Purpose of this guide. This guide explains when to use a single master brand, an endorsed brand, a sub-brand or a standalone brand, and how to manage the transition without losing brand equity, search visibility, customers or rights.
Portfolio architecture is not an illustrated organisation chart. Customers rarely care about internal divisions; they want to know what they are buying, whom they can trust and whether a new offer will deliver the promise they already know.
The problem emerges as the company grows. One product becomes a range. An acquisition keeps its name. A team launches a label. Three offers carry almost the same name, while two others appear to come from entirely different businesses. Costs spread across websites, content, campaigns, domains, brand guidelines, contracts and tools.
1. Key figures: a dense namespace and limited memory
The World Intellectual Property Organization recorded 15.2 million trademark filings by class in 2024. The five most active offices accounted for 61.5% compared with 47% a decade earlier. Adding a name is therefore never free: prior-rights searches, protection, monitoring and the ability to stand out add to the creation budget.
The Madrid system recorded 64,150 international applications in 2025. French applicants filed 4,026. These figures do not measure the value of an architecture, but they illustrate the dense legal environment in which each new brand must operate.
An academic study published in 2026 analysed 1 162 distinctive assets, in 21 categories, four countries and on nine years. It demonstrates why attribution must be measured actively by asset and market by market: a colour or shape mentally available in a country is not automatically available elsewhere.
Research from the Ehrenberg-Bass Institute highlights a counter-intuitive finding: the major brands do not necessarily have more assets, but concentrate their exposure on a stronger and more widely recognised set. Brand architecture should support this cumulative effect. Ten weakly supported names do not become a strong brand simply by being added together.
2. The question that architecture must solve
The right model maximises three values under four constraints.
Factors to optimise include trust transfer, clarity of the offer and investment efficiency. Constraints include reputational risk, distance between propositions, available rights and the practical capacity to support each entity.
The decision can be formulated as follows: What needs its own place in memory to support purchase, and what would benefit from sharing the same brand memory?
This question avoids two reflexes: grouping everything for internal convenience, or creating a standalone brand simply to make a launch appear more important. Neither reflects audience behaviour.
3. The four basic models
Real architectures are hybrid, but four models provide a common grammar.
3.1. The monolithic brand
A parent brand carries the whole: unique name, descriptive offers, strongly shared identity. Trust transfer is fast and investments are cumulative.
This model is appropriate when offers share a promise, close audiences, experience standards and a comparable level of risk. It simplifies SEO, governance, recruitment and cross-selling.
Its weakness is reputational spillover. An incident affects the whole. A very different offer can dilute perception. The descriptors must then clarify without creating a proliferation of pseudo-brands.
3.2. The sub-brand
The main name remains visible, accompanied by a proper name: Make X. The sub-brand gains a distinct personality and proposal while borrowing trust.
This option is useful for an important range, innovation or segment that requires its own language. It costs more than one descriptor, because it is necessary to build memory for two brand cues and their relationships. If the second name helps neither navigation nor preference, it becomes an expensive decoration.
3.3. The endorsed brand
The entity has its own name, with an explicit endorsement: "by...", "a company...", signature or graphic system. Commercial autonomy remains, while the group provides insurance.
The acquisition of a recognised business, entry into a new class or regulated context sometimes justifies this compromise. The strength of the bond must be tested: too discreet and it adds no value; too dominant and it removes the desired autonomy.
3.4. House of brands
Every brand looks independent. The group remains invisible or institutional. This model allows separate positioning strategies, prices, channels and risk profiles.
However, it requires resources. Each entity needs a strategy, assets, content, distribution, protection and monitoring. A house of brands without dedicated budgets often becomes a house of names.
4. The decision tree in twelve questions
The decision process begins with a series of gates. No single answer determines the decision; their combination creates the decision record.
- Is the buying situation the same? If the need arises at the same time, a common memory has more value.
- Does the central promise remain credible? An extension that contradicts the heart weakens both offers.
- Are experience standards comparable? The bond commits the parent brand to the perceived quality.
- Does the public already know the entity? Removing an acquired brand can destroy a built availability.
- Do the buyer audiences overlap? A shared decision-maker benefits from a clear portfolio; unrelated audiences can ask for more autonomy.
- Is cross-selling real? A business hypothesis alone does not justify a visible connection.
- Is the risk of reputation correlated? A controversial, experimental or highly regulated activity may require separation.
- Does the parent brand bring anything? Trust, distribution, recruitment or evidence must be explained.
- Can the entity finance its difference? Without a media budget, product and content, autonomy will remain nominal.
- Are rights available? Strategic reasoning precedes legal validation, but it never ignores it.
- Will the name survive evolution? A name linked to a technology or segment can quickly lock up.
- Will the relationship be understood in five seconds? If a long oral explanation is required, the architecture lacks signage.
A team can score each question from 0 to 3, attach the evidence and report uncertainty. The total is not a machine to decide; it reveals the assumptions that require research.
5. Map the portfolio before redrawing it
Visible entities include corporate brand, products, services, programmes, labels, events, communities, subsidiaries, domains and applications. Internal names that never reach the public remain out of business architecture.
