Marketplace or Your Own Online Store? The Wrong Question

Ask ten online sellers whether they should sell on marketplaces or build their own store, and you'll get two camps.

Camp one: "Why bother with a website? The customers are all on Amazon, Shopee and TikTok Shop." Camp two, usually someone who just received a fee increase notice: "We need our own site. Get off the marketplaces. Stop working for the platform."

Let me be blunt: both camps are wrong. And they're wrong in the same way. Both treat this as an either/or decision, when a marketplace and your own store do two completely different jobs. Asking one channel to do the other's job is what quietly eats a seller's margin.

A confession before we go on. I build websites for a living, so you have every reason to be skeptical when I say a seller needs one. And there was a time I genuinely believed every shop should have its own site, the sooner the better. That belief was wrong. A shop with no repeat customers that builds a website has built a beautiful storefront on a street nobody walks down. It took looking hard at the numbers for me to see that the real question isn't "marketplace or website". It's "where is the money going on every single order?"

I'll use two markets as evidence: the US, where Amazon has had two decades to mature, and Vietnam, where I work, which is running the same experiment at high speed. Two platforms there now control almost the entire market, and the data shows exactly what that does to sellers.

Truth one: the marketplace really is where the customers are

Fairness first. The "just sell on the marketplace" camp has data on its side.

According to Metric, combined revenue across Vietnam's four main marketplaces (Shopee, TikTok Shop, Lazada and Tiki) grew 34.75% in 2025. Shopee took 56.04% and TikTok Shop 41.31%. Two platforms hold more than 97% of the market (Thuong Gia). YouNet ECI measures differently and lands even higher, with Shopee at 57.5% and TikTok Shop at 39.6% (VOV). The two firms disagree on the total, not on the conclusion: online shoppers live on the platforms.

And shoppers stay there for a very real reason. VnExpress reported on a buyer who found a card in her parcel offering 30% off if she ordered from the brand's own website next time. She ordered through the marketplace app anyway, because the platform guaranteed her refund (VnExpress).

A 30% discount wasn't enough to move her. The "quit the marketplace" camp needs to sit with that. A marketplace doesn't just sell shelf space. It sells trust: safe payment, returns with a referee, thousands of reviews from earlier buyers. A brand-new website has none of that.

Truth two: the bigger the market gets, the fewer sellers survive

Now the part the "just sell on the marketplace" camp rarely mentions.

The same Metric report shows that in 2025, while marketplace revenue grew by almost 35%, the number of shops that made at least one sale fell 7.43%, roughly 50,000 fewer shops than the year before (Thuong Gia). In the first half of the year alone, the year-on-year gap was more than 80,000 shops (VnEconomy).

The pie got bigger, and fewer people got a slice. So where did it go?

Metric answers that too: official brand stores ("Mall" shops) were just 2.12% of sellers but brought in 32.6% of revenue. A marketplace is not a level playing field for small sellers. It rewards brands: names shoppers already know, already trust, and search for directly.

Here's the paradox worth remembering: the brand that helps you win on a marketplace can't be built on the marketplace. Shoppers remember "I bought it on Amazon", not your shop's name. The more you sell, the more the name that sticks is the platform's.

Truth three: on a marketplace you own nothing, not even your customers

Selling on a marketplace is like renting a stall in the busiest mall in town. The foot traffic is real. But management can raise the rent whenever it likes, and it keeps your customer list.

The rent only moves in one direction

No guesswork needed, just history. In Vietnam, Shopee's fixed fee for a typical clothing seller was 2% in 2020. From April 1, 2025, it was 10%, before the 5% payment fee and add-on service programs. A clothing shop was then paying at least 16.5% in fees and tax before spending anything on ads (VnExpress).

Another round arrived in May 2026. According to fee schedules reported in the press (neither platform had replied to the reporter at publication time), platform fees alone came to roughly 24.4–27.6% per order. Add ads, shipping, packaging and returns, and some sellers put their total cost at about 39.5% on Shopee and up to 45.6% on TikTok Shop (VOV).

