Li Feng Column 14: Why Should We Thank Chinese Manufacturing?
Does Xiaomi or Three Squirrels Need to Build Their Own Factories?
Column No. 14 — I want to explore Xiaomi and Three Squirrels' business models, and Chinese manufacturing with you.
Over the past couple of years, a common topic surrounding Xiaomi and Three Squirrels has been: why don't they build their own production factories? For a long time, "building your own factory" was seen as a shield, equated with "quality assurance."
Looking back at the 30-year evolution of consumer retail in China, I've discovered some interesting facts I'd like to discuss with you. Before diving in, let me pose four questions:
- In China's home appliance industry, we saw Changhong → Gome, Suning → JD.com → Xiaomi emerge in sequence. Why this order? What did the landscape look like when each emerged, and how did they rise?
- For the home appliance industry, why did a channel model like "JD.com" give way to a brand model like "Xiaomi"?
- In the nuts and snacks industry, we saw Qiaqia → Laiyifen, Bestore → Three Squirrels emerge in sequence. Why this order? What did their eras look like?
- If we draw an analogy with the home appliance industry, why didn't the nuts and snacks industry produce a pure channel model like "JD.com," jumping straight to a brand model like "Xiaomi"?
- And why should we thank China's manufacturing?
I hope this brings fresh perspectives. I welcome your thoughts at the end.
01
A company's success is always created by the times. At any given moment, supply-demand dynamics are among the macro factors that create opportunities for business model innovation.
Every industry we know goes through three sequential stages: "supply shortage," "supply-demand balance," and "oversupply." And when an industry reaches "oversupply," it typically begins gestating the next phase's "supply shortage." The scales of supply and demand are almost always imbalanced.
Within supply-demand dynamics, two situations usually hold: demand consistently leads supply; quality supply is always scarce. As a result, we can see that the greatest entrepreneurial opportunities often emerge when the gap between supply and demand is largest.
The development of consumer retail over the past 30 years can be described as follows: urbanization and income growth created new retail and consumption demand; players from the previous wave of retail and consumption weren't prepared to meet this demand, creating a gap between supply and demand; this gap could be filled by new players leveraging new channels, new traffic, and new business models; meanwhile, the supply chain capabilities that China's manufacturing sector itself developed could coordinate on the production end to help these new players satisfy this demand gap.
02
Take the home appliance industry as an example. Over the past three decades, home appliances in China have completed a full cycle from "supply shortage" → "supply-demand balance" → "oversupply" → new "supply shortage." In this process, we saw Changhong → Gome, Suning → JD.com → Xiaomi emerge sequentially.
In the 1980s, when I was still in elementary school, home appliance products were extremely scarce in China, with imported appliances even harder to get. Even domestic products required ration tickets and came with quotas. It was clearly an era of supply shortage — appliances weren't something every household could easily own. They became an important, even sole, indicator of a family's economic level and quality of life.
In the 1990s, industry began developing, and with income growth, consumer demand started rising too. However, commercial real estate hadn't yet become widespread, goods remained in short supply, and consumers lacked brand selection experience. The home appliance brands that emerged at this stage were all production enterprises that became brands.
Since products were scarce, anything you could make, you could sell; whoever could produce the most was the strongest. Haier, Midea Group, and Changhong were major appliance producers at the time, veritable viral home appliance brands of their day.
Facing the massive home appliance consumer market, the industry's capacity grew at unprecedented speed, with brands multiplying rapidly — reaching a peak in 1995, when over 400 brands were registered in the air conditioning industry alone. Intensifying competition pushed the market toward supply-demand balance, then gradually toward oversupply. We entered a buyer's market. This was evident in the brutal, raging price wars in the home appliance industry around the turn of the millennium.
At this stage, the companies that began commanding话语权 became those that could better reach consumers and offer them more value for money.
So for roughly the decade starting from 2000, the most formidable companies in China's home appliance sector were channel giants like Suning and Gome. During the "Suning-Gome rivalry" era, various home appliance products filled the stores, promotions were dazzlingly diverse, and salespeople enthusiastically explained features. Retail stores became the most important sales channel for small appliance brands — in 2008, Gome and Suning recorded revenues of 45.9 billion and 49.9 billion yuan respectively.
