Put the two poles side by side and the trade-offs become visible. Branded house (one masterbrand across categories): every rupee of marketing compounds into one memory, new launches start with borrowed trust, and one scandal anywhere is a scandal everywhere. The efficiency and the contagion are the same wire. House of brands (standalone names per category or segment): each brand can be perfectly fitted to its segment and price tier, failures die quietly without infecting siblings, competitors face a wall in every niche, and every brand pays full price for its own awareness, forever. Between them, sub-brands (parent visible, offspring distinct) and endorsed brands (independent name, parent's signature in the corner, 'by Tata') tune how much trust transfers and how much risk flows back.
The stretch decision is where architecture stops being taxonomy and starts being money. Three questions, asked in order. Transfer: what does the parent's memory actually say about the new category? 'Titan' carried precision and trust toward watches, but jewellery purchases run on purity anxiety and tradition, associations a watchmaker's name didn't hold, hence Tanishq. Contagion: if the new venture stumbles (a quality scare, a positioning embarrassment), does the damage route back up the name to businesses that were healthy? Distance protects; proximity infects. Economics: a new name costs years of building from zero, while the family name is free on day one and mortgaged forever after. Most stretch mistakes are transfer optimism, assuming the memory says more than it does, audited afterwards by disappointing launches that 'had everything going for them'.
Portfolio strategy assigns each brand a job beyond its own P&L. The cash cow funds the system and is managed for margin, not applause. The flanker exists to fight: a value brand absorbing price warfare so the premium brand never discounts (the detergent ladder's bottom rung is armour, not embarrassment). The silver bullet lends associations upward (a halo product whose job is what it says about the family). The strategic brand is where the future revenue lives and gets funded beyond its current size. Naming these roles out loud is what prevents the classic portfolio pathology: every brand manager fighting for equal investment as if every brand had the same job.
And architecture needs periodic demolition. Portfolios accrete (acquisitions, line extensions, someone's 2014 innovation initiative) until marketing money is spread across brands the strategy forgot. The audit questions are blunt: which brands would we launch today if starting fresh? which have distinct jobs and which merely have histories? what would migrating this equity into that one cost, versus feeding both forever? Consolidation done well (equity migrated, codes retained, customers walked across) funds the survivors. Done lazily (a memo and a logo swap), it donates the orphaned equity to competitors.