Vietnam e-commerce market size, and how to size your own opportunity without buying a report

Ask four respectable sources how large e-commerce in Vietnam is and you will get four different answers, the largest more than twice the smallest. Head office usually notices this the week before a board meeting, picks the number that looks most official, and builds the plan on a share of it.

Neither step is unreasonable, and both are usually wrong for the purpose. The spread between the figures is not a sign that someone is mistaken. It is a sign that they measure different things. And even the right total is the wrong number for an entry plan, because no brand sells into the whole of it.

This piece does three things: explains what each published figure counts, puts Vietnam beside Thailand on the most comparable basis available, and sets out how to size one category from the bottom up with evidence that costs far less than a commissioned study.

Four published figures for one market

The two marketplace trackers are the closest to each other, and they still differ by a fifth, because one covers a calendar year and the other a twelve-month window that starts and ends mid-year, and because each estimates sales from platform data in its own way. Metric’s half-year figure for 2026, 291.6 trillion dong, up 44% on a year earlier, shows how quickly any one of these numbers dates.

Why they differ: scope, period and method

Three differences explain nearly all of the spread.

Scope. A marketplace tracker counts what is sold on Shopee, TikTok Shop, Lazada and Tiki. It does not count sales through a brand’s own website, through Facebook pages and chat, or through retailers’ own apps. A regional report counts e-commerce more broadly. A national index of business-to-consumer online retail is broader again; the summary page of the VECOM report does not set out every line it includes, which is itself a reason to quote it with care.

Period. The four figures cover three different twelve-month windows. In a market growing by a fifth or more a year, a year’s difference in the window is a large difference in the number.

Method. Trackers read sales from what platforms display. Regional reports model the market from industry data and interviews. National indexes survey businesses and, in VECOM’s 2025 edition for the first time, used provincial postal and delivery data. Gross merchandise value, in every version, counts orders placed before cancellations, refusals at delivery and returns, so it describes where buying happens rather than what sellers keep.

The practical rule follows: quote one figure, with its source, its scope and its period, and never add or compare figures from two sources in the same sentence.

Vietnam vs Thailand: the comparison head office will ask for

Regional teams often weigh Vietnam against Thailand. The most comparable basis is a single source that measures both countries the same way, and e-Conomy SEA does.

Another tracker points the same way. Momentum Works’ report on platform e-commerce in Southeast Asia, published in April 2026, ranks Vietnam third in the region for 2025 behind Indonesia and Thailand, with Thailand’s platform sales growing faster that year.

Read carefully, the table says less about which market is better than about where the growth comes from. Vietnam has more people and a smaller share of them online, and each internet user spends a little over half of what a Thai user does. Thailand’s market is larger per buyer. For a premium product, the Thai figure may be the more relevant one; for a mass product, Vietnam’s larger population still coming online may be. Neither total answers that for a particular category.

The comparison also leaves out most of retail. Vietnamese ministry officials put online retail at between 11% and 12% of total retail sales of goods and services in 2025. Most purchases in most categories still happen in a shop, a market or a supermarket, and that share differs far more between categories than between Vietnam and Thailand. A brand that will sell mainly through distributors and modern trade should treat any e-commerce total as a measure of where buyers research and compare, not of where its own revenue will come from.

What a national total cannot tell you

A national total is useful for one decision: whether a market is large enough to deserve a look at all. Vietnam clearly is. After that, the total stops helping, for four reasons.

• Price band. In many categories, much of the value sits in price bands below where an imported brand will sell. A share of the total is mostly a share of buyers who will never consider the product.
• Channels. The brand will start in two or three channels, not all of them. Sales on the others are not available to it in year one, whatever their size.
• Cities. Most entries begin in Hanoi and Ho Chi Minh City. Buyers elsewhere are part of the national figure and not part of the first plan.
• Category. The total is a sum of categories that grow at very different rates. The category’s own size and growth matter; the average does not.

This is the mechanism behind one of the most common planning errors in Vietnam, described among the mistakes foreign brands make here: taking a modest-looking share of a national total and sizing a budget to it.

Sizing from the bottom up: five steps

A bottom-up estimate starts from buyers rather than from the total, and narrows them in the order the brand will meet them.

