2025.

For a brand with only packaging and printed documents, tokens in the technical sense are unnecessary.

But the principle is not overkill: every color in the system must have a clear name, an exact value (HEX code, CMYK for print, Pantone if needed), and a documented role.

• Name colors by role, not by feeling. "Primary foundation color" is clearer than "signature green" when handing over to a new team.
• Specify which color is used at which tier, and under what conditions.
• Specify which colors must NOT be used together, and why.
• Update the documentation every time a new product line launches, not later.

When to start over instead of just adding on

You can't always add to an existing system. In some cases, starting over is the more sensible choice.

Sign one: your current foundation color is no longer distinctive in your industry. Many competitors use nearly the same color, and customers no longer associate that color with your brand.

The second sign: the business strategy shifts so much that the parent brand needs a complete repositioning, not just a wider range.

Sign three: the current color layers are so tangled that you can no longer tell one layer from another, and the cost of reorganizing is higher than the cost of a fresh, structured rebuild.

If none of the three signs above applies, the problem isn't bad color. The problem is a missing architecture. Fix the architecture, not everything.

A brand is the sum of product, environment, communication and behavior. Color does not stand alone: it operates within that system.

Wally Olins, On Brand

What Sinh Vũ sees again and again in practice

In most cases Sinh Vũ takes on, the problem does not come from a brand that launched a new line and then ran into trouble.

It's a brand that launched new lines without a process, and by the time it noticed, it had three or four lines with unrelated palettes.

At that point there are two paths. The first is to accept the cost of rebuilding the entire identity system.

The second is to build the architecture from what you already have, accepting that some colors will need slight adjustments to fit the right tier.

The second path is usually more economical and less risky, but it requires a clear decision about what is fixed and what is flexible.

It is strategic work before it is design work.

If you are preparing to launch a second or third line, this is the best time to ask this question. Not after you already have six lines and the palette has broken down.