After Years of Flailing, Why Has the Fight in Fresh-Food E-Commerce Moved Back Offline?

The "stores" opening offline now will be worth more in the future as "warehouses." But this round of land-grabbing offline runs on an entirely new system.

Brick-and-mortar retail has been shrinking for years. A string of news over the past month suggests it is coming back to life.

On November 15, Hema Xiansheng (Freshippo), backed by Alibaba, entered the Fujian market with its first store in Fuzhou.

On November 25, Super Species, a new format from Yonghui Superstores, opened its first Shanghai store. Yonghui is a Fuzhou company.

That same month JD.com, which owns 10% of Yonghui, said it would open a 7Fresh fresh-food supermarket in Beijing.

On November 20, Fresh Legend, a community supermarket chain with just 30 stores in Hefei, raised a 200 million yuan Series A round at a 1 billion yuan valuation — enough to get the industry's attention.

The same day, Sun Art Retail Group, China's largest brick-and-mortar retailer, sold a 36.16% stake to Alibaba, making Alibaba its second-largest shareholder.

On December 1, Meituan said it was creating a Greater Retail business group to oversee fresh-food retail and related businesses. In July this year, Meituan's Zhangyu Fresh opened its first store in Beijing's Wangjing area.

By incomplete count, brick-and-mortar retail saw more than 100 financings and acquisitions in 2016 and 2017, worth more than 20 billion yuan — and that is a conservative figure. The most willing to spend is Alibaba, which likes to talk about "New Retail." Look back a little further and it has spent 73.5 billion yuan on acquisitions in four years, which makes it China's largest offline retailer.

Encouraged by all this, the O2O boom that pushed everything online a few years ago has swung back offline, and many brick-and-mortar retailers now want to be invested in. A joke going around the industry: a business that lives or dies store by store has developed a restless eye for investors — the so-called "B2VC model."

Take Quanshi, the convenience-store chain with the most outlets in Beijing, which we reported on recently. It is expanding fast in the hope of finding a buyer in the capital markets.

On October 29, Good Neighbor (Hao Lin Ju), the convenience-store brand with the second-most outlets in Beijing, was bought outright for US$84 million (about 560 million yuan) by a group that includes Xianshenghuo, Greentown Service and Yiguo Fresh. They plan to turn it into fresh-food convenience stores that can also handle e-commerce orders.

You have probably spotted the pattern by now: behind all this retail news sit big capital and the internet companies, and the direction is unmistakable. Everyone wants to sell fresh food.

"Groceries are the most tempting and the most dangerous category," as Nadia Shouraboura, a former Amazon executive, put it. "Their volume and frequency pull in investors and retailers, then sink them in a swamp of low margins."

But fresh food is treated as the last "blue ocean." According to figures from the China E-Commerce Research Center, gross transaction value in fresh-food e-commerce has grown every year and should reach 165 billion yuan in 2017, up 80.5%. The fresh-food market as a whole is worth trillions: the National Bureau of Statistics put it at 1.36 trillion yuan in 2016, which works out to an e-commerce penetration rate of less than 7%.

Ten years have now passed since the first wave of fresh-food e-commerce companies appeared, and still no national leader has emerged. And the fight over this business looks like it is moving back offline.

First: buying customers online has become expensive

This is not a new problem. Apparel, the category with the deepest e-commerce penetration, saw it at least three years ago, which is why so many Taobao-native brands started opening physical stores. Today the slowdown in online traffic growth is worse across the whole internet industry — it is also why Alibaba has started talking about "time spent."

As the chart below shows, at China's two biggest e-commerce platforms, Alibaba and JD.com, advertising and marketing spending per newly added user has risen in a straight line over the past five years, while the profit each user contributes has not grown much at all.

So retailers went looking for new places to sell, and even set about creating them. Fresh food looks like the piece hardest for the internet to take, and it is also the most frequent purchase in daily life. Yonghui Superstores, which was never badly hurt because fresh food makes up so much of its business, and the newly arrived Hema Xiansheng both showed the industry that going back offline could work.

So how did the earliest vertical fresh-food e-commerce model get pushed out?

Yiguo Fresh, one of China's earliest vertical fresh-food e-commerce companies, marked its 12th anniversary in late November by tying up with FamilyMart convenience stores — that is, by moving offline.

