The Post-Mortem Library: Four DTC Collapses, Three Repeating Mistakes
We keep a library of other people's funerals. Not because failure is entertaining, but because in this industry the tuition for these lessons was paid by someone else — billions of dollars of someone else — and the notes are lying around for free. When we underwrite an acquisition, the diligence checklist is mostly a list of ways companies on this page died.
- Channel concentration: one rented acquisition engine (an algorithm, a marketplace, a flash-sale list) treated as if it were owned.
- Growth priced ahead of economics: capital raised against trajectory, forcing growth-at-any-cost long after the unit math said stop.
- Fixed costs scaled to the best quarter: headcount, warehouses, and leases sized to peak demand, turning a slowdown into a death spiral.
The case files
| Company | Peak | What Broke | The Mistake |
|---|---|---|---|
| Thrasio | ~$10B valuation, 200+ brands | Chapter 11, 2024 | 4:1 debt-to-equity buying spree; volume over diligence; Amazon dependence |
| Casper | $1.1B private valuation | IPO'd at less than half; sold off cheap | CAC outran LTV in a commodity category; growth marketing as the product |
| SmileDirectClub | $8.9B IPO (2019) | Liquidated, 2023 | Scaled a clinically contested model on paid acquisition; never reached durable unit economics |
| Zulily | $9B valuation (2014) | Shut down, 2023 | Flash-sale model with structurally slow shipping in a Prime world; repeat behavior decayed |
Mistake one: rented land, treated as owned
Thrasio had Amazon rankings. Zulily had a daily-deal email list whose habit decayed the moment shipping expectations changed. In each case the company's most valuable "asset" lived inside someone else's system, subject to repricing without notice. The test we apply: if the platform halved your reach tomorrow, what's left? If the answer is "the product and a spreadsheet," the brand is a channel position, not a business.
Mistake two: capital that demands the wrong behavior
Casper's S-1 made the math public: hundreds of dollars of CAC against a product most people buy once a decade, with double-digit return rates. Everyone could see it; the structure didn't care. Venture pricing demanded growth, so growth was purchased at negative margin, quarter after quarter. SmileDirectClub ran the same loop with higher stakes. The lesson isn't "capital is bad" — it's that the return profile of your capital becomes your operating strategy whether you like it or not. A brand bought with patient money can decline a bad growth dollar. A brand priced on trajectory cannot.
Mistake three: fixed costs sized to the best quarter
Thrasio staffed teams of ten for the work of two. Zulily built fulfillment for a model whose volume assumption broke. Operating leverage works in both directions: every dollar of fixed cost added at the peak is a dollar of loss locked in for the trough.
What survives
Run the same lens backward over the brands that endured and the pattern inverts: multiple channels with owned relationships at the core, capital structures that tolerate flat quarters, and variable cost structures that breathe with demand. None of it is glamorous. All of it is buyable — which is the entire point of our model. We don't need to invent durable brands; we need to recognize them, price them honestly, and not import the three mistakes after closing.
- Diligence weights channel concentration as heavily as profitability — a profitable brand with one channel is priced like a fragile one.
- We use modest, sensible structures. No deal math that only works if next year is the best year.
- Post-close, fixed costs stay variable as long as possible — shared infrastructure across the portfolio exists precisely so no single brand carries peak-sized overhead.
Sources: Thrasio Chapter 11 filings and coverage (PYMNTS, 2024); Casper S-1 and subsequent reporting; SmileDirectClub liquidation coverage (2023); Zulily shutdown reporting (2023); Starter Story interview with John James. Valuations as publicly reported at peak.
Own a brand this applies to?
We talk to founders at every stage — long before they're ready to sell.
Start a Conversation →