Integrated vs. Modular Ecommerce in Vietnam


Suppose you run a corner store in 2021, and you want to start selling goods online. There are two ways you could go about it. You could set up your own storefront — a Facebook Shop, a WeChat mini program, or your own website. For the back end, you’d cobble together several services, from payments to order management to delivery. Or, you could link up with an existing marketplace, such as Amazon, Taobao, or Shopee. These sites offer all the services you need to instantly start selling online, in exchange for a portion of your profit.

Which would you choose?

It’s not necessarily a question with a clear-cut answer. The marketplace approach gives you instant market access and a seamless customer experience. But the direct selling approach allows for much greater brand control and profit share. In the West, this dichotomy underlies the tension between Amazon’s integrated value chain and Shopify’s modular offerings. It also raises the question: will the Internet empower everyone to share in the gains, or will tech giants come to hog all the spoils?

I want to consider this question in the context of Vietnam, where small- and medium-sized enterprises (SMEs) make up 98% of all businesses. While Vietnam has the fastest-growing ecommerce economy in Southeast Asia, retail is still dominated by low-tech, family-run corner shops. The future of ecommerce depends on the successful digitization of this massive traditional sector. Marketplaces like Shopee and Lazada have been making headway into the SME opportunity. But will modular tools emerge to empower merchants to compete on their own? And more broadly, is there a place for the entrepreneurial middle class in a world dominated by big aggregators on one side, and direct-to-consumer superstars on the other?

The only way I see it happening, is through social commerce tools purpose-built to localize the level of ecommerce competition. I know it’s a mouthful. Let’s dig in.

A note on terminology: for the rest of the article I’m going to use the term “marketplace” interchangeably with “aggregator.” The terms “modular tools” and “the modular stack” will be used interchangeably with “infrastructures.” My thinking on this topic was heavily informed by Ben Thompson’s writing on Tech’s Two Philosophies and The Anti-Amazon Alliance, as well as by the oft-quoted Jim Barksdale line: “there’s only two ways I know of to make money: bundling and unbundling.”

The Tradeoff

First, let’s consider the math for selling via a marketplace vs. selling direct. To set up shop on Amazon, businesses typically pay a monthly subscription of $39.99. Every item sold comes with a commission rate, usually 15%. Most merchants then also pay for Amazon to store, pack, and ship products — a service known as Fulfillment by Amazon. On top of these fees, the site charges extra for ad placement, inventory storage, early product reviews, and dedicated account managers. Amazon doesn’t publicize the income it makes from merchant services vs. merchants’ income from selling on the site, but analysts estimate the company’s total cut to be around 40%.

As an aggregator, Amazon is most useful for its ability to match supply and demand at scale. But Amazon’s scale incentive is at odds with each seller’s profit incentive. The more sellers there are on the platform, the less bargaining power each seller has against Amazon. That’s why the company’s third-party marketplace is its highest-earning subsidiary, bringing in even more cash than AWS. Meanwhile, the majority of Amazon sellers earn lifetime profits of less than $100,000.

The math is a little more straightforward for merchants going through Shopify. Unlike Amazon, Shopify’s infrastructure is invisible to the end consumer. Its software quietly powers the storefront of millions of small businesses, allowing them to set up an online presence, connect to back-end systems, and handle online payments. Merchants typically pay third-party developers an upfront cost to build a customized app or website, but they also have the option to go at it themselves using Shopify’s toolkit. Ongoing fees are minimal, ranging from $29/month for a basic plan to $299/month for the high-volume option.

Shopify’s best interest is for its merchants to do well. As merchants level up, they also start using more infrastructure, move to a higher subscription tier, and bring in more business for Shopify’s developer and agency partners. Shopify’s modular value chain aligns the incentives of every single partner: in 2018, the company made $1B; its merchants earned $1.7B; and its community of developers pulled in $7B.

The same math (and logic) apply for the decision between selling on Taobao (owned by Alibaba) vs. selling direct via a WeChat mini-program. In China, the distinction between these two approaches is also known as public vs. private traffic. As an aggregator, Alibaba and its subsidiaries handle every single piece of the ecommerce value chain for merchants, from inventory and logistics, to marketing and financing. For every dollar that merchants generate in GMV, Alibaba takes a 96¢ cut. By contrast, WeChat miniprograms cost a few thousand dollars to build, but practically nothing to run. The only cost is a 0.6% transaction fee for payments made via WeChat Pay — something merchants would already have to bear regardless of the payment service.