Chen Jiajie, Yiguo's vice president of marketing and public relations, said at a press conference that the company expects sales of 10 billion yuan in 2017, the largest among its peers. But two B2B businesses — Anxianda in logistics and Yunxiang in supply chain — already account for 5 billion yuan of that, half the company's revenue. Half its customers are no longer individual shoppers.

Benlai, which made its name on Chu oranges, and FruitDay, which turned imported cherries into a hit, are moving the same way. These companies opened up the fresh-food e-commerce market by selling not cheap pork and vegetables but expensive imported produce and seafood. Investors and the industry now call them "traditional fresh-food e-commerce."

When Meiwei77, which had taken strategic investment from Amazon, collapsed in 2016, it was read as the moment vertical fresh-food e-commerce was pushed out. The China E-Commerce Research Center counts 14 fresh-food e-commerce companies that died in 2016 and 2017, most of them still in their early days.

The 2016-2017 fresh-food e-commerce "death list"

Founded: 2010.1

Jiduo Fresh (a general e-commerce platform built on fresh and cooked food; later pivoted to a medical website)

2011.5

Caiguanjia (a food sales and delivery platform)

2012.1

Xianpinhui (a fresh-food O2O platform)

2013.5

Meiwei77 (fresh-food e-commerce; took strategic investment from Amazon)

2013.8

Huayang Shenghuo (a community and e-commerce site built around flowers)

2013.8

Zhengyuan Shipai Guoshubang (a fruit-and-vegetable delivery service that handled procurement for restaurants in Beijing)

2013.12

Houchu.com (a mobile mall for farm and sideline products)

2014.3

Qingnian Caijun (an e-commerce company selling washed, pre-cut vegetables)

2014.5

Tetu.com (a nationwide e-commerce platform for local specialty foods)

2014.6

Guoshibang (a mobile e-commerce company selling quality fresh fruit and imported food)

2014.8

Caigou Xiongdi (a B2B fresh-food procurement platform)

2014.8

Qiangxiangou (a vertical fresh-food e-commerce brand)

2015.6

Yizhuo (a fresh-food e-commerce platform serving Shanghai and the Jiangsu-Zhejiang region)

2015.8

Benlai Bianli (fresh food and fruit delivered within the hour)

This first wave of traditional fresh-food e-commerce was, like e-commerce in every other category, really in the traffic business, and it faced the same rising cost of acquiring customers. These companies were happy to publicize falling logistics and cold-chain costs; they rarely disclosed what a new customer cost them. But a remark by Yiguo Fresh CEO Jin Guanglei — "we were still making money before 2013" — says something.

Vertical fresh-food e-commerce also ran almost entirely on "hit products," which can bring in a flood of traffic and orders in a short time. The trouble is that average order value swings wildly, and the low-spending customers a hit attracts are unlikely to stay — unless you can keep producing hits, which for fresh food is close to impossible.

Thin margins and heavy spoilage are the industry's common ground. Yan Le, a former VP of operations at MissFresh, spent more than two years in fresh-food e-commerce before leaving recently for a startup called Huoqiu Maishou. He told QDaily that this follows largely from the "central warehouse" model that traditional fresh-food e-commerce used. "Goods sit out of cold storage too long in a central warehouse, which pushes up logistics costs and causes serious spoilage. In cold-chain logistics, any order that needs ice packs costs about 50 yuan just for that. Keeping things fresh costs another 30 yuan or so."

Logistics that expensive need a high average order value to carry them, which is why the minimum order at vertical fresh-food e-commerce companies was 68 yuan at the lowest and 100 yuan at some, with most offering only next-day delivery. Customers could not shop often on platforms like that, so few of them stayed.

Even among the survivors, then, "basically nobody in the whole industry is making money."

Chen Junqi, who joined Yiguo Fresh more than two years ago as director of innovation, spent most of our interview on the company's turn into a "fresh-food enabling platform." He used a lot of technical vocabulary to explain what Yiguo will do next, but it comes down to B2B. Benlai is likewise focused on business customers now, offering online wholesale and delivery.

Yan Le called the cold-chain operation at Yiguo's Anxianda "one of the best in the industry." Without enough orders to spread the cost, the advantage means nothing.

