VNG and Zalo: Vietnam’s First Unicorn and the App That Beat Facebook Messenger

VNG is the company that proved a Vietnamese consumer-internet business could reach a billion-dollar valuation, and then spent a decade discovering how hard it is to turn that into a second act. Its games business has generated cash for twenty years; its messaging app won a market that Facebook, WhatsApp and Viber all wanted; and yet the group has struggled to find a business beyond those two that earns a return. This article traces how a games publisher built the country’s dominant chat app, why Zalo beat Messenger, what the money has been spent on since, and what the last three years of listings, losses and investigations say about the risks of scale in Vietnam’s tech sector. It is part of the Vietnam Company Stories hub.

How did a games company become Vietnam’s first unicorn?

VNG became a unicorn because it controlled the distribution of online games in Vietnam at the moment the market exploded, and used that cash to build a portfolio of consumer products. A 2014 estimate by the World Startup Report valued the company at about $1 billion, the first for a Vietnamese startup.

The origin story is specific. In 2004 Lê Hồng Minh, a former PwC and Vina Capital analyst who spent his spare time at internet cafés, co-founded VinaGame with a group of friends and a small amount of seed money. Their insight was that Vietnamese gamers were playing pirated or foreign-hosted titles with terrible latency, and that a locally hosted, officially licensed game with local payment could capture a market that did not officially exist.

The product was Võ Lâm Truyền Kỳ, the Vietnamese edition of Kingsoft’s Swordsman Online, launched in 2005. It became a national phenomenon, with hundreds of thousands of concurrent players at its peak, and it was monetised through prepaid scratch cards sold at internet cafés, phone shops and convenience stores across the country. VinaGame built the card distribution network itself, and that network, plus the café relationships that came with it, was the real asset. It let the company launch further titles and, from 2007, the Zing suite of web products, including Zing MP3, the Zing Me social network and Zing News.

The company renamed itself VNG in 2009 to signal that it was no longer only a games publisher. Tencent, which was itself built on a similar mix of games and messaging in China, had taken a stake early through an affiliate, and other investors including GIC and later Temasek followed. By the time of the 2014 valuation VNG had more than a thousand employees, a dominant position in online games, the largest domestic music-streaming service and a two-year-old messaging app that was about to become the most important thing the company had ever built.

Why did Zalo beat Facebook Messenger?

Zalo won because it was engineered for the network conditions, devices and habits of Vietnamese users in 2012, when Facebook Messenger was a heavy app designed for American smartphones on fast networks. Speed, data efficiency and phone-number contacts gave Zalo a lead that network effects then locked in.

Zalo launched in December 2012 under Vương Quang Khải, a VNG co-founder and engineer who had led the Zing products. The first version was rebuilt from scratch after an initial cross-platform attempt proved too slow. The team’s obsession was performance on the phones and networks Vietnamese people actually had: low-end Android devices, congested 3G in the cities and 2G in the provinces. Messages sent instantly, voice notes compressed well, and the app did not drain battery or data. That sounds trivial; in 2013 it was the whole product.

The second factor was identity. Zalo used phone numbers as the contact graph, as WhatsApp did, rather than requiring a Facebook account. In a country where most people had a mobile number long before a social-media profile, and where small traders, drivers and family groups communicated by phone, that made adoption frictionless. The app added group chats, stickers designed for Vietnamese humour, and a timeline feature that gave it a social layer without turning it into a Facebook clone.

Growth compounded. Zalo reported 10 million users in 2014, 100 million registered accounts by 2018, and market-research firms such as Decision Lab began recording it as the most-used messaging platform in Vietnam around 2020 and 2021, ahead of Messenger. By 2024 VNG cited on the order of 75 million monthly active users. Facebook remains the largest social network in Vietnam, but for the act of sending a message, Zalo is the default, which is a position no local app holds against Meta in any other Southeast Asian market. Viber, which had a strong early position, faded; WhatsApp never gained traction.

How does VNG actually make money?

VNG makes most of its money from online games, which contribute well over half of revenue, followed by advertising and business services on Zalo and Zing, and a small but growing cloud and enterprise segment. Payments and other ventures lose money, and have done so for years.

The games business has evolved from licensing Chinese role-playing titles to publishing global franchises. VNGGames distributes PUBG Mobile in Vietnam, took over League of Legends and related Riot titles from Garena in 2023, and publishes a rotating catalogue of mobile games across Southeast Asia. It is a publishing and operations business rather than a developer: VNG localises, markets, runs servers and handles payments, and shares revenue with the studio. Margins are healthy but the hit-driven nature of the business means revenue swings with the catalogue.