To compete for customers, Suning and Gome grabbed territory in core cities through chain operations, opening stores aggressively. In some prime locations, consumers could compare both stores within a 100-meter walk. Paving the way for Suning and Gome's offline strategy was the urbanization process, and the commercial housing reform that began in 1998 — after several years, both residential and commercial real estate began下沉 from first-tier cities to second-tier, third- and fourth-tier cities, and even county towns.
Housing commercialization was a major driver of appliance consumption — it's not hard to understand why, since appliances are essential for outfitting commercial housing. Real estate commercialization, in turn, provided chain brands like Gome and Suning with abundant leasable space. Put bluntly: whoever could open chain stores, commercial real estate would let them in and offer them favorable rents.
China's real estate golden age lasted 20 years, and the home appliance industry's growth momentum continued for 20 years as well. In 2008, China's home appliance sales totaled 434.5 billion yuan; ten years later in 2017, sales nearly doubled to 812 billion yuan. From 2008 to 2018, when discussing home appliance consumption growth, Suning and Gome contributed significantly, and JD.com also made outstanding contributions.
JD.com entered through 3C (home appliances, digital communications, computers), dealing in standardized product e-commerce — early on, it was essentially the online version of Gome and Suning. The online home appliance shopping model worked for two reasons: first, the rapid spread of the internet enabled supply-demand information to become digitized, creating room for more efficient, user-friendly business models like JD.com; second, after being educated by Gome and Suning, consumers knew what features appliances had and how to choose — appliances had become standardized products.
Once appliances became standardized products, how to pick them mattered less — what mattered was who offered lower prices and better service, like delivering to your home, installing for you, even accepting returns and exchanges.
This was JD.com's advantage. It moved appliance retail from offline to online, offering consumers greater convenience, more choices, better prices, and higher transaction efficiency. Clearly, JD.com kept improving for users because in the home appliance industry's supply-demand relationship, we had evolved from the ration ticket era to one where we had to court consumers.
After the home appliance industry reached oversupply, it kicked off a new supply-demand balance cycle. From iPhones being out of stock, even requiring long queues, to Xiaomi phones selling out in early days, to Huawei's certain high-end series becoming sought-after in recent years — these everyday examples all testify that the consumer electronics industry has returned to a stage of user demand upgrading, calling for product innovation, where good products are in short supply.
Xiaomi is the typical representative of this stage. It extended its reach from phones to TVs, rice cookers, air purifiers, air conditioners, washing machines, LED lights, electric vehicles, robot vacuums and other home appliance categories, capturing part of the market from traditional home appliance companies. In April 2019, Xiaomi-branded smart TVs topped the smart TV online retail rankings with approximately 18% market share. In June that same year, Xiaomi announced the establishment of its Major Appliances Division.
So here's the question. Both are brands — what's the difference between Xiaomi and the first wave of appliance brands like Changhong?
You could say Xiaomi managed to break out from the home appliance industry's crowded market because of its differentiated product strategy: emphasizing value for money, design sensibility, smart connectivity, and user experience — all coinciding with upgraded consumer demand. Indeed, today's young and mature consumers have their own requirements and standards for home appliances, with their own insights and opinions on quality and brand.
Beyond this, I want to emphasize another key distinction: Changhong was a production enterprise; Xiaomi has no factories, relying almost entirely on OEM (original equipment manufacturing) and ODM (original design manufacturing).
Why doesn't Xiaomi need to build its own factories? Because China has long been a home appliance manufacturing powerhouse, and existing supply chains can fully satisfy personalized, flexible design demands. As early as 2004, China's international market share ranked first globally for six home appliance products: air conditioners, refrigerators, rice cookers, microwave ovens, vacuum cleaners, and electric shavers.
Put bluntly, China's home appliance supply side today is both good enough and plentiful enough — Xiaomi has no need to build its own factories. By amplifying its strengths in product design, IoT, channels, and other capabilities, and collaborating closely with backend suppliers, it can secure its place in the vast home appliance market. We really should thank China's manufacturing.
03
China's nuts and snacks industry followed a similar pattern. Over the past 30 years, it went from not being able to buy snacks, to being fine with whatever was available, to wanting to pick good snacks, to knowing how to pick good snacks, then back to redefining what good snacks are — demand upgrading, with quality snacks in short supply.
In this process, China saw Qiaqia → Laiyifen, Baicaowei, Bestore → Three Squirrels and other representative brands emerge sequentially.
Why this order, and what did the landscape look like when each emerged?
