House of brands: when a second brand is honest
Founders reach a point where the obvious next move is a second brand. The reasons usually given do not justify it. Three of them do, and the cost is larger than the pitch deck says.
- If the proposal can be written as another rung or another occasion inside the existing range, it is range work, not a second brand. Only three cases genuinely need a separate name: a price tier the brand cannot carry, a buyer blocked by positioning rather than by range, and a channel the first brand cannot enter without damaging its own price.
- Trade mark protection is per mark, not per company. Section 18(2) of the Trade Marks Act, 1999 allows one application across several classes but the fee is payable for each class, and the series provision in Section 15(3) covers variants of one mark, not a new name. Amazon Brand Registry enrols per brand against that brand's own registered or pending trade mark, and Flipkart requires a brand approval per brand.
- The roll-up record is checkable. Thrasio, at roughly 180 brands, filed for Chapter 11 on 28 February 2024 seeking to cut about 495 million dollars of debt. Benitago filed on 30 August 2023 with 94.6 million dollars of debt. In India, 10Club began insolvency proceedings at the Bengaluru bench of the National Company Law Tribunal, reported by Entrackr in May 2025. BRND.ME, formerly Mensa Brands, told Inc42 it bought about 20 brands and now concentrates on four.
- There is no reliable published benchmark for the overhead of running a second brand in India, so build the number bottom up: a second trade mark, second artwork and labels, second brand approvals on every platform, reviews and search history at zero, acquisition at full price, and founder attention taken from the brand that already works.
A second line is range work. A second brand is a new company
The brand works, the founder has capacity, and somebody says the next move is a second brand. Usually it is not. Usually it is a second line, which is a different and much cheaper decision.
Range architecture already handles most of what founders bring here: another rung on a price ladder, another need state, another pack size, a variant a retailer asked for. That post sits one level below this one. If your proposal can be written as a position inside the existing range, it is range work. Stop there.
A second brand is a different object: a new trade mark, new artwork, a new catalogue, a review base at zero, a brand approval on every platform, and a fresh claim on the input you cannot buy, which is attention. Treat it as founding a company that shares your warehouse. The mechanics of then running two are a separate discipline.
The reasons founders give, and which ones hold
Four come up constantly. None is sufficient.
The first brand has plateaued. A second brand does not fix a demand problem, it divides the team that was going to fix it. If growth has stalled, the work is in the range and the acquisition cost.
We already have the team and the warehouse. That is a statement about cost, not demand. Spare capacity makes a second brand cheaper. It never makes it sell.
An investor wants a platform story. Real for the investor, expensive for you.
The name was available. Naming energy is not a market.
Three cases where a separate brand is the only honest answer
A price tier the existing brand cannot carry. A ladder inside one brand works while the shopper can see what the extra money buys. Past that point a value line under a premium name teaches your premium buyer to wait, and a premium line under a value name is not believed. When your pricing architecture cannot hold the gap, the gap needs its own name.
A different buyer, where positioning is the obstacle. If the person you want is rejecting what your brand means rather than what it sells, no amount of range fixes it, and a position cannot be widened indefinitely without becoming nothing.
A channel the first brand cannot enter without damage. A value platform, a modern trade entry price, an institutional pack. Channel conflict is governable inside one brand until the price that channel requires becomes visible to your best customer. Then it is a second brand or nothing.
What two of everything costs in India
Trade mark protection is per mark, not per company. Section 18(2) of the Trade Marks Act, 1999 lets one application cover several classes, but the fee is payable for each class, and the series provision in Section 15(3) covers variants of one mark, not a new name. So brand two means a second application, examination and opposition window. Settle the naming question first.
Then the platforms. Amazon Brand Registry enrols per brand against that brand’s own registered or pending trade mark. Flipkart makes you clear a brand approval before you can list under a name, with a trade mark or an authorisation letter naming your seller ID. Both are per brand, on every platform, permanently, as are artwork, identifiers, label compliance and the document pack.
Then the parts nobody budgets. Reviews restart at zero. Search and conversion history restart at zero. Acquisition restarts at zero, so brand two pays full price for customers brand one now gets cheaply, and the per SKU arithmetic does not transfer. There is no reliable published benchmark for what a second brand’s overhead costs an Indian D2C business, so accept no percentage from anyone. Build it bottom up, and put an honest number on founder hours.
The roll-up version, and what the record shows
The aggressive form of a portfolio is the roll-up: buy many small brands, run them on one back office, sell the group. Thrasio, which had assembled roughly 180 brands at a reported peak valuation near 10 billion dollars in 2021, filed for Chapter 11 on 28 February 2024, seeking to cut about 495 million dollars of debt. Benitago filed on 30 August 2023 with 94.6 million dollars of debt. In India, 10Club began a corporate insolvency resolution process at the Bengaluru bench of the National Company Law Tribunal, reported by Entrackr in May 2025, on the ground that its assets were insufficient to pay its debts.
The version that survives is smaller. BRND.ME, formerly Mensa Brands, told Inc42 in February 2026 that it bought about 20 brands, sold the ones that were not core, and now concentrates on four. Inc42 reported FY25 revenue of 168 million dollars and a loss of 60.4 million dollars, with FY26 guidance of 1,600 crore to 1,700 crore rupees and adjusted EBITDA breakeven expected. Read that as the model correcting itself, not a verdict on one company.
What changed is what gets bought. On 29 September 2026 Prashant Parameswaran, formerly managing director of Soulfull, launched Arovia with a 100 crore rupee commitment from Fireside Ventures to take largely majority stakes in regional packaged food businesses, initially two to four. The stated filter is businesses past 100 crore rupees of revenue that are already cash flow positive or profitable. That is a portfolio of working companies, not of listings. The operator version of the same arithmetic is Snitch buying Berrylush, where the case rests on fixed cost spread across stores the buyer already runs, and Honasa putting a chief executive over an acquired brand rather than a brand manager.
The test
Three questions, in writing, before anybody registers a name.
Name the buyer the first brand cannot serve, and say why positioning rather than range is the obstacle. If you cannot separate those two, you are diluting.
Name what the first brand loses. Attention is the honest answer and the best predictor of failure. A hero product that drifts on availability or content while a new brand launches has quietly been defunded.
Then the stand alone question. If this brand belonged to somebody else, and you funded it at full cost with no shared name and no borrowed goodwill, would you write the cheque. A portfolio is a set of businesses that each pass that test. A dilution is one business wearing two names.
A house of brands is also a distinct acquirer, closer to a strategic than a financial buyer, and it prices you on whether you slot into a back office it already owns.