Sub-brand and product line identity
Three layers of variability: what must not be touched and what must change
A working model that is actually applied in branding practice divides the system into three levels. The first is the invariant layer: elements that remain the same for all sub-brands without exception. The second is the layer of controlled variations: what changes according to strict rules. The third is local variability: contextual solutions that are only permitted in specific media or platforms.
| Layer | What is included | Example of a rule |
|---|---|---|
| Invariant | The logo form of the master brand, font family, proportions of the sign, basic composition | The master brand sign is not scaled or distorted in any sub-brand |
| Controlled variations | Accent color, descriptor, iconography, tone of communication | Each sub-brand is assigned one accent color from a closed palette |
| Local variability | Packaging layout, adaptation for social networks, seasonal graphic inserts | Permitted is one's own illustrative style within the platform, but not a custom font |
The error of most systems is not that there are too many variations, but that there is no clear boundary between layers. A designer changes what should remain immutable (the shape of the sign) and leaves unchanged what must differ (the color). The result is that sub-brands look either like clones or like foreign companies to each other.
Descriptor: a word that encodes specialization
A descriptor is a verbal or graphical element that is attached to the name of a master brand or sub-brand and explains its specialization without creating a new name. It is not a tagline or a slogan — it is a functional label that works as a navigational indicator within the architecture.
A demonstrative example is the transition of the company Paintgroup from the 'house of brands' model to the 'brand-house' model. The historical names of the sub-brands were preserved — this was fundamentally important for the continuity of culture and audience loyalty, which over the years had associated itself with a specific name. But descriptors were added to the names, which indicated the specialization of each direction. The result: the audience does not lose its attachment to the familiar name, but receives a clear hint as to why this sub-brand exists.
A descriptor works best when it:
- describes a function or segment, not an abstract emotion;
- is visually subordinate to the master brand name — smaller font size, different font weight;
- is used systematically across all touchpoints, not only on packaging;
- does not duplicate the meaning already carried by color or iconography.
Color coding: one accent — one sub-brand
Color is the fastest visual signal, and precisely for this reason it is most often used incorrectly: shades are distributed arbitrarily, without connection to meaning. The working principle is simple: each sub-brand is assigned one accent color, while the base color of the master brand remains fixed in navigational and 'holding' elements — the logo block, header, and service graphics.
The retail network «Magnetic» built its system exactly like this: «Magnetic Cosmetics» is coded in magenta, «Magnetic Pharmacy» — in blue, «Magnetic Opt» — in orange. At the same time, the graphical structure of the logo and typography remain unified — only the accent changes. The customer instantly distinguishes the signs from one another, but does not doubt that all three points belong to the same network.
A more illustrative example of two-level encoding is the VTS system. The upper blue part of the logo indicates subordination to the master brand and remains unchanged across all sub-brands. The lower colored part is variable and distinguishes sub-brands from one another. Thus, the color works on two levels at once: the top maintains affiliation with the brand family, while the bottom differentiates within it. This is the answer to the question "How to look like a family but not twins": the fixed part of the composition maintains integrity, while the variable part maintains difference.
Case of Red Hat: three-level sub-brand hierarchy in practice
Red Hat is one of the rare examples where the sub-brand architecture is built not declaratively but engineering-wise, in layers of inheritance. The system is built from three levels.
- Platforms — four sub-brands with a full visual system: their own palette, iconography, and compositional rules. This is the top layer, visually as autonomous as possible, but structurally tied to the master brand through a shared font and graphic language.
- Products — each has its own logo and icon, but they inherit the visual language of their platform: the same principles of sign construction, the same logic of color. A product does not create a new system — it develops the platform's system.
- Components — the lowest level, which has only an icon inheriting the brand language, but without a separate logo. The component does not claim independent identity — it is dissolved within the platform's system.
The value of this case lies in the fact that it demonstrates: the lower the level in the hierarchy, the less visual autonomy is allowed. This is a direct illustration of the layers of variability model — simply applied to the scale of an entire product line, rather than to a single product.
