In the AI-native era, is House of Brands the better model?
Originally written in Chinese. Translated with AI, reviewed by Gary. Read the Chinese original →
A few days ago, in a meeting with a potential client, we got talking about brand architecture. I found it interesting and took a screenshot. Flipping through it again over the long weekend, the more I thought about it, the more interesting it got.
Honestly, it's hard to say any one model is always better. Branded House, Sub Brand, Endorsed Brand, House of Brands: each has pros and cons, and each fits a different stage of a company.
But in today's environment, especially as we enter the AI-native era, I lean towards House of Brands.
Why?
1. AI agents rewrite the purchase path
In the AI era, shoppers no longer search and compare on their own. They hand the need to an AI, which recommends brands directly. Nobody says "I want a Unilever product". They say "I want a gentle shampoo". The AI recommends Dove, not Unilever.
2. AI recommends by tags, not by brand tree
AI tags brands (audience, price, scenario, emotion) rather than understanding them through a parent-brand hierarchy.
Cornetto = romantic ice cream for young people. Hellmann's = kitchen condiments. Under a Branded House these tags blur easily, but under House of Brands they stay very clear.
3. Less guilt by association
Under a Branded House, when one category has a problem, the whole parent brand takes the hit.
Imagine an Apple Watch exploding: it could drag down trust in the iPhone and the Mac. Under House of Brands, the risk is ring-fenced. Even if one brand runs into trouble, the others keep running normally.
4. Fitting diverse personas
Consumer needs are more and more fragmented:
- Young people want trendy → Supreme
- Mums want practical → Anakku
- Couples want emotion → Victoria's Secret
House of Brands can catch each group precisely with a different brand, instead of relying on one parent brand to "drag" every audience along.
5. More flexible experiments and iteration
In the AI-native era, product life cycles keep getting shorter and iteration keeps getting faster.
P&G can quickly launch a new shampoo brand as an experiment and shut it down if it fails, without touching the long-term value of Pantene or Head & Shoulders. House of Brands lets a group work like an investment portfolio: incubate fast, cut fast.
6. A natural edge in crossing categories
A Branded House finds it hard to cross categories; users feel it doesn't fit. House of Brands is completely different. Each brand stands on its own, and consumers only recognise the brand itself, not who the parent company is.
L'Oréal is the classic example:
- Lancôme → premium skincare and beauty
- Maybelline → mass-market make-up
- Kiehl's → natural, healthy skincare
- Garnier → mass hair care and hair colour
- L'Oréal Paris → upper-mass beauty and skincare
One group, but each brand runs independently with its own positioning, covering everything from drugstore shelf to luxury, skincare to make-up, young to mature consumers, across many categories.
My early judgement
AI-native retail is no longer "one parent brand carries everything", but "one group running a set of precise, independent brands behind the scenes".
That makes House of Brands the lower-risk, better-matched, more sustainable model for growth.
Of course, this is only my current thinking. Let's test it together from here.