Nobody asks for a sub-brand because the audience is confused. They ask because the org is. A new vertical needs to feel like a real launch, a team wants something of its own to run, a facility cost enough money to deserve its own identity. All of those are internal problems, and a sub-brand pays for them with the one thing the audience actually had: a single clear idea of who you are.
That does not mean never build one. It means the bar is higher than it feels in the meeting where it gets proposed.
What a new mark actually costs
The useful way to think about this comes out of the Ehrenberg-Bass Institute’s work on distinctive assets. Jenni Romaniuk’s framework scores any brand element on two axes: fame, meaning how many people connect it to you, and uniqueness, meaning how exclusively it points at you rather than a competitor. Strong assets score high on both. That is what makes a color, a shape, or a wordmark do work without being explained.
A brand new mark starts at zero on both. Every impression it earns is an impression that could have reinforced the parent instead. You are not launching a brand, you are taking on a debt and hoping the budget outlasts it. The organizations that get away with a house of brands are the ones large enough to fund each one properly, forever. Most teams proposing a sub-brand are not that, and the honest version of the proposal is that the parent brand will get quieter so the new one can exist.
Three questions
So before designing anything, three questions, in this order.
Does the audience already sort these separately? Not the org chart, the audience. If the same person moves between the two offerings without thinking about it, they are one brand with two rooms. If the two attract genuinely different people who would not consider the other, you have a real case.
Does it need its own distinctive assets, or just its own name? These are very different asks. A named tier inside a parent system costs almost nothing and stays legible. A separate color, type, logo, and voice is a second brand to fund. Most cases that feel like the second one are satisfied by the first.
Can the parent’s assets carry it? If the parent is well known in its category, borrowing that recognition is the cheapest distribution you will ever get. Spending it on a fresh identity to look serious is usually the expensive way to look smaller.
What this looked like in practice
At SPORTIME I owned marketing across 13+ clubs and 25+ social accounts, and the architecture question was live constantly. Tennis was the established core. I helped launch the pickleball vertical, which by every internal signal wanted to be its own brand: different energy, different players, a real growth story, and a lot of enthusiasm for making it distinct.
It stayed inside the parent system. The two sports share facilities, staff, and in a lot of cases the same member walking from one court to the other. Sorting them into separate brands would have described an internal distinction the audience did not experience. What they got instead was their own content behavior inside one identity, which is the first question resolved as a named tier rather than a second brand.
Where a distinct identity was warranted, the reason was structural rather than tonal. The John McEnroe Tennis Academy is a named institution with its own reputation, and I directed creative around its $75M flagship expansion. The Johnny Mac Tennis Project is a nonprofit, which is a genuinely different promise to a genuinely different audience, donors rather than members. Those pass the first question honestly. Pickleball did not, however much it wanted to.
The tell, across all of it, was that the pressure to split almost always came from inside. It came from a team wanting ownership, or a launch wanting a moment. Those are real needs, and they are almost always better served by giving people a distinct lane inside one system than by handing them a brand to maintain.
The part that keeps it from becoming chaos
If you do split, the discipline is deciding up front what is shared and what is not, and writing it down. Which assets are the parent’s and never get modified. Which are the sub-brand’s alone. Who approves a new one. Twenty five accounts do not go feral because someone had bad taste. They go feral because nobody ever wrote down which decisions were already made, so every account made them again locally.
That is the whole thing, really. Brand architecture is not a diagram of boxes. It is a decision about what you are willing to fund and for how long, plus a written record of what has already been settled. Get those two right and a sub-brand is a tool. Skip them and it is 25 different answers to the same question.
SPORTIME is the proof that the restraint is the harder call, and usually the right one.
Sources
- Jenni Romaniuk, Building Distinctive Brand Assets (Oxford University Press, 2018): global.oup.com
- Ehrenberg-Bass Institute on distinctive asset measurement: marketingscience.info