How to market in Vietnam: the first ninety days

Once a foreign company has decided to enter Vietnam, the next request is usually for a media plan: which platforms, how much, starting when.

A more useful request comes first. What should we know at day ninety that we do not know today? A quarter designed around that question produces fewer sales than one designed to sell, and far more to plan the rest of the year from. The four decisions that belong before any spending — which revenue definition to use, who owns the accounts, whether marketplaces are a channel or the channel, how chat sales get recorded — are assumed settled here. This is about what follows them.

What ninety days can tell you, and what they cannot

A quarter of spend, used well, answers a specific set of questions about your own product in this market:

• Where the category is discovered, and where the purchase actually completes.
• What an accepted order costs in two channels, measured end to end.
• Whether the offer covers its acquisition cost at small scale.
• What buyers ask before they buy, and how long they take to decide.
• What share of orders starts in a conversation rather than on a page.

It does not answer whether the brand will be recognised in two years, how large the full opportunity is, what a customer is worth over a lifetime, or what happens at five times the spend. Those questions matter. They are answered by research and by time, and a first-quarter plan that pretends to answer them usually answers none of the questions above either.

The distinction shapes everything that follows. A quarter built to learn holds variables still, funds few channels properly, and judges them on cost per accepted order. A quarter built to sell changes everything at once whenever the numbers disappoint, and ends with a sales figure nobody can explain.

Where day one falls matters more than it looks

The Vietnamese calendar has stretches when a test measures the season rather than the plan.

The weeks before Lunar New Year change three things at once: media gets more expensive as every advertiser competes for the same attention, buyers behave differently, and delivery capacity tightens. The weeks after Tết are slow across most categories. The large marketplace sale days — the double dates such as 9.9, 11.11 and 12.12 — distort traffic, advertising costs and the mix of buyers inside the platforms for days either side.

None of this is a reason to delay a launch. It is a reason to place the test month, the middle thirty days, in an ordinary stretch of the calendar, and to extend the plan rather than read a test that ran through a peak.

There is also a set of lead times marketing does not control: the legal entity, advertising account verification, marketplace store approval, and, for categories such as cosmetics, supplements, healthcare and education, advertising content rules that should be cleared with counsel before any creative is made. These gate day one. Start them first and put their dates on the same page as the media plan.

Days 1–30: build what the market needs, then check it works

The first month is mostly construction. What gets built is specific to this market:

• The place where the purchase completes. A marketplace store with complete listings, or a site that accepts the payment methods buyers here actually use, including cash on delivery and bank-transfer QR codes. Whichever it is, it has to be finished before traffic arrives.
• Assets made for this market. Not translated: the offer, the way the price is presented, the proof a first-time buyer needs, and a product explanation in the short-video formats where discovery happens.
• Conversation handling, staffed. Who answers Zalo, Messenger and marketplace chat, in which hours, how quickly, and where they record whether each conversation ended in an order.
• Measurement. A source attached to every order, delivery outcomes joined to orders, marketplace settlement exports pulled per order. The detail of what that involves sits under measuring marketing in Vietnam.

Then comes a small calibration spend — enough to generate real orders across the intended channels, not enough to judge anything. Its only purpose is to prove the chain works: that an order from each channel can be followed from the ad to the accepted delivery.

The exit condition for month one is simple and strict. If an order cannot be traced end to end from every channel in the plan, the tests do not start. Running them anyway produces a month of spend with results nobody can attribute.

Days 31–60: two tests, written down before they start

The second month runs two tests: one channel whose job is discovery, and one whose job is closing. Which channels those are depends on how your category is bought, and the reasoning is set out under which digital channels matter.

Each test gets a single sheet, completed before the first unit of budget moves:

Three disciplines make the month worth its cost. Hold the offer and the price still, because a test that changes the offer halfway through measures two things and learns neither. Rotate creative only within a set prepared in advance, so that "trying something new" does not become a third variable. Do not add a third channel because the first two look slow in week six; slow in week six is normal, and a third channel makes both results unreadable.

