Should a foreign brand sell on Vietnamese marketplaces?

A head office asks the question in a form that has no good answer: are Vietnamese marketplaces worth it? The number that comes back is a commission rate, someone compares it with a margin, and the decision gets made on one line of arithmetic.

The arithmetic is not wrong. It is just answering a smaller question than the one being asked. What a marketplace costs is knowable and stable. What it gives you, and what it quietly stops you from building, is the part that decides whether the first year was well spent.

There are three routes in, not two

Most board discussions frame this as marketplace versus own store. In practice a foreign brand entering Vietnam is choosing between three routes, and they carry very different consequences.

None of these is the correct answer in general. They are correct for different things, and the failure mode is picking one for a reason that belongs to another.

The most common mistake is choosing distributor-led for speed and then being surprised, eighteen months later, that nobody at head office can answer basic questions about who is buying. Speed was real. The blindness was the price, and it was not written down anywhere at the time.

What a marketplace actually rents you

It is worth being precise about the thing being bought, because the word "channel" hides it.

A marketplace rents you access to an audience that already trusts the platform. Not the audience itself. The trust does not transfer to you, and neither does the habit. If you stop paying, the access stops with it, and what remains is a review count on a page you do not own.

That is not an argument against it. Renting access is often exactly right, particularly when the alternative is spending a year and a large budget to build the same access yourself. It is an argument for being honest in the board paper about which line of the plan is an asset and which is a rental.

The second thing being rented is infrastructure: payments, delivery, returns handling, and a dispute process buyers already understand. For a brand with no local entity and no logistics, this is genuinely hard to replicate and easy to undervalue from a distance.

The question is sequencing, not preference

Once you accept that the routes answer different questions, the decision stops being ideological.

A sequence that works for a large share of brands looks like this: use the marketplace to find out whether demand exists at a price that works, then build the owned channel on top of the answer rather than in parallel with the question.

The reverse sequence — building the owned channel first, then adding marketplaces when traffic proves expensive — is not wrong, but it is slower and it front-loads the spending. It makes sense when the product needs explanation before someone will buy it, or when the brand arrives with an audience already in place.

What rarely works is running both at full effort in year one with a small team. Not because of budget, but because the two routes need different operating rhythms, and the team ends up doing neither properly while reporting on both.

What you can learn there that you cannot learn elsewhere

The underrated argument for starting on a marketplace is not revenue. It is the speed and honesty of the feedback.

• Whether the price works. Not whether a survey says people would pay it. Whether they do, next to visible alternatives, on the same screen.
• Which variants actually move. Assortment assumptions carried in from another market are wrong more often than they are right, and this is the cheapest place to find out.
• What the real competitive set is. Often it is not the brands head office listed. It is whoever appears beside you in the same search.
• How seasonal the category is. Campaign peaks are heavily concentrated here, and a plan built on a flat monthly average will be wrong in both directions.

Answers to these questions are worth having before committing a full year of budget, and getting them any other way is slower and less reliable. Where this fits against the wider entry decision is covered under Vietnam market entry.

What it will not tell you

The same channel is close to silent on a second set of questions, and they are the ones a brand needs in year two.

Who bought. You will see orders, not people you can speak to again. Communication is expected to stay inside the platform.

Why they bought, and why others did not. There is no view of the people who considered you and left, which is where most of the useful information about positioning lives.

Whether they come back. Repeat behaviour is visible to the platform far more clearly than to you, and the platform is under no obligation to show it.

This is the structural reason marketplace-only strategies get harder rather than easier over time. Every year of growth increases the share of demand that exists only inside someone else's system, and the cost of moving that demand later rises with it.

A twelve-week test that answers the question

Rather than debating the strategy, most brands are better served by running a bounded test and deciding from what comes back. The shape below is the one I would propose; the numbers are illustrative, not benchmarks.

