Brand architecture: monolith, sub-brands, hybrid
Three models instead of one formula
Before examining selection criteria, we need to distinguish three basic logics. Monolith (Branded House) — all products live under one name and one visual system, as with Apple: iPhone, iPad, MacBook are sub-products of one promise. Brand portfolio (House of Brands) — each product gets its own name, its own consumers, often its own design system, and the link to the parent company is hidden or minimal — the classic P&G model with Tide, Pampers, Gillette. Hybrid combines both principles within one company, applying different models to different parts of the business.
| Model | Logic | Example | Main risk |
|---|---|---|---|
| Monolith | One brand — all products | Apple, Google | Failure of one product affects all |
| Brand portfolio | Each product — its own name | P&G, Marriott | High support cost (Brand Tax) |
| Hybrid | Different models for different segments | Amazon, Microsoft | «Drift» in architecture without documented rules |
Selection criteria: where the math converges
The key criterion is the linkage of products by demand and by supply. Demand linkage means that the same buyer purchases both products in a similar context. Supply linkage means that the products are produced using common technologies, capacities, and R&D. It would seem that the higher both linkages, the more logical it is to combine everything under one name — this is the reasoning of most articles in the top of the search results. But there is a paradox here that is rarely explained.
At P&G, cleaning products are linked both by demand and by supply — the same household, the same chemical base. It is precisely this high degree of linkage that made the company separate the brands into individual names rather than unite them. The logic is simple: if Tide and, say, a less successful detergent were to coexist under one name, management would be tempted to cut costs on the quality of the second product at the expense of the reputation of the first. Separate brands here work as a disciplining mechanism — they prevent internal savings from turning into reputational debt. Apple, on the contrary, keeps all its products in a monolithic structure precisely because their linkage is technological and systemic (a single iOS ecosystem, a unified design language), not production-based in the style of 'one factory, different formulas'.
The second criterion is budget. A brand portfolio requires money for each individual marketing, SEO, content, and design system — this is the so-called Brand Tax, about which hardly anyone writes in general articles. For an undercapitalized business in the digital age, maintaining ten separate brands is economically unprofitable: each requires its own traffic and brand recognition from scratch. The third criterion is risk: the higher the cost of error for one product to the reputation of the entire company, the stronger the argument for separating the brands.
When a hybrid is not a choice from an abundance of options, but the only viable variant
Hybrid architecture is often described as a compromise for those who couldn't make up their minds. This is incorrect. There are at least three situations where a hybrid is the only working model, not an intermediate option.
- M&A. The acquired company comes with its own recognizability and client base — cutting it down to the parent brand means losing the transaction's value. Amazon did not rename Whole Foods and Zappos to "Amazon Foods" — the recognizability of these brands was part of the asset for which money was paid.
- Transition periods. When a company changes its positioning or enters a new market, full rebranding in one step is risky. A sub-brand with support from the parent name (endorsed branding) gives time to build trust without losing connection to the main brand.
- A complex product line with different audiences. Amazon Prime, Video, Kindle — consumer-facing sub-brands, while AWS is a standalone brand for B2B-audience with a completely different decision-making cycle. Mixing their logic under one architecture would mean ignoring the difference in consumer behavior.
Microsoft has a similar setup, but it's organized differently: Office and Azure operate as a monolith, LinkedIn has a supported status (the parent brand is visible but not dominant), while Xbox and Minecraft remain fully independent. This is not chaos, but the result of different decisions for each segment — and that is exactly the essence of a healthy hybrid.
Reputational risks: why a crisis of one brand can destroy an entire portfolio
The tighter the architecture, the higher the correlation of reputations. In a monolith, a scandal around one product automatically becomes a scandal around the entire company — headlines are not written as "problem with model X", but as "problem with brand Y". This is the price a monolith pays for saving on Brand Tax.
In a well-constructed brand portfolio, a crisis of one product is theoretically isolated — a consumer of Pampers may not know that the same P&G produces a problematic product under a different name. But this only works if the link to the parent company is truly private. If the press and exposing journalists quickly connect the brands to the corporate owner — the risk isolation stops working, and the company suffers damage as "that same corporation that was hiding it".
From this practical conclusion: the decision to disclose or conceal the maternal brand must be a conscious element of the architecture, not a coincidence. This is especially sensitive in B2B: there the buyer almost always sees and evaluates the corporate brand directly, so a clean House of Brands in B2B is rare — Branded House or Subbranded model is usually used, where the connection to the parent company works as a reliability guarantee, not a risk.