• Who could buy. The households or people in the target cities who have the need at all. Official population and household statistics, and published counts of internet users, are enough for this step.
• Who buys online, in your channels. The share of those buyers who purchase the category online, and on the platforms you will open first. This is where evidence about where Vietnamese consumers buy does the work.
• Who buys at your price. The share who buy in the price band the product will sit in, after platform fees and promotions are taken into account rather than at the list price.
• How much they buy. Purchase frequency times the price actually paid, for a year.
• What share is defensible. The share the brand can plausibly win by year two, given the competitors already in the band.

Each step should carry its source and a confidence level. The value of the method is not precision. It is that every assumption is written down, so the board can argue with a specific line instead of with the conclusion.

A worked example: one category, two numbers

The figures below are invented for the example. An imported baby skincare line plans to sell at about USD 15 a bottle.

The top-down version is the one that usually reaches head office: e-commerce in Vietnam is USD 25 billion, and 0.1% of it is USD 25 million. Modest, by the look of it.

The two answers differ by a factor of eighteen. Neither is proven, but only one of them can be checked.

The check is against observed sales. Suppose the twenty best-selling listings in that price band on the two largest marketplaces sold about 52,000 units between them over a recent month, read from marketplace analytics data or from the change in their units-sold counters. At USD 15, that is roughly USD 9.4 million a year, or a little over half of the USD 17.3 million band estimate. For a category where a few leading products take most of the sales, that is plausible, and the estimate stands. If the same listings had sold 15,000 units, they would account for about a sixth of the estimate, which is hard to believe in a concentrated category; the online share or the price-band share would be too high, and the range would come down.

Presented to the board, the result is a range with its weakest assumption named: for example, year-two sales of USD 0.9 million to USD 1.8 million, most sensitive to the share of buyers at the price band. That is a smaller number than the top-down one. It is also one the Vietnam team can be held to.

Three checks before the number reaches the board

A bottom-up estimate can be wrong in quieter ways than a top-down one. Three checks catch most of them.

Does each step use the same buyers? The most common slip is to take the number of households from one definition, the online share from a survey of a different group, and the purchase frequency from interviews with a third. Each step should narrow the group before it, not switch to a new one. Where the sources do not line up, say so in the assumption column rather than smoothing it over.

Can the share be named? A defensible share is one that can be pointed at: the brands whose sales it would come from, and why buyers would move. If the year-two share is larger than what the third-best listing in the band sells today, the plan is assuming the brand will outsell established competitors within two years, and that assumption deserves its own line and its own evidence.

Does the number survive the cost of getting it? Year-two sales that do not cover the cost of the channels needed to reach them are not an opportunity, whatever their size. The estimate should sit next to a rough view of what reaching that share costs, including platform fees and promotion depth, so the board sees both sides on the same page.

None of the three requires new data. They require the person who built the estimate to read it again as if someone else had written it.

Where the inputs come from without buying a report

Most of the inputs for a first estimate are public or cheap.

Primary consumer research has a place, usually later and narrower than expected: to test a price point, or to measure a buyer segment the public evidence cannot see. When that point is reached, choosing a market research firm in Vietnam covers what to ask before signing.

Where the figures in this article come from

Q&Me’s Vietnam E-commerce Market Report 2026, as reported by The Investor in June 2026, covering May 2025 to April 2026. Metric’s full-year 2025 marketplace data and its first-half 2026 figures, as reported by the Vietnamese business press. e-Conomy SEA 2025 by Google, Temasek and Bain, published in November 2025, with country figures as reported by Vietnam Investment Review and Thai business press. VECOM’s Vietnam E-commerce Index 2025, published in April 2025. Statements on online retail as a share of total retail by officials of the Ministry of Industry and Trade, as reported by the Vietnamese press between December 2025 and February 2026. DataReportal’s Digital 2026 reports for Vietnam and Thailand. Momentum Works’ fourth report on e-commerce in Southeast Asia, published in April 2026.

Gross merchandise value counts orders placed, before cancellations, refusals and returns. Dollar conversions of dong figures are approximate and rounded. The per-user figures are our calculation from the two sources named in the table. The worked example uses invented numbers.

Where this work stops

MWY does not run consumer fieldwork, own a panel or resell third-party reports, and takes no commission from research firms or data providers. What we do is build the estimate toward the decision it has to support.

Vietnam Market Research is six weeks on one product line: the bottom-up size of the category, the price band that holds, competitor presence by channel, and a written answer on whether Vietnam deserves a budget, with what could not be established marked as such. The same research opens Go-to-Market Strategy, which adds the entry plan, so a brand that continues does not pay for the research twice.