If the modular stack seems too good to be true, it’s because of how much harder it is to succeed as a direct-to-consumer business. What the math for the modular approach doesn’t reveal is the cost of marketing and customer acquisition. A common CAC tactic is to pay for targeted ads on social media to direct customers to one’s Shopify website. However, as social ads become more saturated, paid media’s ability to drive conversion declines. D2C brands lacking a strong value proposition can often end up losing money from social ads.

The two paths to online selling correspond to two broad categories of online sellers. Merchants selling generic consumer products are better off going on Amazon to meet their customers. Brands with differentiated products and strong organic marketing would benefit more from selling exclusively on Shopify. The forces of centralization and decentralization keep each other in check.

The Social Commerce Opportunity

When people talk about ecommerce in Vietnam, they are usually referring to the major search-driven aggregators: Shopee, Lazada, Tiki, Sendo. And, to a lesser extent, the major consumer electronics retailers: thegioididong, FPT Shop, Bach Hoa Xanh, etc.

It’s true that marketplaces are currently driving the lion’s share in official ecommerce GMV. What is less often talked about is the emerging social commerce opportunity. If we look closely, a lot of the foundational pieces are falling into place for social commerce to be the next frontier of retail transformation in Vietnam.

For one thing, social networks are already monopolizing the mindshare of Vietnamese consumers. While this is generally true of emerging economies, penetration and usage are particularly high in Vietnam. The country ranks seventh in the world in number of Facebook users, first in the world in number of viewers of video games livestream on Facebook, and has a higher social media penetration than both the regional and the global average. Social platforms by far the most powerful distribution channel to have ever existed, online or off.

What’s more interesting is the extent to which social media are already facilitating a makeshift form of ecommerce. A survey by the market research firm Q&Me finds that 85% of Vietnamese buyers who shop online have used Facebook Messenger to contact business owners. Another report by NextTech and Visa estimates that 60% of online orders in Vietnam are placed directly with business pages on social networks. While surprising, these figures point to social’s ability to solve three main pain points associated with search-driven marketplaces:

  • Discovery: people go on Shopee and Lazada if they have something to purchase, but rarely for entertainment or discovery. Social is the perfect medium for higher-funnel activities, such as brand awareness and product consideration.
  • Tech literacy: not every mom-and-pop is capable of setting up its own website and managing all the backend integration required of a true D2C business. But the majority of shopowners are technologically literate enough to start a Facebook business page, manually take orders via Messenger, and ship them cash-on-delivery.
  • Consumer trust: ecommerce in Vietnam is a trust-deficit market, with fake reviews, fake goods, poor consumer protection, and weak return logistics. Social media makes it easier for consumers to interact with sellers prior to buying, as well as to draw on social circle recommendations. Trust can also be addressed with livestream ecommerce, which, when done right, can mimic the experience of offline shopping.

A new ecosystem is coming into view. In the future modular stack, Vietnamese merchants host their storefronts on Facebook Shops and Zalo Shops. They take payments via Momo or Zalopay, and use GoStream’s and Uiza’s technologies to livestream product demos across social platforms. Some businesses might choose to work with influencers, who they contract with via multi-channel networks such as NextOn and MeTub. Logistics & delivery are taken care of by NinjaVan, GHTK, and Grab. Along the whole value chain, an entire ecosystem of B2B infrastructures spring up to address specialized needs: Telio (procurement), Haravan (omnichannel), KiotViet (inventory management), Sapo (order management), etc.

Social platforms are only a few integrations away from becoming the competing modular stack.

Platforms and Aggregators

Let’s think back for a second to our original question: will social commerce tools empower corner stores to compete online? They can, but not always.

Perhaps infrastructures’ greatest contribution is that they help SMEs avoid the fate of customer disintermediation. The aggregator playbook is one of eventual disaggregation: once suppliers are thoroughly commoditized and reduced to competing on price, aggregators can begin wrestling customers away from merchants. When consumers buy from or, it’s not individual merchants they’re building relationships with — it’s the marketplace. In the US, half of product searches originate on Amazon’s website, not on Google.

To add to this power imbalance, aggregators routinely withhold customer data from sellers, then use this informational advantage to come out with much cheaper private-label products. Over time, merchants have no choice but to abide by the aggregator’s dictates in order to reach the end consumer. If Shopee suddenly decided to raise commission, remove shipping discounts, or exclude certain sellers from marketing campaigns — as Shopee is known to have done in the past, there’s nothing that merchants could do to protest or take their business elsewhere. Remember, merchants never owned their customer relationships to begin with. Modularization, by contrast, provides an alternative to power being concentrated in the hands of a few monolithic aggregators.