Three rules make the difference between a test and twelve weeks of activity:

• Change one variable at a time. Running a price change and a promotion together produces a number that cannot be attributed to either.
• Hold the budget steady. Varying spend inside the test window makes week-on-week comparison meaningless.
• Decide the decision rule in advance. Write down, before launch, what result leads to expanding, what result leads to changing price, and what result leads to stopping. Deciding afterwards means deciding from the result you got rather than the result you needed.

Twelve weeks of this costs less than most entry research exercises and answers a narrower set of questions far more convincingly. It does not replace research into the category and the competitive set, which is a different job with a different method — see Vietnam market research.

Four things to write down before launch

Most of the argument that follows a marketplace launch can be prevented by settling four things in advance, while nobody is defending a result yet.

The first is the price floor, stated as a number below which you will not discount regardless of what the platform's campaign calendar offers. Promotional participation is a recurring decision here, the invitations are frequent, and each one in isolation looks reasonable. Without a floor agreed at the start, a year of individually sensible decisions ends somewhere no one would have chosen deliberately.

The second is the share of revenue you are willing to have inside one platform. This is a strategy question wearing the costume of a reporting metric. Say it out loud early: is forty per cent acceptable, is seventy, is there any level at which the answer becomes no. A brand that never sets the number does not avoid the decision, it just makes it by accident and finds out later.

The third is which part of the plan is a rental and which part is an asset, one sentence each. This is the line that keeps a board discussion honest two years later, when marketplace revenue is the largest number on the page and somebody asks what the company actually owns.

The fourth is who holds the seller account. If a distributor or a service provider holds it for practical reasons, write down now what happens to the sales history, the reviews and the store rating if that relationship ends. This is the item most often left out and the one that costs the most to fix afterwards, because a rating built over two years does not transfer.

All four belong in the same document as the launch plan, not in a separate risk register that nobody reads.

What changes after the first year

The economics of renting access do not stay still, and they move in a direction that is easy to miss while the revenue line is going up.

In year one, platform access is clearly cheaper than building your own. In year three, two things have usually changed. The share of your demand that exists only inside the platform has grown, which means the cost of leaving has grown with it. And the cost of acquiring the same customer has typically risen, because more sellers are competing for the same placements in a category you helped prove out.

Neither of those is a reason to avoid the channel. They are reasons to start building something you own while the marketplace is working, rather than after it stops working. The brands that handle this well tend to begin the owned channel at the point when marketplace performance is at its best, which feels counterintuitive and is precisely why it rarely happens.

A practical version of that: once the twelve-week test has answered the demand question, begin collecting whatever direct relationship the platform rules permit — packaging inserts, post-purchase registration, anything that gives a reason to come to you next time. The volume will look trivial against marketplace orders for a long while. It is the only part of the picture that compounds in your favour.

Three conditions that make marketplace-first the wrong call

Three situations turn the usual recommendation around.

The product needs explaining before anyone will buy it. Marketplace formats are built for comparison, not for teaching. A product that requires a specification conversation will convert badly, and the resulting data will read as absent demand rather than as the wrong venue.

Price positioning is the strategy. If the brand's position depends on not being seen next to cheaper alternatives, a marketplace listing costs more than it earns, and the damage is not visible in the first year's numbers.

The category is dominated by price-led local sellers with a cost base you cannot match. Then the honest conclusion is that the route exists but the economics do not, and it is better to find that out from a twelve-week test than from a year of it.

Where this work stops

This is a marketing decision, not a legal or a tax one. Entity requirements, import duties and product registration sit with local counsel, and the answers can change the commercial picture enough that the marketing plan should follow them rather than lead.

It also stops at profit after advertising cost. Landed cost, inventory, returns handling and the cost of local operations sit outside the scope of independent marketing oversight, so any margin used in the modelling above is a figure the company supplies rather than one derived from marketing data.

Deciding the route and building the first plan around it is the substance of go-to-market strategy. Rebuilding the numbers once a channel is already running, and saying plainly which of them can be trusted, is what a digital marketing audit is for.