Zalo’s monetisation has been cautious. For most of its first decade the app carried no advertising in chats and earned money mainly through Zalo Official Accounts, a business-messaging product used by banks, retailers, government agencies and small traders to reach customers. In 2022 VNG introduced paid tiers for business accounts and restricted some features for free users, which triggered a backlash from small traders who had built their customer lists on the platform, but the company held the line. Zalo also became the delivery channel for public services during the pandemic, which cemented its role but did not directly generate revenue.

The financial picture at group level has been uncomfortable. Revenue was on the order of VND 8.6 trillion, roughly $350 million, in 2023 and higher again in 2024, but the group reported net losses in 2022 and 2023 in the region of VND 1 trillion or more each year, driven by ZaloPay, losses at associates and the write-down of its Tiki stake. Management guided toward break-even in 2024 and reported a narrowed result, and the market reaction suggested investors had discounted the losses but had not been persuaded that the ventures would ever pay.

What happened when VNG tried to build a second and third business?

VNG’s attempts to build businesses beyond games and Zalo have mostly consumed capital. ZaloPay has lost money since launch, the stake in e-commerce platform Tiki was written down to a fraction of its cost, and smaller bets in logistics and B2B commerce have been quietly reduced.

ZaloPay, launched in 2016, was the obvious move: a payments layer on top of the country’s biggest messaging app, following the Tencent playbook with WeChat Pay. It has not worked as planned. MoMo had a five-year head start, deeper merchant coverage and more capital, and the bank-led QR standard that arrived in 2021 made free bank-to-bank transfers available inside every banking app, eroding the case for any wallet. ZaloPay has grown, but at the price of cumulative losses in the trillions of dong, and VNG has repeatedly had to fund it from games profits.

Tiki was a different kind of misjudgement. VNG invested in the e-commerce platform in 2016 and increased its stake in subsequent rounds, at a time when Tiki was the most credible local rival to Shopee and Lazada. By 2022 Tiki’s share had collapsed under the weight of Shopee’s subsidies and TikTok Shop’s arrival, and VNG had written its holding down to close to zero. The lesson, shared with several other Vietnamese investors, is that being local is no defence against regionally funded competitors in a market with no switching costs.

Other ventures show a pattern of small stakes in adjacent sectors: Telio in B2B wholesale, Ecotruck in logistics, Got It in gifting, Haravan in merchant software, plus a data-centre and cloud business that competes with Viettel and FPT for enterprise customers. None has yet become material. The most recent bet, GreenNode, offers Nvidia GPU capacity from a data centre in Thailand for AI workloads, a business with obvious demand but capital intensity of a kind VNG has never managed before.

Why did the Nasdaq IPO stall?

VNG filed for a Nasdaq listing in August 2023 through a Cayman Islands holding company, VNG Limited, then postponed it within weeks, citing market conditions. The underlying problems were a loss-making profile, a complex ownership structure and a valuation gap between the 2014 unicorn label and what US investors would pay.

The listing plan was long in the making. VNG had explored a Nasdaq IPO as early as 2017, signing a memorandum with the exchange, and the 2023 filing was the culmination of a restructuring in which Vietnamese shareholders, including the founder, exchanged shares in the Vietnamese operating company for shares in the offshore vehicle. The prospectus disclosed a structure familiar from Chinese listings, with foreign investors holding economic rights through a chain of entities designed to comply with Vietnam’s foreign-ownership limits on internet businesses.

The timing was poor. Global technology IPOs were weak in 2023, the company was reporting losses and the filing arrived shortly after VNG’s January 2023 listing on Vietnam’s UPCoM market, where the thin float had sent the share price on a wild ride from an initial reference in the low hundreds of thousands of dong to well over VND 1 million within weeks, before falling back. That volatility made the domestic price a poor anchor for an international book-build. In September 2023 the company announced the offering would be delayed.

The deeper issue is that the two things investors valued, games cash flow and Zalo’s reach, do not combine into a growth story without a profitable third business, and VNG had not yet found one. A Nasdaq listing remains a stated ambition, and the upgrade of Vietnam’s own market may eventually make a domestic main-board listing more attractive, but as of mid-2026 neither had happened.

What happened when the founder was called in by the police?

In September 2024 Vietnamese media reported that Lê Hồng Minh had been summoned by police in connection with an investigation, and VNG confirmed it was cooperating with the authorities. The company appointed a long-serving executive, Kelly Wong, as acting chief executive, and in early 2025 restructured leadership with Lê Hồng Minh as chairman and Wong as CEO.