For each, the card indicates promise, public, purchasing situation, income, margin, geography, reputation, brand trafficking, rights, operating cost, dependencies and incidents. Really recognised assets are added. This photograph prevents deleting a value that does not appear in the organisational chart.
The work often reveals three anomalies: duplicate offers, names without own proposal and invisible relationships. It also identifies "phantom brands", always present in search results, contracts or client language after their official abandonment.
6. Measuring trust transfer
A parent brand is useful only if it alters the perception of the entity. The test presents an offer with and without bail, in a realistic context, then measure understanding, trust, consideration and attribution.
The protocol avoids the question "Do you like this architecture?". Participants perform a task: identifying the supplier, explaining relationships, choosing an offer or finding the right support. Timeliness and errors are as informative as declared preference.
For an important migration, we compare clients, prospects and collaborators. Customers sometimes know the old system to the point of compensating its defects; the new audiences reveal the real clarity.
7. Choose the right level of endorsement
The relationship is not binary. It has several levers: name order, size, proximity, formulation, colour, domain, navigation, legal signature and narrative.
An endometrial matrix can define four intensities.
| Intensity | Standard expression | Use |
|---|---|---|
| Institutional | group mentioned in the footer | strong autonomy, separate risk |
| Security | Brand name shown | useful trust, stand-alone proposal |
| Sub-brand | Make Name | highly shared memory and experience |
| Monolithic | Make + descriptor | offer integrated with central promise |
The rules are applied by family, not renegotiated channel by channel. Otherwise, each team rebuilds the hierarchy according to its interests.
8. Verbal architecture and signage
A legible portfolio has naming conventions. The descriptors describe a category. The proper names identify the entities that merit standalone brand memory. The versions, levels and modules follow a predictable syntax.
Digital signage then translates the relationship: domain, Breadcrumb feed, navigation, titles, metadata, structured schematics, social accounts and email signatures. An architectural drawing that does not appear across the customer journey does not help anyone.
Search engines require the same clarity. A migration provides canonicals, redirections, business listings, structured data, mentions and entity pages. Merging two sites without inventory can lose search queries, links and historical evidence.
9. The particular case of acquisitions
The decision is not made on the day of the announcement. It depends on the value of the name acquired, the timing of integration, contracts, markets and the truly unified experience.
Four trajectories are common: retention, phased endorsement, transition to a sub-brand, absorption. Each has transition criteria. For example, absorption can wait until support, product, billing and service commitments meet the standards of the brand name.
The transitional period must end. An eighteen-month lock without measure creates two identities, two sites and contradictory messages. Conversely, an immediate shift can sacrifice local trust and make former clients orphaned.
A sentence of deliberately long sums up the necessary discipline: before removing an acquired name, the company must be able to show, through research data, interviews, routes and commercial metrics, what this name still brings, what the target brand will add, how customers will be oriented, what assets will be preserved, what risks of confusion have been tested and when the decision will be reassessed if the results contradict the initial hypothesis.
10. Build the business case
The calculation compares a status-quo scenario and one transformation scenario over three to five years. It includes research and creation, repositories, sites, applications, content, signage, packaging, contracts, training, SEO migration, transition campaigns and support.
The expected gains are formulated separately: reduced duplicate costs, increased consideration, accelerated cross-selling, clarification of recruitment or reduction of confusion. Each receives a metric and a hypothesis. Savings on brand guidelines are not an increase in income.
The opportunity cost counts. During migration, teams produce less growth or product. If this mobilisation is not funded, the project stretches and coexistence costs more.
11. Migrate without losing brand memory
The transition follows controlled waves.
Wave 1 — evidence and foundations. Legal validation, asset inventory, baseline, nomenclature, governance and technical plan.
Wave 2 — critical touchpoints. Product, sale, support, contracts, billing and security. The promise must be true before the campaign.
Wave 3 — identity and discovery. Sites, domains, content, accounts, press relations, partners and search engines. Previous search queries have an explicit destination.
Wave 4 — withdrawal. Obsolete assets are archived, redirections monitored and residual uses processed. The legal evidence of the old name is retained.
For several months, a table follows old/new brand traffic, support contacts, attribution errors, conversion, cross-selling, customer sentiment, internal usage and asset coverage. Success is not "100% of the files have the new logo".
12. Governance: Who can create a name?
A simple rule protects the portfolio: any request for a proper name must prove a separate purchase situation, proposal and maintenance budget. Otherwise, the entity receives a descriptor in the existing system.
The Architecture Committee brings together strategy, brand, product, trade, legal, digital and, as appropriate, country. It does not validate every campaign title. It arbitrates new entities, exceptions, acquisitions and withdrawals.
The central registry retains ownership, status, promise, relationship, rights, domains, assets and review date. Exceptions expire. A living architecture governs itself like a portfolio, not like a frozen set of brand guidelines.
13. Test three scenarios rather than one drawing
The committee compares as a minimum the status quo, partial consolidation and ambitious target architecture. Each scenario uses the same criteria: understanding, trust transfer, risk, rights, cost, delay and governance capacity.
