Brand architecture sounds like a problem for conglomerates. It is actually the second product decision you make, it is made silently the day you name your second line, and reversing it later costs more than almost anything else in marketing.
Summary
Brand architecture is the decision about how many brands you operate and how much trust flows between them. A branded house puts everything under one name, so credibility transfers and marketing spend compounds. A house of brands keeps names separate, so risk is contained and each brand can target a different buyer at a different price. The right choice depends on whether your existing reputation helps or hurts the new thing you are selling.
Internet Culture · Playbook · Brand 002
Helps you decide: whether your next product gets your name, a new name, or a name attached to yours.
Why this is worth ten minutes
- Most companies do not choose a brand architecture. They accumulate one, product by product, and discover the structure years later when the marketing budget stops working.
- The real question is not organisational. It is whether trust transfers in your category, and trust transfers in some categories and actively reverses in others.
- Brand architecture determines your cost of launch forever. Under one brand, every new launch inherits awareness. Under many, every launch starts from zero.
1 The room
A direct-to-consumer brand has done well selling one thing. Coffee, say. The team wants to launch a second line: tea, or protein, or something adjacent.
Two slides appear. One says the new line should carry the existing name, because you have spent four years and a lot of money making people trust it. The other says the new line needs its own name, because the existing brand is known for one thing and stretching it will confuse people.
Both arguments are correct in general and only one is correct here. Which one depends on a question nobody in the room has asked: does being known for coffee make a customer more likely to buy tea from you, or less?
That question has a testable answer, and it is the whole of brand architecture.
2 What are the four brand architecture models?
The standard framing, drawn from David Aaker’s work on brand portfolios, gives four positions on a spectrum.
Branded house. One master brand across everything. Google. Apple. Amul. Tata. New products are described, not renamed. Amul Butter, Amul Cheese, Amul Ice Cream. Maximum compounding, maximum contagion.
Sub-brands. A master brand with named children that borrow its credibility. Titan and Tanishq sit close to this. Marriott and its tiers. The parent vouches, the child specialises.
Endorsed brands. Separate names, visibly backed. “A Tata Enterprise.” “By Marico.” The child has its own identity and the parent’s signature underwrites the risk.
House of brands. Separate names, parent invisible to the buyer. Hindustan Unilever running Dove, Surf Excel, Lifebuoy and Kwality Wall’s. Most consumers do not know they are related and the company would prefer it that way.
The brand architecture spectrum is really a single dial: how much trust and how much risk do you want to flow between your products?
3 How do you choose between a branded house and a house of brands?
Forget the models for a second. Ask three things about the new product.
Does the existing brand’s reputation help? If you are known for reliability and the new product’s main barrier is trust, use your name. You are converting an existing asset into a lower cost of acquisition. This is why a broker known for transparency can launch adjacent financial products under one name and spend almost nothing introducing them.
Would failure here damage the core? If the new product is riskier, cheaper, or aimed at a buyer who would embarrass your existing buyer, separate it. This is why premium brands launch diffusion lines under new names, and why an enterprise vendor moving downmarket often does not put the same logo on the small business product.
Is the buyer the same person? This is the one teams get wrong most often. Two products can share a category and not share a buyer. If the new product’s buyer has different reference points, different price expectations and different competitors, a single brand forces one positioning to serve two comparison sets, and it will serve neither well.
THE TRUST TRANSFER TEST
Write one sentence: “Because we are known for ________, buying ________ from us feels ________.”
If the last blank is “safer”, use your name. Trust transfers, and a branded house is a compounding machine.
If the last blank is “strange” or “cheaper than I expected” or “not really their thing”, build a separate name. You are about to spend your reputation as working capital and get a poor rate on it.
The reverse test, which is the important half: “If this new product fails publicly, what happens to the original?” If the answer is nothing, one brand is fine. If the answer is a support queue full of confused existing customers, you have found your firewall.
4 The mechanism stack
01. One brand compounds marketing spend. Many brands divide it. Every rupee spent on a branded house lifts everything in the portfolio. In a house of brands, that same rupee lifts exactly one product. This is the single strongest argument for consolidation, and for most companies below a certain size it is decisive, because they do not have the budget to build two sets of awareness.