The "stores" opening offline now will be worth more in the future as "warehouses"

After all those years of flailing, in the words of Zhang Guohong, assistant to the president of Hema Xiansheng, speaking to QDaily: "The part between buying goods and selling goods — for new entrants and old players alike — everyone is even now. This is the moment when it really comes down to your ability to sell." Zhang was vice president of Lianhua and president of Hualian before this. "I joined three years ago, when I felt this — Hema — might be the direction of the future."

By figures Hema released earlier, its first store, in Shanghai's Jinqiao area, currently splits sales 7:3 between online and offline. Zhang said most stores run about half and half, up from the 35%-40% online share last December. "But my judgment is that online can eventually get to 90%."

If it gets there, only a tenth of sales will come from people shopping in the store. Stores would then mostly do marketing work, providing the experience, and their greater value would lie in serving as "warehouses."

Li Guo, a vice president of Yonghui Superstores, has said much the same: that all its physical stores, large and small — Super Species and Yonghui Life included — will serve as "forward warehouses" for home delivery.

The term got popular thanks to the second wave of fresh-food e-commerce startups. Put simply, it solves what plagued the first generation: high order minimums and long delivery times. The clearest example is MissFresh, mentioned above. Founded in 2014, it raised a US$100 million Series C round in January 2017 from Tencent, Far East Horizon and Huachuang KTB.

In the summer of 2015, MissFresh built its first forward warehouse on Huguang Middle Street in Beijing's Wangjing: a cold-storage room of just over 100 square meters holding up to about 800 products, kept in different temperature zones according to what each one needs.

Beijing now has more than 300 of them, each serving about 20,000 people nearby. Orders peak around dinnertime every day, largely because MissFresh delivers within two hours on a 39 yuan minimum. Yan Le went as far as to call the forward warehouse "an advanced mode of production": it raises delivery efficiency, which in turn speeds up inventory turnover. And because the delivery radius is small, there is no need for foam boxes or even ice packs, so logistics costs fall sharply.

MissFresh now delivers 90% of orders within an hour, and plans to cut that to 30 minutes.

"We are one of the few companies that sells a lot of pork, and one of the few e-commerce companies selling scallions, ginger and garlic online," Xu Zheng, MissFresh's founder and CEO, said in a recent speech. "At traditional fresh-food e-commerce companies, a customer orders one or two SKUs per order. At MissFresh it is about 10."

That is the other big difference from the first wave: the core product is not high-end fruit but the ingredients people actually need to cook with. Zhang Guohong told QDaily this is also the core of Hema Xiansheng. "What we have to solve is not which fruit you eat, but how to meet a young person's need for dinner. They don't like going to the wet market."

This is really a question of distance from the customer. We have used Amazon and Walmart to explain the logic before: only by being closer to shoppers do you get the chance to serve them better.

Shixing Fresh and U Zhanggui run similar forward-warehouse operations. Shixing says it is profitable in Suzhou; U Zhanggui says it took 15 months to break even in Shanghai. Yiguo Fresh began building forward warehouses in Shanghai this summer.

MissFresh has expanded to Shanghai, Guangzhou and other cities, says it is profitable in first-tier cities, and puts annual sales at more than 10 million yuan per forward warehouse.

Zhang Guohong sees it differently. "A forward warehouse is only a hundred-odd square meters, 300 to 500 SKUs at the absolute most. Second, the storage conditions aren't good — a cold room with an air conditioner in it is bound to have plenty of flaws. In the end you need enough orders out of each location to spread the rent. A store with 300 SKUs versus mine with 3,000 — which one has the higher average order value?"

The forward warehouse does demand a great deal of the operation. "Because inventory depth is shallow, the goods are basically all fast-turnover items, and they have to be restocked daily to stay fresh," Yan Le said. That is why Xu Zheng, who studied math, has had to design an extremely complicated algorithm — weather, day of the week, historical weighting, correlations between category promotions — to set the restocking level for each forward warehouse and hold down spoilage. Renting apartments inside residential compounds as warehouses also carries some regulatory risk, and renting commercial space pushes the rent up.

Hou Yi, Hema's founder, thinks opening physical stores is the safer answer. He once won over Alibaba's Daniel Zhang with the line, "I come out of traditional supermarkets in the first place — what is there to be afraid of?" Online, Hema offers much what MissFresh does — delivery within 30 minutes inside a 3-kilometer radius — but the 21 stores it has opened all double as warehouses for filling those orders, and as new gateways for traffic.