When a sub-brand is truly needed, and when it is an unnecessary entity
A sub-brand is justified only in one case: when the target audience, value proposition, or market position genuinely differ from the parent brand — not slightly, but structurally. If the difference comes down to taste, package size, or price, a sub-brand is not needed — a modification of the umbrella brand through the same descriptive logic and local color accent is sufficient.
The problem is that sub-brands are often created not for the market, but for the company's internal organization: a new division wants its own identity, a manager wants 'their' project in the portfolio. This solves an organizational issue, but creates an external one — the consumer doesn't understand why another brand appeared if the old one perfectly covered this need. The result is budget fragmentation and dilution of the entire architecture's recognizability.
Nike and the Russian example of 'Inferit' show the opposite, working approach: a limited number of sub-brands, each with a clearly defined role and audience, without excessive splitting. This is a key practical conclusion — the system works not because of the number of sub-brands, but because of the discipline of their differentiation.
Differentiation matrix and typical architecture errors
Before drawing logos, one should build a differentiation matrix — a table where rows represent sub-brands and columns represent parameters of difference: accent color, descriptor, iconography, voice tone, packaging type. Such a matrix immediately shows gaps: if two sub-brands match on three out of five parameters, it made no sense to separate them at all.
| Error | Why it occurs | How to avoid |
|---|---|---|
| Too many sub-brands | Each division wants its own identity | Introduce a sub-brand only when audience or positioning difference is proven |
| Chaotic color coding | Color is assigned 'by eye', without a system | Closed palette, one accent per sub-brand, fixed master brand color |
| Mixing layers of variability | A designer changes the shape of the sign instead of the color | Explicitly specify what is invariant and what is governed by rules |
| A sub-brand solves an internal problem | Organizational logic is more important than market logic | Test the sub-brand hypothesis through the audience and positioning, not the company structure |
The differentiation matrix does not replace the brand book, but it is that missing tool that allows you to design the sub-brand system before money is spent on design. Without it, the brand architecture is built on intuition, and intuition does not scale well to ten product lines at once
Three layers of variability: what must not be touched and what must change
A working model that is actually applied in branding practice divides the system into three levels. The first is the invariant layer: elements that remain the same for all sub-brands without exception. The second is the layer of controlled variations: what changes according to strict rules. The third is local variability: contextual solutions that are only permitted in specific media or platforms.
| Layer | What is included | Example of a rule |
|---|---|---|
| Invariant | The logo form of the master brand, font family, proportions of the sign, basic composition | The master brand sign is not scaled or distorted in any sub-brand |
| Controlled variations | Accent color, descriptor, iconography, tone of communication | Each sub-brand is assigned one accent color from a closed palette |
| Local variability | Packaging layout, adaptation for social networks, seasonal graphic inserts | Permitted is one's own illustrative style within the platform, but not a custom font |
The error of most systems is not that there are too many variations, but that there is no clear boundary between layers. A designer changes what should remain immutable (the shape of the sign) and leaves unchanged what must differ (the color). The result is that sub-brands look either like clones or like foreign companies to each other.
Descriptor: a word that encodes specialization
A descriptor is a verbal or graphical element that is attached to the name of a master brand or sub-brand and explains its specialization without creating a new name. It is not a tagline or a slogan — it is a functional label that works as a navigational indicator within the architecture.
A demonstrative example is the transition of the company Paintgroup from the 'house of brands' model to the 'brand-house' model. The historical names of the sub-brands were preserved — this was fundamentally important for the continuity of culture and audience loyalty, which over the years had associated itself with a specific name. But descriptors were added to the names, which indicated the specialization of each direction. The result: the audience does not lose its attachment to the familiar name, but receives a clear hint as to why this sub-brand exists.
A descriptor works best when it:
- describes a function or segment, not an abstract emotion;
- is visually subordinate to the master brand name — smaller font size, different font weight;
- is used systematically across all touchpoints, not only on packaging;
- does not duplicate the meaning already carried by color or iconography.
