The models you will be shown are older and shakier than they look. What actually decides it, and the three costs of opening a second brand.
Start from the cost, not the model. Measured against the sources behind them, the 4 canonical models are quoted more confidently than their evidence supports: the taxonomy everyone cites comes from an article rather than the book it is attributed to, and “hybrid” is not one of the four at all.
What you are actually choosing between
Strip the vocabulary away and there are two positions with a slope between them.
Everything can carry the parent name, so each new product inherits the recognition you already have. Or products can stand on their own names, so a failure or a category shift does not touch the parent. Most real ranges sit somewhere along that slope, and the decision is rarely permanent.
What the models will not tell you
The literature here is quoted with more confidence than it earns, and a founder should know that before paying for a strategy built on it.
The taxonomy that gets drawn on every whiteboard comes from a journal article, not from the book it is usually credited to. “Hybrid” is a later addition and not one of the original four. The single most-repeated sentence in this field is an opinion offered without data. And the most-cited finding linking architecture to company value is routinely repeated backwards from what the research said.
None of that makes the models useless. It means they are a vocabulary for discussing your decision, not evidence for it.
The three costs of opening a second brand
This is where the decision actually gets made, because each cost lands on a different part of your business.
| Cost | Who pays it | When it shows up |
|---|---|---|
| Recognition restarts | Marketing | Immediately, and for years |
| Identification and listings multiply | Operations | At the first retailer conversation |
| Every system doubles | You | At the second range extension |
The middle row is the one founders underestimate. A separate brand means a separate identity, separate artwork, and its own identification on every pack: ask GS1 what a new variant or a changed pack requires of your product numbering before you commit, since the answer arrives with the retailer and not with the design.
Range architecture, the same decision on the shelf
Range architecture is the same decision expressed on the shelf, and it is more concrete.
Decide what stays identical across every pack and what is allowed to vary, then decide what the varying element is: colour, a name, an icon, or a position. Colour is the usual answer and it runs out faster than founders expect: the range grows and the palette does not. A structural answer, where the hierarchy or a fixed field carries the difference, survives a bigger range than a colour code does.
What the shelf costs, in numbers you can check
Two of the costs above have figures you can look up before you decide, and they are worth pulling into the room.
A second brand on the same shelf needs its own identification, and identification has its own size floor: a retail barcode runs from 80 to 150 percent magnification, and at 100 percent it occupies 37,29 mm of width. On a small pack, two brands means two panels that each have to carry that block plus the mandatory information at its own legal floor, which in the UK is a 1,2 mm x-height. Those two numbers decide more small-pack architectures than any model does, and MarkaWorks checks them against the smallest pack in the range before the architecture is agreed.
A test you can run this week
Write down the products you will sell in twenty four months, not twelve. Then ask three questions of that list.
- Would a buyer expect these to come from the same company?
- Does any product carry a risk you would not want attached to the parent name?
- Does your distribution treat them as one relationship or several?
If all three point the same way, you have your answer without a model. If they split, the split tells you exactly where the boundary in your architecture belongs.
Where this lands in practice
A masterbrand consolidation that looks decisive in a case study is usually slower in reality; one documented consolidation came three years after the acquisition that prompted it.
So treat the decision as reversible and expensive rather than permanent and free. MarkaWorks writes the architecture decision into the brief with the range list beside it, and revisits it at the second extension rather than at the fifth, because the list is what makes the choice concrete, and a choice made against a list can be defended a year later when somebody asks why.
The architecture line in a MarkaWorks brief
MarkaWorks writes the architecture decision into your brief with the range list beside it, and revisits it at the second extension rather than at the fifth. Before the architecture is agreed, we check the barcode floor and the mandatory panel against the smallest pack in your range. Range architecture is part of the brand strategy and visual identity service.