An underfunded test is worse than no test. It produces an inconclusive number that still gets quoted in the board pack as though it were a finding.

Days 61–90: read the results, then write one page

Two things happen in the third month, and the order matters.

First, the data matures. Orders placed in weeks five to eight are still being delivered, refused, returned and settled. Reading them in week nine counts orders that will never become money. So spend continues at a steady level on both channels while the earlier orders complete, and the reading is done on orders old enough to have finished their journey.

Then the reading itself, with each channel set against the most the company can afford to pay for an accepted order — a ceiling calculated from the margin the company supplies, after local fees. An illustration, with invented numbers:

Read channel by channel, the conclusion looks obvious: cut discovery, scale closing. But suppose most of the closing channel's orders came from people searching for the brand by name, and that searches for the brand barely existed before the discovery test began. Then the closing channel is harvesting demand the discovery channel created. Read as a pair, the two sit comfortably under the ceiling. Cut the first, and the second loses the demand it was harvesting.

This is the most common misreading at day ninety, and it is avoidable only if the pair was planned as a pair from the start — including a way to see how many closing-channel buyers had met the brand before.

The output of the month is a day-90 memo, one page, with five parts:

• What we now know, with the numbers and how confident each one is.
• What we still do not know, and what it would take to find out.
• The ceiling, and where each channel sits against it.
• The decision: which channel to scale, in what steps; which to hold; which to stop; whether to test a third.
• The date of the next decision.

The reserve held back at the start of the quarter is released against this memo, not against the original plan.

How long launching a brand here takes

The question gets asked as if it had one answer. It has three, running at different speeds.

The setup clock covers the entity, the accounts, the marketplace store and any regulated-category approvals. Its pace is set by parties outside marketing, and it is the one most often underestimated because nobody in the marketing plan owns it.

The learning clock is roughly one quarter of spend to reach one reliable read on what a customer costs. It cannot be compressed much, because orders need time to be delivered, accepted and settled before they can be counted, and a test needs weeks to leave the platforms' learning period.

The brand clock is how long it takes for recognition to lower the cost of every other channel: fewer discounts needed to win a comparison, more benefit of the doubt in a chat thread, more people searching by name. That is measured in years. Day ninety cannot show it, and a day-90 memo should not be used to dismiss it either.

A launch plan that promises all three in one quarter is committing to the second and hoping for the others.

What to leave out of the first ninety days

• Annual commitments — retainers, media contracts, full-year creative. They lock in the plan before the plan has been tested.
• A brand campaign at scale. It runs on the brand clock, and it will be judged on the learning clock.
• More than two funded channels. Coverage feels safe and makes every result unreadable.
• Targets copied from another market. The break-even here depends on local margin after local fees; a benchmark from elsewhere is a number nobody built for this market.
• A launch built around a marketplace sale day. It produces volume with a discount-driven mix of buyers, and leaves no baseline to compare the next month against.
• Creator fees paid flat for reach with no way to track what followed. Creators and affiliates paid per result are a different matter: that is variable acquisition cost, and it can be judged like any other channel.

The plan on one page

A Vietnam go-to-market plan that cannot fit on one page usually has its important parts buried in the appendix. The page needs these rows and not many more:

The last row does the most work. Every item without an owner becomes an assumption, and assumptions are what the day-90 memo ends up explaining.

Where this work stops

MWY builds this plan and does not execute it: no media buying, no content production, no commission from any platform, agency or partner. That is what allows the plan to recommend fewer channels, a smaller first quarter, or a later test month when the calendar calls for it.

Sizing the opportunity — how many buyers, at what price, served by whom — comes out of Vietnam Market Research. The channel sequence, the partner shortlist and the ninety-day launch plan with owners and dates are the deliverables of Go-to-Market Strategy.

Once spending is live, reading the day-90 memo and each month after it from the company's side of the table is Digital Marketing Advisory & Oversight. What MWY will not do is forecast a return before the measurement exists to check it against. The first quarter is where that measurement gets built.