Color coding: one accent — one sub-brand
Color is the fastest visual signal, and precisely for this reason it is most often used incorrectly: shades are distributed arbitrarily, without connection to meaning. The working principle is simple: each sub-brand is assigned one accent color, while the base color of the master brand remains fixed in navigational and 'holding' elements — the logo block, header, and service graphics.
The retail network «Magnetic» built its system exactly like this: «Magnetic Cosmetics» is coded in magenta, «Magnetic Pharmacy» — in blue, «Magnetic Opt» — in orange. At the same time, the graphical structure of the logo and typography remain unified — only the accent changes. The customer instantly distinguishes the signs from one another, but does not doubt that all three points belong to the same network.
A more illustrative example of two-level encoding is the VTS system. The upper blue part of the logo indicates subordination to the master brand and remains unchanged across all sub-brands. The lower colored part is variable and distinguishes sub-brands from one another. Thus, the color works on two levels at once: the top maintains affiliation with the brand family, while the bottom differentiates within it. This is the answer to the question "How to look like a family but not twins": the fixed part of the composition maintains integrity, while the variable part maintains difference.
Case of Red Hat: three-level sub-brand hierarchy in practice
Red Hat is one of the rare examples where the sub-brand architecture is built not declaratively but engineering-wise, in layers of inheritance. The system is built from three levels.
- Platforms — four sub-brands with a full visual system: their own palette, iconography, and compositional rules. This is the top layer, visually as autonomous as possible, but structurally tied to the master brand through a shared font and graphic language.
- Products — each has its own logo and icon, but they inherit the visual language of their platform: the same principles of sign construction, the same logic of color. A product does not create a new system — it develops the platform's system.
- Components — the lowest level, which has only an icon inheriting the brand language, but without a separate logo. The component does not claim independent identity — it is dissolved within the platform's system.
The value of this case lies in the fact that it demonstrates: the lower the level in the hierarchy, the less visual autonomy is allowed. This is a direct illustration of the layers of variability model — simply applied to the scale of an entire product line, rather than to a single product.
When a sub-brand is truly needed, and when it is an unnecessary entity
A sub-brand is justified only in one case: when the target audience, value proposition, or market position genuinely differ from the parent brand — not slightly, but structurally. If the difference comes down to taste, package size, or price, a sub-brand is not needed — a modification of the umbrella brand through the same descriptive logic and local color accent is sufficient.
The problem is that sub-brands are often created not for the market, but for the company's internal organization: a new division wants its own identity, a manager wants 'their' project in the portfolio. This solves an organizational issue, but creates an external one — the consumer doesn't understand why another brand appeared if the old one perfectly covered this need. The result is budget fragmentation and dilution of the entire architecture's recognizability.
Nike and the Russian example of 'Inferit' show the opposite, working approach: a limited number of sub-brands, each with a clearly defined role and audience, without excessive splitting. This is a key practical conclusion — the system works not because of the number of sub-brands, but because of the discipline of their differentiation.
Differentiation matrix and typical architecture errors
Before drawing logos, one should build a differentiation matrix — a table where rows represent sub-brands and columns represent parameters of difference: accent color, descriptor, iconography, voice tone, packaging type. Such a matrix immediately shows gaps: if two sub-brands match on three out of five parameters, it made no sense to separate them at all.
| Error | Why it occurs | How to avoid |
|---|---|---|
| Too many sub-brands | Each division wants its own identity | Introduce a sub-brand only when audience or positioning difference is proven |
| Chaotic color coding | Color is assigned 'by eye', without a system | Closed palette, one accent per sub-brand, fixed master brand color |
| Mixing layers of variability | A designer changes the shape of the sign instead of the color | Explicitly specify what is invariant and what is governed by rules |
| A sub-brand solves an internal problem | Organizational logic is more important than market logic | Test the sub-brand hypothesis through the audience and positioning, not the company structure |
The differentiation matrix does not replace the brand book, but it is that missing tool that allows you to design the sub-brand system before money is spent on design. Without it, the brand architecture is built on intuition, and intuition does not scale well to ten product lines at once
Recent articles
MegaFon, BMW, VTB and more 240+ companies. See how it looks in the showcase reel:
Our branding agency is approached to order:
Branding agency services