The episode was alarming because VNG had, more than any other Vietnamese tech company, been identified with its founder. Lê Hồng Minh had been chief executive and public face since 2004, held a large personal shareholding and was the architect of every strategic decision from Swordsman Online to the Nasdaq plan. The share price fell sharply on the news, and the company’s disclosure was minimal, saying only that it was working with the authorities and that operations continued normally. Details of the inquiry were not made public in a form that allowed outsiders to assess it.

What followed was a demonstration of institutional depth that surprised some observers. Kelly Wong, who had joined VNG in 2013 from a background in games publishing and had run the games unit, took over day-to-day management. The business did not stall; the games catalogue continued to perform, Zalo kept growing and the 2024 results showed the narrowing of losses that management had promised. By January 2025 the board had formalised a split of roles, with the founder moving to chairman and Wong confirmed as chief executive.

The broader context is Vietnam’s anti-corruption campaign, which since 2021 has reached into banking, property, healthcare and technology, and which has made every large private company nervous about the boundary between commercial activity and regulatory exposure. The cases described in the analysis of Van Thinh Phat and SCB are of an entirely different scale and nature, but they set the climate in which a police summons of a tech founder becomes a market event. For VNG the practical consequence was that the succession question, which investors had long asked, was answered under pressure rather than by design.

What does the VNG story teach founders and investors?

The lesson is that distribution and product excellence can win a national market against global platforms, but converting that win into adjacent businesses requires either patient capital or a capital-light model, and VNG has had neither in sufficient quantity. Founders should be honest about which of their businesses is a platform and which is a hobby.

For founders, Zalo is the model of how to beat a global incumbent: pick the dimension where the incumbent is structurally weak, in this case performance on poor networks and cheap phones, and go all-in on it before the incumbent notices. Facebook could have optimised Messenger for Vietnam; it did not, because Vietnam was small in its priorities, and by the time it mattered Zalo had the contact graph. The same logic applied to Viettel’s rural coverage and to FPT’s Japanese-language engineers. Local champions win by choosing the ground.

The second lesson is about the cost of the platform ambition. VNG looked at Tencent, whose WeChat became a payments, commerce and services super-app, and assumed Zalo could do the same. But Tencent had a market twenty times larger, a payments environment with no bank-led alternative, and a decade of growth capital. Vietnam’s banks moved first on QR payments, the e-commerce market was captured by regionally funded platforms, and VNG’s games profits, while substantial, were not enough to subsidise several loss-making ventures simultaneously. A super-app strategy is a capital strategy before it is a product strategy.

For investors, the practical point is that VNG’s value lies in two assets, games and Zalo, and everything else should be priced as an option. The postponed IPO, the UPCoM volatility and the 2024 leadership episode all reflect a market that has not decided whether the options are worth anything. Investors who bought the unicorn narrative in 2014 have waited a long time; those who buy the cash flow and treat the ventures as free may find the company more attractive than its headlines.

What could go wrong for VNG from here?

The main risks are regulatory pressure on messaging and data, a games catalogue that depends on foreign licensors, continued losses in payments and cloud, and the unresolved question of how a founder-built company governs itself once the founder is no longer chief executive. Any one of these could stall a company that has already lost momentum.

Regulation is the most structural. Zalo operates under Vietnam’s rules for over-the-top messaging and social networks, which have tightened steadily under the 2018 Cybersecurity Law and subsequent decrees requiring data localisation, user verification and content removal on request. Zalo’s local status has been an advantage relative to foreign platforms, which face the same obligations with less goodwill, but the same rules give the state considerable leverage over the company, and any change in political mood could translate quickly into commercial constraints.

The games business carries licensing risk. VNG does not own the intellectual property behind its largest titles; it publishes them under agreements with Krafton, Riot and other studios that can be renegotiated or withdrawn, as happened when Garena lost League of Legends in Vietnam. VNG’s own development studios have produced modest hits but nothing on the scale of its licensed franchises. A publisher without owned IP is always one contract renewal from a revenue hole.

Finally there is the question of whether the AI and cloud push can be funded. GreenNode’s GPU business competes with hyperscalers and with FPT’s Nvidia-backed AI factory, requires continuous capital investment, and serves a market whose pricing is falling. VNG has the engineering talent and a domestic customer base through Zalo, but it has entered every capital-intensive business so far with less money than its rivals. The next few years will show whether the games engine can carry one more bet.

Frequently Asked Questions