Written by: Michał Sobieraj, Operations Manager, Digital Colliers
Shein lost roughly $5B in market value in the days after its Hong Kong IPO, dropping to around $21B. One of the worst opening weeks for a major listing. The stumble isn't just about Shein. It's a signal that marketplace platforms can wobble fast, and brands selling through them rarely have a margin model that shows what happens when one channel halts.
Why brands skip the stress test
Growth hides structural risk. When a marketplace channel is growing, you don't model the downside. You scale into it. Customer acquisition cost across DTC brands has risen roughly 40% since 2023, so marketplaces feel like a cheaper channel. You get distribution without paying Meta's CPMs, which have climbed year over year. The platform takes a commission, but you don't fund the top of funnel.
The problem is most operators don't know their real unit economics split by channel. You know your blended margin. You might know margin by SKU. But you probably don't have a model that shows revenue, contribution margin, and P&L by channel with and without each platform. When a marketplace stumbles, you're guessing how bad it gets.
What the margin scenario model looks like
The model is simpler than it sounds. You need three layers. First, revenue by channel. Marketplace A, marketplace B, your own site, retail if you have it. Second, contribution margin after all costs. That means platform fees, ad spend if you're buying placement, returns, payment processing, fulfilment. Third, the stress test. What happens if one channel drops 50%? What if it goes to zero?
Most brands stop at layer one. They know how much revenue comes from each channel. But they don't know if that revenue is profitable once you account for returns and the cost to acquire the customer. Online return rates run roughly 19 to 20% of gross sales. For apparel, it's higher. UK apparel return rates run 25 to 40% depending on category. If your marketplace channel skews toward high-return categories, the gross margin on the dashboard is fiction.
The unit economics most brands miss
Returns are obvious, but there are two more. First, roughly 30% of SKUs at a typical multi-channel eCommerce brand lose money per order after returns and ad spend. You might be profitable overall, but a big chunk of your catalogue is a loss leader. If your marketplace channel skews toward those SKUs, you're subsidising growth. Second, ad spend is rising everywhere. Meta CPMs are up. Google is up. Even if you're not buying ads on the marketplace itself, you're paying to drive traffic to your site, and some of that traffic leaks to the marketplace listing instead.
The pattern I keep seeing is brands assume marketplace revenue is incremental. It's not always incremental. Sometimes it's cannibalising your owned channel, where the margin is better. Sometimes it's selling SKUs that don't make money. You don't know until you model it.
Build the model before the platform stumbles
You don't need a complicated tool. A spreadsheet works. Start with revenue by channel for the last quarter. Then subtract the platform fee, the cost of returns, and any ad spend allocated to that channel. That gives you contribution margin. Now stress it. What if marketplace A drops 30%? What if it drops to zero? Can you shift that volume to another channel, or does it just disappear?
UK eCommerce grew roughly 3% in 2024 versus 2023. Single-digit growth is the new baseline. You can't assume another platform will replace lost volume. The model should show you how much of your P&L depends on each channel and what happens if one wobbles. Then you can decide if that dependency is tolerable or if you need to rebalance.
The operators who weather platform wobbles are the ones who built the model before the wobble happened. They know which channels are profitable, which SKUs are profitable, and how much runway they have if a marketplace stumbles. The ones who didn't build the model are scrambling to figure out the damage while the market cap is falling.