One seller of handmade wool goods, on marketplaces since 2015, estimates her fees climbed from about 5% to 30% over a decade. The head of a Vietnamese business association said about 90% of her members who sell on marketplaces report that "the more we sell, the more we lose" (VnExpress).

If that sounds like a young market's growing pains, look at the mature one. Marketplace Pulse analyzed seller P&Ls and found that Amazon keeps more than 50% of a typical seller's revenue once referral fees, fulfillment and advertising are counted, up from about 40% five years earlier (Marketplace Pulse). Every mature marketplace follows the same arc: attract sellers with low fees, then raise them once sellers have nowhere else to go.

The customers belong to the platform. You're only borrowing them

This is the most expensive part, and it's not on any fee schedule.

People buy from you on the marketplace, but their data belongs to the marketplace. You don't have their email, you can't tell them about a new product, you don't know who's on their fifth order. To sell again to someone who already bought from you, you pay the platform's cut all over again, plus ad spend to get back into their feed.

And the platform won't let you take them elsewhere. Shopee's position is explicit: a thank-you card or warranty slip with a phone number is fine, but content that advertises, redirects or solicits buyers to transact outside Shopee breaks its policy (VnExpress).

When sellers resort to stuffing discount cards into parcels, it tells you something: they're trying to buy back what should already have been theirs.

The three truths in one line: the marketplace has the shoppers and the trust. But its costs only rise, the winnings flow to brands, and the customers you find there never really become yours.

The opposite trap: "our own store gets us off the marketplace"

The "quit the marketplace" camp is probably nodding by now. Not so fast. I've seen enough sellers pour money into a store and walk away disappointed to know how deep this trap goes.

A website doesn't bring its own traffic. A marketplace has tens of millions of people opening the app every day; your new store has exactly one visitor on day one, and it's you. To get shoppers, you pay for ads, content and SEO. In other words, you take on the job the marketplace fee was paying for. Sellers who move everything to their own store just because they resent the fees often discover that buying traffic costs even more.

Remember the 30%-off card: the shopper still chose the marketplace. A store with no compelling reason to buy there, no clear returns policy, an awkward checkout and a slow mobile experience won't beat the platform's trust, whatever the discount.

So what is your own store for? The answer is an old number that has never stopped being true. Research by Frederick Reichheld of Bain & Company, cited in Harvard Business Review, links a 5% increase in customer retention to a 25% to 95% increase in profits. And depending on the industry, acquiring a new customer costs 5 to 25 times more than keeping an existing one (Harvard Business Review).

Read those two numbers together and the division of labor is obvious:

Finding new customers is the most expensive job. The marketplace does it better than you, and you pay a fee for it. Fair enough.
Keeping existing customers is the most profitable job. Yet on the marketplace you pay the same full fee on every repeat order. That's where money gets burned for nothing.

Use the marketplace to find customers. Use your own store to keep them. That's the whole strategy. Everything else is execution.

The "hunt there, keep here" framework: which stage are you at?

Not every seller needs a store today. I split it into three stages, and to be honest, at stage one my advice is don't build a store yet.

Stage	Signs	Marketplace's role	Own store's role
1. Finding the product	Still testing products, few repeat buyers, nobody searches for your name	Main channel, nearly the only one	Not needed yet. A social profile and a messaging channel are enough
2. Regulars appear	Customers reorder, ask "do you sell outside the marketplace?", the product has a story	Channel for new customers	Home for regulars: reorders, member offers, exclusive bundles
3. A brand	People search for your name on Google, you have exclusive products	Paid marketing channel, for reach	Main sales channel for loyal customers, where your data and margin live

Five questions that tell you it's time

Answer honestly. If you say "yes" to three or more, you're paying marketplace fees on orders that should be yours.