02. One brand also compounds damage. A quality failure, a data breach or a bad news cycle in one line contaminates all of them. House of brands structures exist primarily as risk containment, which is why they dominate in categories where products fail in visible, personal ways.
03. Category associations are sticky and they set a ceiling. Once a name is filed under a category in memory, extending it outside that category requires the customer to re-file it, which is effortful and mostly does not happen. A brand strongly linked to one buying situation struggles to be retrieved in another, no matter how much you spend telling people it now does other things.
04. Price is the hardest thing to stretch. A brand extension can often work across adjacent products. It struggles enormously to stretch across price tiers, because price is a positioning signal and the customer has already learned yours. Stretching down is faster and more damaging than stretching up.
05. Brand architecture drives operating cost, not just messaging. Every separate brand needs its own creative system, its own social presence, its own team attention and its own agency relationship. Founders consistently underestimate this. The architecture you choose becomes a headcount decision within eighteen months.
5 The counter-weight
Consolidation has become fashionable and it is not always right. Renaming a group entity to a single masterbrand looks decisive in a press release. Whether it clarifies or dilutes depends on whether the consumer-facing brands keep their own equity, and that answer usually takes years to arrive.
Small companies overthink brand architecture. If you have one product and four hundred customers, you do not have an architecture problem. You have a product market fit problem wearing a costume. The question becomes real at the second product, not before.
House of brands is expensive in a way that is easy to hide. The cost does not show up as a line item. It shows up as thin teams, diluted attention and four half-built brands instead of one strong one. Unilever can afford it. A company with a fifteen person marketing team usually cannot.
Aaker’s framework predates platform businesses. Companies where the product is a marketplace, an app ecosystem or a set of interlocking services do not sort neatly into these four boxes, and forcing them to is a common source of bad workshops.
Internet Culture verdict desk
Your brand architecture is not a diagram. It is a decision about where you are willing to let one failure travel.
6 Steal this
01. Run the Trust Transfer Test on your next launch before naming anything. Twenty minutes, one sentence, two directions.
02. Map your current portfolio honestly. Draw every product, sub-brand, campaign name and internal label your customers have ever seen. Most teams discover they already operate three brands they never approved.
03. Count your buyers, not your products. If two products serve the same buyer in adjacent situations, one brand. If they serve different buyers at different price points, consider separation.
04. Estimate the second brand’s real annual cost. Creative, social, content, team time, agency fees. Compare that to the incremental revenue it protects. This kills most unnecessary sub-brands in one meeting.
05. Default to fewer brands than feels right. Consolidation is reversible. Fragmentation is expensive to undo, because merging two brands means killing recognition that someone already paid for.
7 Frequently asked questions
What is brand architecture?
Brand architecture is the structure that defines how a company organises its brands and how much credibility and risk flows between them. The four main models are branded house, sub-brands, endorsed brands and house of brands, ranging from one shared name across everything to fully separate identities.
What is the difference between a branded house and a house of brands?
A branded house uses one master brand across all products, so marketing spend and reputation compound but risk also spreads. A house of brands keeps products under separate names, so each brand can target a different buyer and a failure in one does not damage the others, but each must build awareness independently.
When should a company launch a sub-brand instead of a new brand?
Use a sub-brand when the parent’s reputation genuinely helps the new product but the new product needs its own identity to speak to a distinct buyer or price tier. Use a fully separate brand when the parent’s associations would confuse or deter the new buyer, or when failure would damage the core business.
Can a brand stretch into a new category?
A brand extension works most easily into adjacent categories where the customer already sees a logical connection and the buyer remains similar. Stretching across large price differences or into categories with different buyers usually fails, because existing memory associations pull the brand back toward its original category.
Does brand architecture matter for a startup with one product?
Not yet. Brand architecture becomes a real decision at the second product or the second buyer segment. Before that, a single-product company should focus on positioning and product market fit. The mistake is making the decision silently at launch two rather than deliberately.