In the memories of post-80s generations, snacks in the 1990s were unremarkable: biscuits, instant noodles, hawthorn-based dried fruit rolls, "fig" dried fruit that had nothing to do with figs (made from shredded radish), candied mango pits with only scattered mango flesh, with roasted seeds dominated by messy-to-crack roasted sunflower seeds... Places selling snacks were mainly supermarkets and small shops; specialty stores were rare.
In 1999, Qiaqia switched from roasting to boiling, pioneering boiled sunflower seeds with more flavorful taste. Through yellow-background-red-character kraft paper packaging and thoughtfully designed collectible cards, among other relatively novel marketing methods, it quickly established differentiated recognition in consumers' minds.
Qiaqia Sunflower Seeds is a famous snack brand under Hefei Huatai Group. Founder Chen Xianbao had successfully run a cold drink food factory before entering the roasted seeds market. So Qiaqia Sunflower Seeds was genuinely a factory brand that innovated on product.
Laiyifen and Baicaowei were founded in 1999 and 2003 respectively. That was when snacks began growing — brands started multiplying, but we were still solving problems related to having enough food and clothing; snacks hadn't yet become a刚需. A key reason Laiyifen and Baicaowei could rise from offline channels was that it coincided with the first下沉 of commercial real estate and real estate, providing retail with sufficient leasable ground-floor shops. New channels needed brands to attract traffic, and offline chain brands welcomed a development opportunity. This parallels Gome and Suning.
By 2010, Baicaowei had over a thousand offline stores. By 2011, Laiyifen had 2,447 directly operated stores. Similarly, Bestore, founded in 2006, also started from physical stores — according to public reports, it currently has 2,100 offline stores, with offline sales accounting for 60%.
Drawing comparison with the appliance industry: Qiaqia corresponds to "Changhong" in model — in an era of product scarcity and supply shortage, a production enterprise created a new product and became a brand; Laiyifen, Baicaowei, and Bestore in their early days resembled "Gome and Suning" — representatives of offline snack chain operations, offering consumers convenient selection and purchase channels.
After 2000, as computers became ubiquitous and the internet reshaped retail, e-commerce emerged as a defining retail channel — and with it came opportunities to build new brands. In the snack industry specifically, Three Squirrels, founded in 2012, capitalized on Tmall's traffic dividends, committed to product innovation, and recently listed on the ChiNext board of the Shenzhen Stock Exchange.
Interestingly, sticking with the appliance-snack comparison: although both started online, Three Squirrels didn't evolve into a pure online channel brand like JD.com. Instead, it more closely resembles Xiaomi — a category brand with internet DNA. Why?
This has everything to do with the categorical characteristics and growth trajectories of appliances versus snacks.
Appliances are daily necessities. They became a major category during the first wave of residential commercialization, and they're standardized products. In other words, they were already a massive vertical category by the time the internet arrived, with enough scale to sustain an entire online channel. And indeed, JD.com built its foundation on exactly this.
Snacks are different. First, China's snack market is fragmented, dispersed, and vast — anything but standardized. Second, snacks only gradually became a major category around 2015. That year, China's per capita GDP hit $8,000, and the nature of domestic consumption shifted: spending was no longer centered on necessities, with demand upgrading through successive waves. Specifically when it came to food, the share of staple foods began declining while snacks, health supplements, and other non-staple categories rose. Even within snacking, consumers moved beyond nuts and roasted seeds to embrace dried meats, braised delicacies, cookies, and pastries.
Over the past four years, we've watched snacks evolve into something approaching a necessity — becoming daily and entertainment-driven. Through this process, supply has tipped slightly toward surplus, and consumers have grown fatigued by homogenized products. Three Squirrels invested heavily in product innovation to meet this upgrading demand. Examples include: "Daily Nuts" with wet-dry separation packaging to ensure optimal texture; the "Giant Snack" gift package designed to satisfy both sensory pleasure and psychological gratification; and the innovative bottled nut product "Xiao Ping Guo," which created distinct consumption scenarios for nuts, among others.
The internet was also a crucial catalyst as snacks grew into a major category. Based on Alibaba's sales data, CBNData's 2018 China Online Snack Consumption Trends Report showed that online food industry sales trends and per capita online food spending maintained double-digit growth from 2014 to 2017. Snacks held the largest share of online food category sales, serving as the top pillar category in 2016 and 2017. Online snack sales grew continuously over the five-year period from 2013 to 2017.
