Brand architecture: main building models

Five brand architecture models and how to distinguish them from each other

In the nineties, David Aaker described brand architecture in terms of two poles: Branded House and House of Brands. In practice, companies rarely live purely on one of the poles: Marriott manages 30+ hotel brands, and within this portfolio there are monoliths, sub-brands, and endorsed models at the same time. Apple — rare example of a disciplined Branded House: iPhone, iPad, MacBook live under one name without exception, and this unity did not prevent the company from reaching a 3 trillion dollar market capitalization in January 2022. P&G — opposite extreme: 65 individual brands for 5 billion consumers, and most Pampers buyers don't even know that P&G is the owner.

Sub-brand or endorsed brand: where exactly does the boundary lie

The confusion here is not about the architecture of the portfolio as a whole, but about a specific visual weight of the parent brand name next to the product name. With a master brand, the parent brand occupies a larger portion of the carrier: the Apple logo is larger and more primary than the word 'iPhone' next to it, and the consumer reads this as a line of one company. With an endorsed brand, the proportion is reversed — the product name dominates, and the parent brand becomes a signature in small print or a separate line: 'Courtyard by Marriott' on a hotel facade looks like a standalone establishment with a quality guarantee at the bottom.

A practical test is simple: cover the parent brand name with your hand on a sign or packaging mockup. If the product is still understandable and recognizable — it's endorsed. If without the parent brand name the product loses meaning and context — it's a sub-brand.

How to understand that architecture is broken: four measurable symptoms

The problem almost never looks like an obvious crisis — it accumulates over months, and marketing long attributes the symptoms to 'weak advertising' instead of diagnosing the structure.

• In the brand recognition survey, customers confuse the company's products with one another — if more than 20% of respondents cannot correctly match the name with the category, this is an architectural, not an advertising problem;
• The marketing budget for maintaining brand recognition is growing faster than revenue — the company is paying for constant reminder of who is who, instead of sales growth;
• The sales department detects internal competition: managers of two company products are approaching the same client with different proposals;
• The launch of a new product is delayed at the 'under what name to bring it to market' stage — if this question is resolved at each meeting anew, then a portfolio policy does not exist.

Two matching points out of four — a reason to order an architecture audit not next quarter, but now.

Architecture change: a step-by-step process and the cost of error in speed

Companies change their brand model for three reasons: acquiring a new brand breaks the previous portfolio logic, the old architecture fails to explain new products, or the business enters a market where the familiar name is not readable. A sharp replacement is almost always more expensive than a gradual one: audiences adapt to a name over years, and instant substitution is perceived as the disappearance of a service rather than rebranding.

A telling case is the transition of Docdoc to SberZdorovye (17 June 2020), when the name of the medical services platform was changed. The standard cycle of architecture changes looks like this:

1. Audit of the current architecture: brand recognition survey, revenue breakdown for each brand separately, interviews with 15-20 clients on how they perceive the connections between products;
2. Choosing a new model for a horizon of 3-5 years, rather than for current quarterly sales department tasks;
3. Brand distribution in the growth/market share matrix — which names to develop, which to consolidate under a parent brand, and which to discontinue;
4. Testing the new structure on focus groups before public launch, not after;
5. Informing the audience through the parallel existence of the old and new name for no less than 3-6 months — dual signage, transition banners, letters to clients.

Two mistakes that cost the most

The first issue is mixing models without a system: part of the product line follows the Branded House logic, another part follows the House of Brands logic, and no one in the company can clearly explain why this particular choice was made. The client perceives this inconsistency faster than the marketer can come up with an explanation during the briefing, and stops trusting the brand as a whole, rather than just one product.

The second is extending a premium name into the budget segment without protective separation in the brand architecture. In the nineties, Mercedes-Benz almost undermined its premium reputation by releasing the budget A-Klasse under the same name, without a sub-brand or an endorsed layer: the market interpreted this as devaluing the entire range rather than sensibly broadening the audience. An endorsed model would have been the right solution here: a separate name with a Mercedes endorsement, rather than a monolithic brand.

Who is responsible for the architecture after the model is selected

Choosing a model is the beginning of the work, not the end of the project. Next, the architecture needs to be maintained just as regularly as brandbooks are updated, and not have to deal with it only after the next scandal. In practice, this involves three specific mechanisms:

• The portfolio owner is a person or a small team that is responsible for the consistency of all company brands, not just one product;
• Planned architecture audit once every 12-18 months, regardless of whether a crisis is imminent or not;
• The fixed product launch schedule — what name a new product is released under is decided not by the taste of an individual manager, but by a written portfolio rule;
• Unified visual and verbal standards in the brandbook: how the subbrands and the parent brand relate to each other in terms of size, color, and signature position — with no room for improvisation by branch designers.

Companies that skip governance usually return to the architectural issue within two to three years — but already in a state of crisis, when the cost of rebranding is multiple times higher than the cost of regular audit. The budget difference between planned improvements and emergency rebranding usually convinces management to keep architecture under constant control rather than postpone the issue until the next product failure.