Skip to content
Faisal Mahmud TusherDigital marketing, SEO & analytics
with Faisal Tusher

Branding › Brand Foundations

Brand Architecture

How brand architecture works: branded house, house of brands, endorsed and hybrid models, sub-brands and product naming, and when to add or retire a brand.
Guideintermediate9 min readUpdated

Why architecture matters

Every pound a company spends on communication builds memory for some name. Architecture decides which one. A company with one master brand pours all of its spending into a single memory structure, and every product launch, every advert and every piece of press adds to it. A company with a dozen separately named brands spreads the same spending across a dozen structures, each of which has to reach (Reach) category buyers often enough to be remembered on its own. Neither is wrong; each has a price, and the price is usually paid without anyone having chosen it.

Architecture also decides what credibility transfers. A new product under a trusted master brand starts with the master brand's associations, good and bad. A separately branded product starts clean, which is a cost when the master brand is strong and an advantage when the new product would damage it, serves a different audience or competes at a different price. The decision is therefore never purely about marketing efficiency; it is about which associations the company wants to flow where, which is a question the brand strategy guide answers for each brand and this guide answers across them.

Most companies arrive at their architecture by accident: a founder's product name becomes the company name, an acquisition keeps its old name, a product team invents a brand because it can. The result works until it does not, typically when customers cannot tell which things belong together, or when the marketing budget is split so many ways that no brand is funded to be remembered.

The models

The models sit on a spectrum from one name to many. The table gives the usual four and the trade-offs each carries.

ModelHow it worksStrengthsCostsFits when
Branded houseOne master brand on everything; products are descriptorsAll spending builds one memory; new products launch with borrowed trustOne failure taints everything; hard to serve very different audiences or price pointsOne audience, one promise, limited product range
House of brandsSeparate brands with their own identities; the parent is invisible or minorEach brand can own a position and audience; failures are containedEach brand needs its own budget to be remembered; no shared equityDifferent audiences, price tiers or categories under one owner
EndorsedEach brand has its own name, visibly backed by the parentBrands get their own position plus borrowed credibilityEndorsement must be managed; confusion if the parent's meaning is unclearAcquisitions, premium or specialist lines, markets where the parent is trusted
HybridMaster brand for the core, separate or endorsed brands where neededFlexibility; the core keeps its concentrationNeeds explicit rules or it decays into accidentGrowing companies with one strong core and a few exceptions

The choice is not permanent, but it is expensive to reverse, because changing architecture means re-educating customers about what belongs together. The default for most companies, and particularly for small ones, is the branded house: concentration builds memory fastest, and a single identity is cheaper to maintain. The case for separation has to be made on evidence that the audiences, promises or price points genuinely conflict, not on a product team's wish for its own logo. Share of voice (Share of Voice) is the clarifying lens: each separately named brand needs its own share of category voice to grow, and a company that cannot fund that for every brand should have fewer of them.

Sub-brands and naming

Between a descriptor and a separate brand sits the sub-brand: a named product or range that carries the master brand but has some identity of its own. Sub-brands are where most architecture problems start, because they are cheap to create and each one claims a slice of memory. A naming rule set prevents most of them. The same logic decides whether a sub-brand gets social accounts of its own, which the guide to structuring social accounts works through.

A descriptive name (the master brand plus a plain description of what the product does) builds the master brand and asks nothing of the customer. A sub-brand name asks the customer to remember a second thing, and is justified only when the product needs its own associations: a different audience, a distinct promise or a long enough life to repay the investment. A separate brand is justified when the product should not share the master brand's associations at all. The rule set should state which of the three is the default, who can authorise an exception and what evidence they need. The same discipline applies to the rest of the brand's rules, and the guide to handling exceptions and changes to brand guidelines sets decision rights for each element, from open template areas up to the committed assets only leadership may change.

Naming itself has practical constraints that are easy to forget in a workshop: the name has to be available as a trademark in the markets that matter, pronounceable by the audience, searchable without colliding with something larger, and distinguishable from sibling names in a list. A family of sub-brands with similar prefixes looks tidy on an organisation chart and confuses customers at a glance. For companies selling through marketplaces and catalogues, naming also governs how products appear in listings and feeds, which the e-commerce marketing hub covers; a name that reads clearly in a product title is worth more than one that only works on a logo. The step-by-step tutorial on naming a product or brand turns these constraints into a brief, a screening grid and a buyer test, with legal clearance advice taken before anything is committed.

Adding, merging and retiring brands

Brands are added through launches and acquisitions, merged when a portfolio has grown unwieldy, and retired when they no longer earn their budget. Each decision has a method.

Adding. Before a new brand is created, the question is whether the master brand can stretch to cover the new product. Stretch fails when the new product's audience would not trust the master brand, when its price point would confuse the master brand's meaning, or when its failure would damage the core. If none of those applies, a descriptor or sub-brand is cheaper and builds more memory. If one does, the new brand needs its own strategy, identity and budget from the start, and its launch belongs in a go-to-market plan that treats it as a brand launch, not a product update.