Fresh food back offline is nothing like offline used to be. This is a new system

Since Hema Xiansheng arrived in 2016, a crop of upscale fresh-food supermarkets that cook on the premises has appeared: Yonghui's Super Species, Bailian's RISO, Meituan's Zhangyu Fresh, and even Better Life has started its own, Xianshi Yanyi.

They mix dining and grocery in different proportions. Super Species and Hema Xiansheng get compared constantly, but dining accounts for as much as 80% of the first and under 20% of the second. Zhang Guohong said Hema is a supermarket business in the end, though he puts more stress on its identity as "mobile e-commerce."

"Traditional retail counts inventory by the day. Hema does it in real time." That means higher efficiency. But Zhang also said this could not have worked ten years ago. Beyond the spread of smartphones and mobile payments, fresh food has become less unpredictable as a non-standard product — which is in fact one of the things the vertical fresh-food e-commerce companies achieved by pushing standardization.

It is also why half of Hema's staff are programmers. Three years ago its headquarters had a little over 100 people; now it has more than 800, half of them programmers, whose main job is building the whole digital system from the supplier end through to the consumer. The other half come from offline retail, restaurants and logistics.

Real-time inventory brings its own problems. Say a Hema store has one box of pork left today and a customer has put it in a basket without checking out — how much stock should the app show? Hema's answer for now is to sort popular items into several categories, online and offline, and use an N-A algorithm to hold back a certain amount of stock so an order does not land on an empty shelf. In practice, the stock held back can also spoil.

With every competitor promising to cover a 3-kilometer radius and to shorten delivery as much as possible, though, whether the model is the forward warehouse or the store-warehouse hybrid, whoever has the denser network of stores or warehouses will cover more people.

So expansion, and taking market share, became the first priority

In a sense, the way today's fresh-food players stress offline and grab territory quickly is not fundamentally different from the way e-commerce companies once bought traffic.

Yonghui Superstores' first-half results this year showed 224 stores signed but not yet open. MissFresh says it will raise its Beijing forward warehouses to 1,000, and has set aside 300 million yuan to install Bianligou, its unstaffed office shelves. Asked whether opening stores was the first priority, Hema's Zhang Guohong said: "Of course. We need to take the market. Next year we'll expand much, much faster than we are now." Hema has already set up joint ventures with retailers Alibaba holds stakes in, including Sanjiang Shopping and Xinhuadu, and plans to move quickly into more cities.

Zhang Guohong did the arithmetic. Take Shanghai's 10,000 households per square kilometer: the area within a 3-kilometer radius comes to 280,000 square kilometers, and at Shanghai's 2.6 people per household that is 800,000 people. Put average spending on fresh food at 500 yuan a month, or 6,000 a year, and the area's annual fresh-food market is about 5 billion yuan. "Big-box stores like Carrefour and Walmart average 150 million yuan a year per store. We — the Jinqiao store — can do 300 million now. If our abilities allow it, why not 1 billion?"

Besides pulling in foot traffic with dining you can watch being made, Hema recently launched Day Day Fresh, which advertises that "vegetables, meat, eggs and milk are sold for one day only" and prints the day of the week on the packaging, Monday through Sunday — "in different colors, to make it more thorough."

But brick-and-mortar retail is not the internet industry. It took RT-Mart 21 years to open more than 380 stores in China. This does not happen overnight, and each new Hema store costs 10 million yuan to open, which carries no small risk.

As for the dining that first drew customers to formats like Hema and Super Species: in QDaily's earlier report, readers already complained that "several-hundred-yuan seafood and steak are prepared very crudely." We ate at Hema's largest store, the Xingbao store in Shanghai, and found that the food from partner restaurant vendors in particular was not worth the price. Zhang Guohong said: "It's hard to say everyone is happy with this part right now — dissatisfied would be more accurate." He also said: "If a store we've opened isn't selling well, maybe the property was badly chosen, or maybe we aren't yet good enough to keep a badly chosen property alive."

That is the new problem facing these companies, now at the center of a hot topic: once the novelty wears off, what will keep users coming back, offline or online?

Cover image from Pixabay