A prototype is not limited to a diagram. It materialises home page, navigation, offer, commercial proposal, invoice, application, recruitment and search assistance. Relationships then become visible in real tasks.
The maintenance scenario serves as a reference. It includes current hidden costs: duplicate content, competing campaigns, domains, licences, adaptations, commercial confusion and maintenance. Without this basis, simplification seems costly only because the existing one is treated as free.
The intermediate scenario can preserve recognised names while bringing together technology, content and graphic system. It sometimes generates more value than absorption, especially when audiences and reputations remain distinct.
The target scenario is the transition period. For twelve or twenty-four months, two names may have to coexist, which temporarily increases costs before reducing them. Financial models present this curve instead of announcing an immediate economy.
When three options arrive at the committee, each must indicate the assets retained, the clients exposed, the contracts to be modified, the rights to be obtained, the routes to be reconstructed, the transition expenses, the possible loss of demand and the signal that would prefer another route; this symmetry prevents the team from accurately describing the favourite track while caricaturing the alternatives.
14. International architecture: centralise the rules and localise the evidence
The brand can remain global while descriptors, offers and sureties vary by market. A literal translation is not enough: category, regulation, distribution and local reputation modify the reading.
The core defines authorised relationships, assets, exception criteria and decisions that require review. Countries provide linguistic research, competition, legal constraints and purchasing behaviour.
A local exception has a motive, owner and deadline. If several countries request the same adaptation, the central model may be incomplete. If the derogation no longer serves, it shall be withdrawn.
Before a multi-country deployment, the company tests not only the pronunciation of names, but also the understanding of the portfolio, the transfer of trust and digital availability; a strong bond on the original market may not bring any value in a country where the parent brand is unknown, while a locally acquired entity may bear trust that forced standardisation would destroy.
15. What assets cannot prove — and what architecture can never solve alone
A brand relationship does not correct a weak product or poor distribution. Assets give a memory shortcut; experience provides evidence that makes it credible. When a merger of names does not produce the expected cross-selling, the team reviews the offer, incentives, data, sales journey and service before further modifying the signs.
16. Metrics a year later
The follow-up table compares the reference and the exposed cohorts. It observes brand search, direct traffic, attribution, understanding relationships, consideration, cross-selling, support and cost of assets.
The results are read by segment. A temporary decrease in the old name can be expected; it is acceptable only if the new signal progresses and customers always find their way.
The decision is reviewed. An architecture is not good because it corresponds to the approved scheme. It is if it makes the portfolio easier to understand, choose and sustain over time.
17. The most costly mistakes
Reproduce the flowchart. The customer receives the complexity that the company has not resolved.
Consolidate offer and brand. A feature or tariff formula does not necessarily require a proper name.
Underestimating the acquired brand. Revenue does not measure local confidence and demand traffic alone.
Declare synergy without path. Two logos side by side do not create cross-selling or common experience.
Launch before the law. An advanced creative track becomes emotionally difficult to give up.
Fail to plan retirement. The old domains, documents and designations continue to produce confusion and risk.
18. Frequently asked questions
18.1. Should the portfolio always be simplified?
Remove complexity that serves no purpose. Several brands remain relevant when their propositions, audiences, risks and resources are genuinely distinct.
18.2. Can the same company combine several models?
Yes. The hybrid is common. However, it must follow family-readable rules, otherwise it becomes a collection of historical exceptions.
18.3. When does an offer deserve a standalone brand?
When it has a sufficiently distinct proposal, public or risk, that a separation helps purchase and that a sustainable budget can build its memory.
18.4. How long does a migration take?
From a few months to several years according to contracts, products, country and value of the old name. The calendar is based on milestones of experience and understanding, not just on a campaign date.
18.5. Is brand architecture an SEO concern?
It does not reduce, but migration affects domains, pages, entities, links, search queries and structured data. The SEO plan therefore belongs in the programme from the outset.
19. Logiks recommendations
Start with the public brief and purchase situations, then document the costs and risks of each entity. Give a proper name only when it actually helps to choose and the organisation can maintain it. Successful architecture quickly understands itself, transfers trust to the right place and concentrates investments without masking useful differences.
20. Main sources
- WIPO, World Intellectual Property Indicators 2025 — Trademarks : https://www.wipo.int/web-publications/world-intellectual-property-indicators-2025/en/trademarks-highlights.html
- WIPO, Madrid Yearly Review 2026 : https://www.wipo.int/publications/en/details.jsp?id=4822
- Bejoy John Thomas et al., longitudinal study of 1,162 distinctive assets, 2026: https://www.tandfonline.com/doi/full/10.1080/02650487.2026.2637295
- Ehrenberg-Bass Institute, Brands of Distinction : https://marketingscience.info/news-and-insights/brands-of-distinction
- Kantar, What are distinctive assets and why are they important? : https://www.kantar.com/north-america/Inspiration/Advertising-Media/What-are-distinctive-assets-and-why-are-they-important