The thing is, nine times out of ten, modular tools are also platforms that connect third-party suppliers (developers, influencers, logistics providers, what have you) to end users (merchants). As platforms, they also have a massive capital and informational advantage over merchants. Infrastructures such as WeChat or Shopify may decide at any moment to limit sellers’ access to data, raise their commission fee, or even venture into consumer-facing activities. They may or may not intend to be predatory, and I’m sure most players genuinely believe in arming the rebels. But the fact of the matter is, a lot of B2B platforms are still fostering relationships of dependence with merchants, much in the same way that mobile apps are dependent on iOS and Android.

Merchants’ Dilemma

Businesses have always faced the forces of commoditization. In an analogue economy, commerce is hyperlocal and some amount of commoditization is acceptable because you’re only serving demand locally. If you are the only fabric shop in a small town, you can get away with behaving as a differentiated player even though fabric could be a commodity elsewhere. When ecommerce comes along, the level of competition becomes national. Your average family-run business is now forced to compete with the best and the biggest in the whole country.

On the surface, direct selling will give merchants greater brand control and profit share. But competing on marketing is also a game rigged for the top. Consider the practice of influencer (KOL) marketing in China. It is the most popular form of advertising in China, generating $150B in sales in 2019. The handful of top influencers — Viya, Austin Li, Anny Fan — each pulled in billions of dollars in revenue. Yet 90% of KOLs report not being able to make ends meet. There’s a long tail of influencers who have to supplement their income with a second or third job, as well as a long tail of merchants who lose money on KOL marketing.

These numbers illustrate the fact that the modular approach mostly benefits a handful of cult influencers and D2C superstars. The average business is always going to be fighting an uphill battle for relevance.

From merchants’ perspective, you either risk getting squeezed by the costs of selling on a marketplace, or getting crushed by the demands of direct-to-consumer marketing. What seems like a David and Goliath of ecommerce might just be the age-old business faceoff between commoditization and differentiation. And in a world of superstar economics, neither of these fates bodes well for the average majority.

A Time for Local Stars

In order to help SMEs digitize, we need social commerce that localizes the level of competition.

Social networks have the ability to make product decisions that mitigate winner-take-all outcomes. Why is it that Tiktok and Kuaishou, both short-form video apps, attract such different demographics of users? It’s because one app is hyper-optimized for viral content and superstar creators, while the other promotes one-to-one interactions and local creators.[1] Note in the graph below how even among the top 100 accounts, Douyin (TikTok’s name in China)’s follower distribution is still a lot more skewed towards the top 10–20 accounts.

Or why is it that you always hear about memes going viral on Facebook but never on Zalo? It’s because Zalo’s closed-circle setup more closely approximates real-life social circles and enables more human-scale interactions.[2] By contrast, Facebook’s open news feed and frictionless sharing lend themselves more to power-law effects.

Similarly, many of China’s successful social commerce apps have been able to onboard a mass of small businesses thanks to their localizing strategies: prioritizing content from accounts that a user already follows, encouraging private group chats, limiting the number of users in a group chat, rewarding product referrals within one’s own social circle, and banning incentivized sharing. Apps such as WeChat mini-programs, Kuaishou, Pinduoduo, and Xiaohongshu have been extremely popular among rural merchants going online for the first time — the long tail of small businesses that have none of the social savvy of elite D2C brands such as Perfect Diary or Three Squirrels.

The same can happen in Vietnam. The market opportunity is ripe on both the demand and the supply sides: Facebook and Zalo have both been refining their business solutions, VCs have been betting big on SMB infrastructures, and fruit farmers are starting to livestream on TikTok.

The market for search-driven commerce will still exist in the foreseeable future. We will always have aggregators because of their inherent usefulness and ability to leverage the Internet’s scale. But with the right modular tools, your local corner store can also shine in its own corner of the Internet. ▲


[1] The three main differences between Douyin/TikTok and Kuaishou: (1) While TikTok only recommends the most viral content, Kuaishou surfaces a more randomized mix of popular videos and local livestreams with fewer than five viewers. (2) Tiktok monetizes via ad impressions, Kuaishou monetizes through ecommerce and viewer donations. (3) TikTok promotes one-to-many relationships, while Kuaishou encourages one-to-one interactions by allowing nested comments and local chat rooms.

[2] Unlike Facebook, Zalo only allows news feed interactions between people who have agreed to be friends. If someone comments on your friend’s photo, you will only be able to see it if you’re also friends with the commenter. This setup creates a higher degree of reciprocity and exclusivity than on Facebook.



Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store