When a category's emergence as a major category coincides with the arrival of the internet, brands born from that category take on internet-native attributes: they are simultaneously brand and channel. Three Squirrels exemplifies this, as do MINISO and Xiaomi Youpin.
So where does the snack industry go from here? In my view, since snacks have become a near-necessity category and happen to arrive at the new retail era of online-offline integration, the industry will see a new "Suning" emerge — a new type of branded specialty store.
This conclusion applies far beyond snacks. It holds for many new categories that satisfy needs beyond basic food and clothing. China is now not only the world's factory but also the world's market. When front-end demand becomes mass-market — meaning the category grows large enough — and back-end supply chains mature, any brand that can integrate manufacturing may become a new category brand, with new category brands and branded specialty stores emerging. Going forward, China will see many such category-brand specialty stores.
However, today's offline environment differs fundamentally from the era of Suning, Gome, Laiyifen, and Bestore. Back then, commercial real estate was launching for the first time across cities of all tiers, generating substantial leasable space. Offline chains enjoyed relatively low rents, and medium-to-long-term leases shielded them from the general rise in rental costs. Now, commercial real estate in first-, second-, third-, and fourth-tier cities has gone through at least one round of price increases. When retail ventures go offline today, there's no real estate dividend to capture — they need genuinely high efficiency.
**/ 04 / **
As mentioned, Three Squirrels didn't become a channel brand like JD.com, which built its business on online appliance sales. Instead, it became more like the "Xiaomi" of nuts and snacks. What exactly do they have in common?
Both Xiaomi and Three Squirrels started online, both rode a wave of traffic dividends, and both became representative brands of their respective categories. Both expanded their product lines: Xiaomi from phones to various appliances, Three Squirrels from nuts to snacks of all kinds. Both also adapted to evolving retail environments, evolving from pure internet players to omnichannel retailers.
Additionally, like Xiaomi, Three Squirrels currently owns no manufacturing facilities. If we set aside the conventional assumption that "self-built factories equal quality assurance" and think rationally, it's not hard to conclude that Three Squirrels genuinely doesn't need to build its own snack production plants.
Simply put, as the world's factory, China has the world's best, most abundant, and most comprehensive supply. China's food processing industry is formidable, serving not only domestic demand but global markets. According to statistics from the Ministry of Commerce's Distribution Industry Promotion Center, from 2012 to 2016, China's snack industry exports grew steadily, increasing 15.12% over five years. In 2016, the industry's export delivery value reached 157.467 billion yuan.
Yes, food production capacity is already fully sufficient — even in surplus. What we lack isn't redoing supply and production from scratch. What's truly scarce is connecting user demand with quality supply to create new products. This is where Three Squirrels has focused its efforts over the past two years: user insight and product innovation.
Moreover, unlike abroad where cookies alone account for roughly 30% of the snack market, China's snack consumption is extraordinarily diverse due to regional differences and varied historical-cultural roots.
French fries, pistachios, instant noodles, cakes, preserved plums, duck necks (yes, we've spawned three publicly listed duck neck companies) — the images that pop into our minds when we think "snacks" may be completely different.
Even for the same snack, the same flavor, we have different understandings. Take beef: it appears as air-dried beef, braised beef, meat jerky, and more. Or spiciness: it subdivides into fragrant-spicy, sour-spicy, numbing-spicy, sweet-spicy — and even within the same spiciness profile, there are gradations: barely spicy, mild, medium, hot, and "masochist" levels.
In other words, we can no longer give snacks a precise definition. This means there are no monopolistic single products on the demand side — no absolute winners in snack subcategories. Therefore, competing in casual snacks is no longer about single-product efficiency but about maximizing system efficiency, enabling users to one-stop shop.
From this angle, you can also understand why Three Squirrels not building its own factories is relatively more sensible. How many categories and variable products can one factory process? Building one factory and adding capacity doesn't meaningfully impact or change the overall supply landscape, nor does it confer any particular advantage. In conversations with executives from several globally renowned foreign food brands, they candidly acknowledged that more than half their production capacity doesn't come from owned facilities, and not all capacity needs to be achieved through in-house factories.
At its core, what's most lacking in consumer retail today — far more than self-built factories — is product innovation and R&D capability, and driving the back-end supply chain to upgrade together. As for food safety concerns, I believe this is a matter of process control.