Acquiring. An acquired brand arrives with its own memory structure, which is often the thing that was bought. The options are to keep it as a separate brand, endorse it with the acquirer's name, migrate it over a stated period or absorb it immediately. Keeping it preserves the equity and the budget split; absorbing it captures the budget and risks losing the customers who bought the old name. The honest input is a measure of how much the acquired name is worth to its customers, which a tracking survey can estimate before the decision rather than after. Once the decision is made, the tutorial on combining marketing after a merger or acquisition carries it through the two programmes' websites, accounts, lists, calendars and baselines.

Merging and retiring. A brand is a candidate for retirement when its share of category voice is too small to grow, when its audience is served as well by another brand in the portfolio, or when its upkeep costs more than it returns. Retirement is a migration: a period of endorsement or dual branding, communication to the customers who will notice, and a date after which the old name stops appearing. The architecture KPI (KPI) for the portfolio as a whole is simple to state: fewer brands, each funded to be remembered, with no customer unsure which belongs to whom. What each decision means for the brands' websites, when a brand needs a domain of its own and how to merge brand sites without losing search visibility, is covered in the guide to microsites and multiple domains.

Illustrative example — not a real result

A kitchenware group simplifies a portfolio of six brands

A kitchenware group with revenue of £48 million sells under 6 brand names acquired over 15 years, each with its own website, packaging and small advertising budget. A tracking survey of 1,200 category buyers finds that 2 of the names have unaided awareness above 20%, 3 sit below 5% and 1 is unknown to 97% of respondents. Combined marketing spend is £2.4 million, split 6 ways. The group moves to a hybrid model: the strongest name becomes the master brand for cookware and bakeware, the second-strongest is kept as an endorsed premium line, and the remaining 4 are migrated over 18 months with dual-branded packaging and redirected websites. Marketing spend is concentrated into 2 budgets. Two years later the master brand's unaided awareness has risen from 24% to 37%, the group runs 2 websites instead of 6, and the retired names account for under 3% of branded searches.

Architecture for growing companies

A growing company does not need a full architecture until something forces the question, but it needs the rule that says what the default is and who decides exceptions. Written early, the rule is a paragraph: "We are a branded house. New products take a descriptive name under the master brand. A sub-brand or separate brand needs a case, approved by the leadership team, showing that the audience, promise or price point conflicts with the master brand." That paragraph prevents the accidental portfolio that otherwise accumulates.

The forcing events are predictable: the second product line, the first acquisition, the first move into a market where the company name means nothing or something unhelpful, and the first product at a very different price. Each is the moment to revisit the rule, and each is a point where the brand identity and voice guide has to decide how much of the identity system the new brand shares. Shared typography and voice with a different colour and mark is a common and workable middle ground; a wholly separate identity is a commitment to a second budget.

The brand foundations hub places architecture after strategy and identity for a reason: a company that has fixed what its main brand means and how it looks is in a far better position to decide what should and should not carry that name. Companies that do architecture first tend to design portfolios for products that do not yet exist.

Common mistakes

  • Architecture by accident. Names accumulate because nobody wrote the default.
  • A separate brand for every product team. Each one splits the budget and the memory.
  • Confusing a descriptor with a sub-brand. A sub-brand asks the customer to remember a second thing; most products do not deserve that.
  • Keeping acquired names out of sentiment. The decision needs a measure of what the name is worth to its customers, not to its founders.
  • Retiring a brand overnight. Migration is a period, with endorsement, communication and a date.
  • Endorsement without a clear parent. Backing a brand with a name that means nothing adds clutter, not credibility.
  • Designing a portfolio for the future. Architecture should follow the products that exist and the rule for the next one.

Frequently asked questions

Which brand architecture is best for a small company?

Almost always a branded house: one name on everything, with products described rather than named. Concentration builds memory fastest and a single identity is cheapest to maintain. The case for a separate brand has to rest on a real conflict of audience, promise or price point, not on a product team wanting its own logo.

What is the difference between a sub-brand and a descriptor?

A descriptor is the master brand plus a plain description of what the product does, and it asks nothing new of the customer's memory. A sub-brand has a name and some identity of its own under the master brand, and it asks the customer to remember a second thing. Sub-brands are justified when the product needs its own associations and will live long enough to repay the investment; most products do not.

How should an acquired brand be handled?

By measuring what the name is worth to its customers before deciding, rather than after. If the acquired name carries real awareness and trust in its audience, keep it or endorse it with the acquirer's name and migrate only if the strategy demands it. If it carries little, absorb it through a stated migration period with dual branding, so that customers are not surprised and the budget is concentrated sooner.

Terms used on this page

Where this page is used

Part of these learning paths

Also related

Related