What share of your orders come from returning customers? If you don't know, that's your answer: you don't own your customer data.
After every fee, tax and ad dollar, what's left per order? If your margin is down to single digits, every repeat order you move off the marketplace is margin you get back.
Does anyone search for your brand name on Google? If they do and the top result is your marketplace listing, you're handing customers to the platform at your own front door.
Does your product need a story? Skincare, specialty food, handmade goods, design-led products: people buy the story and the origin. A marketplace listing has no room for it.
Do customers reorder on a cycle? Consumables like skincare, coffee, household goods and pet food: a customer who buys every month is the one most worth keeping.

What your store needs before regulars will buy there

Your store doesn't compete with the marketplace on traffic. It competes on reasons. To bring regulars over, it needs:

A clear reason to buy here. Member pricing, gifts, bundles that only exist on your site, faster local delivery. Without one, shoppers drift back to the habit.
A returns policy as good as the marketplace's or better, stated plainly on the product page. That guarantee is exactly what kept the 30%-off shopper on the platform.
A simple checkout, with the payment methods your buyers already use and as few steps as possible.
Speed on mobile. Your customers are used to the marketplace app. A sluggish site loses them at page load.
Measurement. Without analytics set up properly you won't know how often customers return, what they buy, or where they drop off. I wrote separately about the most common GA4 and GTM setup mistakes.

For a concrete example, see the cosmetics shop template in my template library: a filtered catalog, product pages with size variants, and a cart with a straightforward checkout.

Move customers without breaking the rules

Don't turn your parcels into flyers that pull people off the platform. Besides breaking marketplace policy, it annoys the customer.

The durable route is building a brand strong enough that customers come looking for you: a product with its own name, packaging people recognize, social and YouTube content about your name rather than just the product category. When someone types your name into Google, your store should be the first thing they see. It's why I keep saying that SEO in the AI era is a game of brand names, not just keywords.

The quick math: what does it cost to sell again to your own customers?

This is an illustration with simple assumptions, not survey data. Swap in your own numbers.

Say you have 1,000 regular customers, each reordering 3 times a year, with an average order of $30. Revenue from returning customers:

1,000 × 3 × $30 = $90,000 a year

If total marketplace fees (before ads) are 22%, below the 24–28% reported for Vietnam in 2026 and well below Amazon's all-in take, you're paying:

$90,000 × 22% = $19,800 a year

…just to sell again to people who are already your customers.

Move half of those repeat orders to your own store and you keep around $9,900 a year before the cost of running it. That cost is real: hosting, payment processing, a little marketing to bring customers back. But it's usually a fraction of that figure. See my pricing for what a build costs, or read what a website costs in Vietnam to estimate your own.

And one benefit isn't in the math at all: that list of 1,000 customers is now your asset. If the platform raises fees, changes its algorithm or suspends your account, you still have a way to sell.

The bottom line: whose brand is each order building?

Back to the original question: marketplace or your own store?

My answer: both, with each channel doing its own job. Use the marketplace to find new customers, and treat its fee like an ad budget. Use your own store to keep existing customers, where a repeat order no longer loses a quarter of its revenue on the way in.

The business consequence is stark. A seller who lives only on the marketplace is building the marketplace's brand: every order makes the platform a little more memorable and gives it a little more pricing power. Vietnam's 2025 numbers show where that game ends: the market grew almost 35% while tens of thousands of small shops dropped out. A seller who builds a store at the right moment plays a different game: every repeat order adds to an asset they own.

You don't need to choose between a marketplace and a website. You need to know which customers you're renting, and which ones are yours.

If you're at stage two, with regulars and your own products, and want to know what a "home for your regulars" store should include, let's talk. And if you're still at stage one, my honest advice is: focus on the marketplace and the product first. The website can wait.

Sources: Thuong Gia, Metric 2025 data · VnEconomy, Metric H1 2025 · VOV, May 2026 fees and YouNet ECI data · VnExpress, April 2025 fee increase · VnExpress, sellers pulling buyers off marketplaces · Marketplace Pulse, Amazon's cut · Harvard Business Review, the value of keeping customers

Marketplace or Your Own Online Store? You're Asking the Wrong Question