Real mistakes in building the architecture
Most architecture failures are not about the wrong choice of model, but about the lack of discipline in its application.
- Architecture without rules for decision-making. The company chooses a hybrid, but does not document by what principle a new product gets a sub-brand, and which gets an independent name. In two to three years, nobody in the company can explain the logic of the current structure — this is what "architecture drift" is about, which EquiBrand Consulting writes about.
- Ignoring Brand Tax when launching new products. The launch team launches a separate brand "for differentiation", not counting the cost of a separate design system, SEO from scratch, and paid channels — and a year later it turns out that the budget for supporting two brands is not enough for either one.
- Mechanical copying of someone else's architecture. A startup looks at Apple and decides to make a monolith, without having the technological linkage of products that justifies this model for Apple. Or vice versa — copies the portfolio model of P&G, without having the resources to maintain several brands.
- Silence about the parent brand where the market itself will uncover it. Especially dangerous in B2B and in the era of active social networks — an attempt to hide the connection between brands, when it is easily established, reads as deception, not as a reasonable strategy.
How a hybrid is managed in practice
Hybrid architecture is the most complex to administer, and precisely for this reason it often "drifts" without explicit control. A working hybrid is sustained by three tools.
First — a document with criteria for classifying a new product as one of the models within the portfolio: demand and supply linkage with the existing product line, target audience size, strategic value of brand recognition in case of asset sale. Second — regular audit of boundaries: once every one to two years, it is necessary to review whether individual sub-brands have outgrown their niche and whether it is time for them to either merge with the parent brand or, conversely, acquire full independence. Third — the budgetary Brand Tax model: every new independent brand in the portfolio must have a calculated support cost for 2-3 years ahead, otherwise the decision to launch it is made blindly.
For startups and growing companies, the rule is simple: at early stages, a monolith almost always wins — there's no money for Brand Tax, and recognition needs to be built from scratch once, not in parallel for several names. Portfolio and hybrid architecture only becomes justified when a company accumulates products with different audiences, different pricing positioning, or assets acquired through M&A — that is, the architecture grows along with the business, not the other way around.
Three models instead of one formula
Before examining selection criteria, we need to distinguish three basic logics. Monolith (Branded House) — all products live under one name and one visual system, as with Apple: iPhone, iPad, MacBook are sub-products of one promise. Brand portfolio (House of Brands) — each product gets its own name, its own consumers, often its own design system, and the link to the parent company is hidden or minimal — the classic P&G model with Tide, Pampers, Gillette. Hybrid combines both principles within one company, applying different models to different parts of the business.
| Model | Logic | Example | Main risk |
|---|---|---|---|
| Monolith | One brand — all products | Apple, Google | Failure of one product affects all |
| Brand portfolio | Each product — its own name | P&G, Marriott | High support cost (Brand Tax) |
| Hybrid | Different models for different segments | Amazon, Microsoft | «Drift» in architecture without documented rules |
Selection criteria: where the math converges
The key criterion is the linkage of products by demand and by supply. Demand linkage means that the same buyer purchases both products in a similar context. Supply linkage means that the products are produced using common technologies, capacities, and R&D. It would seem that the higher both linkages, the more logical it is to combine everything under one name — this is the reasoning of most articles in the top of the search results. But there is a paradox here that is rarely explained.
At P&G, cleaning products are linked both by demand and by supply — the same household, the same chemical base. It is precisely this high degree of linkage that made the company separate the brands into individual names rather than unite them. The logic is simple: if Tide and, say, a less successful detergent were to coexist under one name, management would be tempted to cut costs on the quality of the second product at the expense of the reputation of the first. Separate brands here work as a disciplining mechanism — they prevent internal savings from turning into reputational debt. Apple, on the contrary, keeps all its products in a monolithic structure precisely because their linkage is technological and systemic (a single iOS ecosystem, a unified design language), not production-based in the style of 'one factory, different formulas'.
The second criterion is budget. A brand portfolio requires money for each individual marketing, SEO, content, and design system — this is the so-called Brand Tax, about which hardly anyone writes in general articles. For an undercapitalized business in the digital age, maintaining ten separate brands is economically unprofitable: each requires its own traffic and brand recognition from scratch. The third criterion is risk: the higher the cost of error for one product to the reputation of the entire company, the stronger the argument for separating the brands.
