Most ecommerce brands measure marketplaces by a single number: gross sales. The dashboard looks healthy, orders are flowing, and growth feels real. But that number hides a structural problem that builds quietly over years — on a marketplace, you are not building a customer base. You are renting access to one.
The distinction sounds academic until you try to grow profitably. A brand that sells a million dollars through a marketplace and a brand that sells the same through owned channels are not the same business. One controls its margins, its data, and its ability to bring a customer back. The other controls almost none of it — and pays a recurring tax for the privilege.
This is not an argument to abandon marketplaces. They are powerful acquisition engines, and for many brands they are the right place to be discovered. The argument is narrower and more important: discovery and ownership are two different jobs, and most brands let the marketplace do both — then wonder why their margins never improve.
The fee you see is not the cost you pay
Every operator knows the headline marketplace commission. What rarely gets modeled is the stacked cost: commission, plus mandatory advertising to stay visible, plus forced discounting to win the buy box or stay competitive on price, plus fulfillment fees, plus the cost of returns the platform’s policies encourage.
Individually each line looks tolerable. Stacked together, they add up to far more than the commission alone. On Amazon, the all-in cut — referral fee, fulfillment, and advertising combined — has reached roughly half of seller revenue, up from about 40% five years earlier. In Southeast Asia the headline numbers start lower — Shopee Indonesia’s mandatory seller fees on fashion run roughly 4.5–11%, and TikTok Shop Indonesia’s base commission sits around 6–12% by category — but the same stacking logic applies. Once advertising, mandatory promotions, logistics, and returns pile on top, effective seller take rates across the region climb as high as 20–25% of post-discount sales.
And because the marketplace controls visibility, these costs are not optional — they are the price of continuing to sell, not a one-time acquisition cost. You don’t pay once to acquire a customer; you pay every time that customer buys, forever.
You don’t own the customer, so you can’t amortize acquisition
The entire economics of healthy ecommerce depend on one mechanic: spend to acquire a customer once, then earn repeat purchases at near-zero acquisition cost. That second, third, and fourth order is where margin lives — which is why acquiring a new customer is estimated to cost five to twenty-five times more than retaining an existing one.
On a marketplace, that mechanic breaks. You typically don’t receive the customer’s contact details, you can’t message them directly, and the platform actively discourages — or prohibits — taking the relationship off-platform. So you re-acquire the same customer on every purchase, paying the full stacked cost each time. A high repeat-purchase rate, which would transform the profitability of an owned-channel brand, does almost nothing for your marketplace P&L because each repeat order carries the same fees as the first.
You are not building customer lifetime value. You are building the marketplace’s.
Without data, you can’t do retention at all
Retention is not a feeling; it’s an operation. It requires knowing who bought, what they bought, when they’re likely to buy again, and a channel to reach them. Marketplaces sever every link in that chain. You get an order, not a customer. You get a transaction, not a relationship.
This is why so many brands that are “doing well” on marketplaces feel strangely fragile. A pricing-algorithm change, a new competitor, a category fee increase, or a suspended listing can erase a revenue stream overnight — and the brand has no way to reach the customers it spent years serving. Dependency isn’t just a margin problem. It’s a continuity risk.
The shift isn’t “leave marketplaces,” it’s “own the second order”
The brands solving this aren’t fleeing marketplaces. They’re treating the marketplace as what it’s actually good at — first contact — and then deliberately moving the relationship onto a channel they control for everything after.
In practice that means using the first interaction to open a direct line: an owned store, a CRM record, and increasingly a conversational channel like WhatsApp Commerce where a real, two-way relationship can exist. The same logic drives brands shifting paid acquisition from rented forms to owned WhatsApp threads, and brands replacing link-in-bio tools with owned destinations. The first order may still carry full marketplace cost. But the second, third, and fourth happen on owned infrastructure — at a fraction of the cost and with full visibility into the customer.
What this looks like in practice
The figures below are illustrative scenarios, not case studies — but the structure is one any operator will recognize.
The “profitable” brand that isn’t. Picture a homeware brand doing strong revenue, the large majority of it through marketplaces. Leadership celebrates the growth. But the blended take-rate after commission, ads, and discounting eats nearly half of every sale, repeat purchases generate no margin advantage, and the brand holds zero customer contact data. When the marketplace raises category fees a few points, the brand has no lever to pull and no audience to fall back on. Revenue is real; resilience is zero.
The brand that reclaimed the second order. Now picture a skincare brand that keeps marketplaces for discovery but routes post-purchase communication — order updates, replenishment reminders, restock alerts — to WhatsApp Commerce. Within a couple of quarters a meaningful slice of repeat revenue flows through the owned channel at a fraction of marketplace cost, and the brand can now segment, reactivate, and launch to its own audience.
A global pattern, not a local one. This dynamic repeats across very different ecosystems — Southeast Asian brands navigating dominant regional marketplaces, European DTC brands balancing Amazon against owned stores, and Latin American sellers building messaging-first relationships. The platforms differ; the structural trap is identical. Indonesia’s marketplace-heavy market is simply one vivid example of a worldwide pattern: wherever a marketplace owns discovery, it tends to own the customer too — unless the brand deliberately takes that relationship back.
What to actually do about it
- Model your true stacked take-rate. Add commission + advertising + discounting + fulfillment + returns as a single percentage of marketplace revenue. Decisions get clearer once you see the real number, not the headline fee.
- Separate acquisition from retention in your strategy. Let marketplaces do what they’re good at — being found. Don’t ask them to also be your retention engine; they’re structurally incapable of it.
- Capture a direct line on the first interaction. Build a deliberate mechanism — packaging inserts, post-purchase flows, opt-ins — to move willing customers onto a channel you own.
- Make the second order happen on owned infrastructure. Move replenishment, restock alerts, and reactivation to a channel where repeat purchases don’t re-incur acquisition cost.
- Treat customer data as the asset it is. A reachable, segmented audience is the single biggest determinant of long-term ecommerce profitability and resilience. Build it deliberately, not accidentally.
The reframing
Marketplace dependency is rarely a single bad decision. It’s the slow accumulation of a thousand reasonable ones — each order taken on someone else’s platform, each relationship left uncaptured, each repeat purchase paying full freight. The result is a business that looks like it’s growing while its margins and its resilience quietly erode.
The fix is not dramatic. It’s a reframing: marketplaces are for being discovered, not for being owned. The brands that internalize this don’t walk away from volume — they simply stop renting the relationships that volume creates, and start owning them.
Many brands are now rethinking where the customer relationship actually lives — shifting retention and repeat-purchase communication toward conversational commerce channels like WhatsApp Commerce, so that the second order strengthens their own business instead of their marketplace’s. If your repeat customers can’t be reached directly, that’s usually the first place worth looking.