How Vietnamese consumers find and compare brands

A category review for Vietnam usually arrives with a funnel diagram borrowed from another market. Awareness, consideration, purchase, retention. The diagram is not wrong; every buyer passes through those states.

What the diagram does not say is where each state happens. In Vietnam the answer is different from most markets a foreign team has worked in before, and all three of the important answers are surfaces the brand does not own.

Discovery: the feed is the front door

In markets where search-led buying dominates, a plan starts with the question already formed: someone knows they want the thing and goes looking for it. Search volume is a reasonable proxy for demand, and a media plan built on it is aimed at a real audience.

For a large part of consumer goods here, that is not the sequence. The first contact happens in a feed — short video, social posts, live selling — or inside a marketplace's own browse and recommendation surface. The buyer was not looking. The product arrived.

Search still matters, and it matters a lot for particular categories: services, anything expensive or technical, anything with a specific problem behind it, and most business buying. The mistake is not using search. The mistake is using search volume to size a category that is discovered rather than sought, then reading a thin first quarter as weak demand.

There is a second-order effect that entry teams rarely anticipate. Because discovery happens in the feed, the brand is frequently first encountered through someone else — a creator, a live seller, an affiliate. The first impression is mediated by a person the brand does not employ and often has not briefed. That is a governance question as much as a media one, and it has to be settled before volume arrives rather than after.

Comparison happens on a page you do not control

Once the product is known, the comparison usually takes place on a marketplace listing, where your item appears beside competing sellers offering something very similar.

What the buyer sees on that one screen:

Two things follow. First, almost none of this comparison is visible in your own analytics: the decisive moment happens on a page the platform owns, against competitors you may not have listed. Second, price is decisive only when the other signals look equivalent — and on a crowded listing page they very often do, which is exactly why competing on price feels necessary when it is not.

A newer seller with few reviews and a slower delivery estimate is not losing on price. It is losing on signals, and discounting is the most expensive available way to compensate for that. Building the signals — review volume, delivery speed, seller standing — is slower and cheaper, and it does not have to be repeated every month.

Reassurance happens in a conversation

Between deciding and ordering there is usually one more step, and it is the one foreign plans most often leave out entirely.

The buyer opens a chat thread and asks the questions the listing did not answer. Is it genuine. Will it work for my situation. How fast does it reach my district. What happens if it is the wrong size. Can I get a better price.

Three consequences matter for anyone planning a market entry:

• Response time is a conversion variable. A ten-minute reply and a four-hour reply are two different businesses, and the difference does not show up in the media plan.
• The answer is a sales conversation, not a service ticket. Whoever staffs it is doing sales work, and should be resourced and measured as such.
• Nothing about this step is recorded by default. The advertising platform sees the click that opened the thread. What happened afterwards exists only if someone writes it down.

That last point is where the reporting damage happens. A channel whose job is to start conversations shows a click and then silence, while a channel that closes on a website shows a clean conversion. On a dashboard the second looks better, budget moves accordingly, and the move is rational given the data. Which channels matter in Vietnam covers what to do about it.

Paying on delivery changes the buyer's risk

Cash on delivery is ordinary here, not a fringe option, and it changes the psychology of the comparison in a way that is easy to miss from abroad.

For the buyer, paying on delivery removes almost all the risk of ordering from a seller they have never heard of. They do not have to trust the brand, only the courier at the door. That lowers the barrier to trying something new — which is genuinely good for an entrant, and is part of why discovery-led selling works so well here.

It also makes refusing the parcel nearly costless. So the same mechanism that raises order volume also produces a share of orders that never complete, and that share is not random: it is higher where the order was placed on impulse, and higher again where the decision was driven by a deep discount.

The practical consequence for a plan is one sentence: an order is not revenue yet, and the gap has to be measured by acquisition source rather than assumed as a company-wide average. Everything that follows from that — which definition of revenue the business runs on, and what it does to every efficiency figure — is covered under whether the data can be trusted.

What "brand" means when the listing is the storefront

For an entrant, the most uncomfortable implication is this: in many categories, the marketplace listing is the storefront, and the brand's own website is not where the transaction happens.

That does not make the website pointless. It changes its job. The site is where a buyer checks that the company is real before ordering somewhere else — an address, a story, a phone number, products that match what they saw. Judged on its own conversion rate it will look like a poor performer. Judged on its contribution to decisions taken elsewhere, it is doing the work asked of it.

It also changes what brand investment buys. In a listing-mediated market, the returns to being recognised are concrete rather than abstract: a recognised name survives the comparison screen better, needs less discount to be chosen, and gets the benefit of the doubt in the chat thread. None of that shows up as attributed revenue, and all of it shows up in the cost of every other channel.

Who owns the customer after the first order

The last structural point concerns what happens next, and it is the one with the longest consequences.

When the sale completes on a marketplace, the platform mediates the relationship. Repeat purchase is likely to happen through the platform's own surfaces, and the data about who bought what is largely theirs. A business built entirely inside that arrangement can be perfectly profitable and still have no independent way to reach its own customers.

The reasonable response is neither to avoid marketplaces nor to ignore the issue. It is to put two numbers in the monthly report from the start:

• What share of contribution depends on a single platform.
• What share of buyers the company can reach without that platform — through a message channel they opted into, an owned account, or a repeat purchase through your own site.

Neither number requires a decision today. They make sure that when a decision is required — a fee change, a ranking change, a policy change — the company knows where it is standing. The fee side of that picture sits under marketplace advertising.

Three journeys, drawn out

The general shape splits into roughly three journeys, and most categories sit close to one of them. Locating yours settles more planning arguments than any single statistic.

The middle column is where most entrants land and where the reporting problem is worst, because the journey deliberately crosses three surfaces owned by three different parties. Discovery is credited to nobody in particular, comparison is invisible, and the sale lands in a marketplace report.

The right-hand column looks the most familiar to a foreign team and is the most likely to be measured badly for a different reason: an enquiry form and a chat message and a referral all arrive labelled "lead", with closing rates that differ by several times. Averaging them produces a cost per lead that describes no real group of buyers.

One caution about the table. It describes how the purchase is made, not who the buyer is, and the two are frequently confused in entry decks. A category is not "young" or "urban" because its journey runs through short video; it runs through short video because that is where the category is discovered. Segmenting on demographics when the real variable is journey shape is how a plan ends up with precise targeting aimed at the wrong step.

Finding out how your own category is bought

Everything above is the general shape. Categories differ enough that the general shape is not a plan, and the specific answer is cheaper to obtain than most teams expect.

Three sources, none of which requires commissioning anything:

• The search terms used inside your own site or store. Whether people type a product name or a description of a problem tells you which half of the discovery question you are in.
• The first question asked in chat. Fifty of them, read in one sitting, will tell you what the listing fails to answer and what the real objection is.
• The reviews on the three best-selling competing listings. Not your own reviews — theirs. This is the closest thing to a free record of what buyers in your category actually care about, written by them.

Commission research when you need to size the opportunity, test a price, or understand a segment you cannot observe. Do not commission it to discover a sequence that fifty chat threads would have shown you in an afternoon.

Where this work stops

MWY does not run campaigns, produce content, or take commission from any platform or agency.

Sizing a category properly — who buys, how many, at what price, and how they decide — is Vietnam Market Research. Turning that into a sequence of channels and a budget that matches how the category is bought is Go-to-Market Strategy.

What MWY does not do is predict consumer behaviour from aggregate market reports. A published figure about a national market says very little about a specific category, and a plan built on the first is a plan that has skipped the second.