When a hybrid is not a choice from an abundance of options, but the only viable variant
Hybrid architecture is often described as a compromise for those who couldn't make up their minds. This is incorrect. There are at least three situations where a hybrid is the only working model, not an intermediate option.
- M&A. The acquired company comes with its own recognizability and client base — cutting it down to the parent brand means losing the transaction's value. Amazon did not rename Whole Foods and Zappos to "Amazon Foods" — the recognizability of these brands was part of the asset for which money was paid.
- Transition periods. When a company changes its positioning or enters a new market, full rebranding in one step is risky. A sub-brand with support from the parent name (endorsed branding) gives time to build trust without losing connection to the main brand.
- A complex product line with different audiences. Amazon Prime, Video, Kindle — consumer-facing sub-brands, while AWS is a standalone brand for B2B-audience with a completely different decision-making cycle. Mixing their logic under one architecture would mean ignoring the difference in consumer behavior.
Microsoft has a similar setup, but it's organized differently: Office and Azure operate as a monolith, LinkedIn has a supported status (the parent brand is visible but not dominant), while Xbox and Minecraft remain fully independent. This is not chaos, but the result of different decisions for each segment — and that is exactly the essence of a healthy hybrid.
Reputational risks: why a crisis of one brand can destroy an entire portfolio
The tighter the architecture, the higher the correlation of reputations. In a monolith, a scandal around one product automatically becomes a scandal around the entire company — headlines are not written as "problem with model X", but as "problem with brand Y". This is the price a monolith pays for saving on Brand Tax.
In a well-constructed brand portfolio, a crisis of one product is theoretically isolated — a consumer of Pampers may not know that the same P&G produces a problematic product under a different name. But this only works if the link to the parent company is truly private. If the press and exposing journalists quickly connect the brands to the corporate owner — the risk isolation stops working, and the company suffers damage as "that same corporation that was hiding it".
From this practical conclusion: the decision to disclose or conceal the maternal brand must be a conscious element of the architecture, not a coincidence. This is especially sensitive in B2B: there the buyer almost always sees and evaluates the corporate brand directly, so a clean House of Brands in B2B is rare — Branded House or Subbranded model is usually used, where the connection to the parent company works as a reliability guarantee, not a risk.
Real mistakes in building the architecture
Most architecture failures are not about the wrong choice of model, but about the lack of discipline in its application.
- Architecture without rules for decision-making. The company chooses a hybrid, but does not document by what principle a new product gets a sub-brand, and which gets an independent name. In two to three years, nobody in the company can explain the logic of the current structure — this is what "architecture drift" is about, which EquiBrand Consulting writes about.
- Ignoring Brand Tax when launching new products. The launch team launches a separate brand "for differentiation", not counting the cost of a separate design system, SEO from scratch, and paid channels — and a year later it turns out that the budget for supporting two brands is not enough for either one.
- Mechanical copying of someone else's architecture. A startup looks at Apple and decides to make a monolith, without having the technological linkage of products that justifies this model for Apple. Or vice versa — copies the portfolio model of P&G, without having the resources to maintain several brands.
- Silence about the parent brand where the market itself will uncover it. Especially dangerous in B2B and in the era of active social networks — an attempt to hide the connection between brands, when it is easily established, reads as deception, not as a reasonable strategy.
How a hybrid is managed in practice
Hybrid architecture is the most complex to administer, and precisely for this reason it often "drifts" without explicit control. A working hybrid is sustained by three tools.
First — a document with criteria for classifying a new product as one of the models within the portfolio: demand and supply linkage with the existing product line, target audience size, strategic value of brand recognition in case of asset sale. Second — regular audit of boundaries: once every one to two years, it is necessary to review whether individual sub-brands have outgrown their niche and whether it is time for them to either merge with the parent brand or, conversely, acquire full independence. Third — the budgetary Brand Tax model: every new independent brand in the portfolio must have a calculated support cost for 2-3 years ahead, otherwise the decision to launch it is made blindly.
For startups and growing companies, the rule is simple: at early stages, a monolith almost always wins — there's no money for Brand Tax, and recognition needs to be built from scratch once, not in parallel for several names. Portfolio and hybrid architecture only becomes justified when a company accumulates products with different audiences, different pricing positioning, or assets acquired through M&A — that is, the architecture grows along with the business, not the other way around.
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