I discussed this with Zhang Liaoyuan. He added that brand owners only gain negotiating and pricing power with supply chains when they reach sufficient scale. When a brand is small, factories can be arrogant: "This is what I produce, take it or leave it."
Now, with scale approaching 10 billion yuan, Three Squirrels leverages its considerable scale and user demand insights to mobilize substantial quality food manufacturing capacity across China. It works closely with production partners, using digitalization to drive upgrades and advancement in the food industry, making it better suited to present and future retail formats.
Speaking of when supply is already abundant and self-built factories are unnecessary — with supply chain control being sufficient — is there a good example? Tech giant Apple has always been the paradigm. Apple has over 700 component suppliers globally, with roughly half manufacturing in China, the world's factory. We truly have China's manufacturing to thank: it's massive, complete, and efficient.
Here's an example of Apple's supply chain control. There's a Chinese company called AAC Technologies (HK: 02018) that supplies Apple with acoustic components, operating several production lines dedicated to Apple.
To control this critical supply chain, Apple provides all the control software, computers, and ERP systems. AAC's production line managers regularly receive emails from Apple identifying issues at specific points on specific lines; Apple then grants access permissions for AAC's line managers to inspect on-site. Additionally, Apple stations roughly twenty engineers at the facility on rotating shifts. In short, in principle, everything except equipment and workers belongs to Apple's control.
However, snack products differ from hard-tech products like iPhones in that their innovations are more easily copied. When market players all start looking alike, even the original innovator needs to iterate once more.
Industries rise and fall, tides ebb and flow. Far beyond snacks, entrepreneurship in any consumer retail subcategory demands that teams perceive change, adapt to change, and iterate continuously. The reason is simple. Every two to three years, supply-demand dynamics shift in any retail sector: changing demand on the left hand, changing production supply and channels on the right. More brutally, both hands are now accelerating. Those who ultimately emerge and win are those who seize change and maximize comprehensive efficiency.
That said, we've long been the world's factory, and now we've grown into the world's second-largest market behind only the US — a rare opportunity for consumer entrepreneurship.
Key Takeaways
**1 ** Great companies are products of their era.
**2 ** When front-end demand becomes mass-market — meaning the category grows large enough — and back-end supply chains mature simultaneously, any brand that can integrate manufacturing may become a new category brand.
**3 ** China possesses manufacturing capacity that's hard to fathom: roughly 70% of global production capacity for major consumer goods sits here. What we lack isn't redoing supply and production, but connecting user demand with quality supply to create new products.
**4 ** Over the past decade, real estate prices across all city tiers have risen at least one cycle. New retail ventures going offline today need higher efficiency, because the real estate dividend is gone.
Food for Thought
Which categories are currently on the path to becoming necessity-level major categories, with potential to become the "Xiaomi" product brand of their category?
Feel free to share your thoughts in the comments below.
(Welcome to read, share, and hit "like." For reprint requests, please reply "reprint" to learn about our reprint policy and contact FreeS Fund [ID: freesfund] for authorization. Copyright belongs to FreeS Fund.)
▲ Li Xiang x Li Feng: Strip Away the Emotion — What Did China Look Like in 2019?
▲ Li Feng Column | China Is at Its Most Precious Historical Moment
▲ Li Feng Column | A Guess at the Endgame of China's New Retail
▲ Li Feng Column | The Middle Path of New Retail
▲ Li Feng Column | US Dollar and Hong Kong Dollar Rate Hikes, and the Unicorn Winter
▲ Li Feng Column | The US-China Tech Rivalry: China's Secret Weapon
▲ Li Feng Column | What Was the Secret Trick That the Only Major Country in History to Successfully Transform Managed to Grasp?
▲ Li Feng Column | A 2 Trillion Market and 10 Trillion in Output — Why Are We Bullish on Chinese Chips?
▲ Li Feng Column | The Essence of New Retail, in 10 Words
▲ Li Feng Column | Lipton 120 Years Ago, Walmart 50 Years Ago, Muji 30 Years Ago — How I Think About Consumption Upgrades and New Retail
▲ Li Feng Column | Mimeng, 8 Million Followers, and Martin Luther, 500 Years Ago — How I Think About Content Entrepreneurship
▲ Li Feng Column | From iPhone X and Smart Speakers to New Drug R&D and Environmental Monitoring — How I View Tech Innovation
▲ Li Feng Column | Bike-Sharing Enters the IoT Track — Where Does the Data War Go Next?
