Lou Wener works damn hard. "Five am to 10pm, seven days a week, including national holidays," the 45-year-old shopkeeper says distractedly behind the cluttered counter of her Xiabao village general store in Zhejiang province, a couple of hours' drive west of Hangzhou city. In fact, she explains as she checks a customer's egg delivery for freshness, she and her husband don't close the shop, not even for Chinese New Year. "Of course we stay open," she says, smiling tolerantly as an elderly customer holding detergent looks on bemused at a western journalist's presence. "That's a very busy day for us."
Lou grew up here in Xiabao, a village of around 1,000 people set among paddy fields and hectares of longjing tea, its outskirts marked by roadside stalls selling locally grown melons and apples. She's run the store for 21 years and lives upstairs with her farmer husband, his father, and their 21-year-old son. "When we first opened, business was good, as people didn't often go to town to buy things," she says. "But in recent years it's got much harder. It's [e-commerce website] Taobao. People started learning how to buy online around 2014; their kids taught them. Then the smartphone arrived. They just stopped buying daily goods. We felt the pressure: it was only old people coming in to the store. If we didn't change, we would be eliminated by the internet."
Then, in July 2015, the local postman offered to turn Lou's simple store into a data-enabled, real-time-responsive, globally connected e-commerce hub of its own. With help from the postman, she plugged an electronic point-of-sale laser scanner, a till-receipt printer and a digital weighing scale into a new Asus laptop that sits between a router, a cash register and a landline towered over by western and Chinese cigarette packs. Now whenever a customer pays for a Funkid Grapefruit Juice or Hazelive Soap, their purchase is tracked instantly on a central database. It is linked to both the shopkeeper and that particular customer, whose membership card is then credited with loyalty points that can be redeemed for a purchase of blueberry juice or rice wine.
A wall-mounted 42" Sanyo TV displays the WeChat group that Lou maintains for the store: a special offer on trainers if ten villagers will commit; prices for latex pillows and organic duck eggs that the postman can deliver to the store by next morning. WeChat lets her tell customers that their order has arrived. On hot days she may deliver locally. There is a China Post logo on the shop awning above a red lantern, which sits alongside a red logo containing the web address "ule.com."
These are clues to the ambitious retail experiment that has embraced this village store. On the floor sit two half-opened boxes of yams that the postman brought today for Lou's customers from a neighbouring province. They lie next to a large box of packed tea brought to the store by a local farmer to be sold on Lou's store website and collected for delivery by the same postman on his way back into town. These are what Lou calls her "virtual SKUs", or stock-keeping units, that give her customers access to thousands more lines than can fit in her cluttered physical store: cotton shirts, denim jeans, dry beef jerky, flowerpots, adhesive tape, chopsticks, dragonfruit, socks, cooking oil, doormats… all brought reliably by next-day China Post delivery.
In one month in 2015, Lou says, her website sold 800 pairs of shoes to this 1,000-person village. And that she credits entirely to her membership of Ule - the fast-growing commerce platform created by the postal service and a Hong Kong multibillionaire, which aims to transform a million village stores into the world's biggest real-time searchable retail database.
So far today, according to Ule's mobile app, Lou's taken 40 orders, earning 1,719RMB (£200) in revenue and 116RMB (£13) in profit. Yesterday, 71 orders brought in 3,295RMB and 180RMB profit. Ule's point-of-sale device lets customers pay utility bills in the store and manage their Postal Bank accounts, further boosting Lou's income. A quarter of her turnover is now online, with a growing trade in outbound sales of local farmers' shredded bamboo, fungus and dried vegetables. There are so many products that she's had to rent the warehouse opposite to store them.
Because Lou's trading data is transparent within Ule's network, the Postal Bank has offered her a 90,000RMB revolving credit line at a preferable interest rate. The offline, non-Ule store next door, meanwhile, "isn't doing well", according to Lou. She says her revenue has doubled since she joined Ule. "I was going to close down as business was so hard," she says. "Young people weren't coming, but now with mobile sharing they know there's a promotion on milk, and we can sell 80 boxes of milk in a day. Or they tell me what they want. I search on Ule for it and it comes in the next day. It may be a bit more expensive than Taobao, but you don't have to worry about fake products. I take care of everything. It's a long, hard day but I feel fulfilled. With the mobile, we're very busy. Before it was so boring that I wanted to cry."
Samson Yeung won't relent in connecting China's village stores until he's reached saturation. "A million stores would be a very good number to dominate the market," he says as we drive along the winding hills that lead to the last remaining store in the village of Yaocun. It's 20km past the nearest small town. "China has 700,000 villages, and we are planning to have one Ule store per village, plus 20 or 30 per city. Then we'll cover all China's rural areas and the best parts of the city," he adds.
As Ule's chief operating officer, Yeung is moving fast to build the world's most ambitious real-time retail-data network. In August, when WIRED tours the villages of Zhejiang province, there are 250,000 stores on the Ule system; by late December, that will rise to 330,000. And because each store owner scans every product variation into the system, to identify everything from Coke to local cabbage, Ule now tracks more than three million individual SKUs. "Point-of-sale is just the beginning," Yeung explains.
"Being on the network makes each store a virtual Walmart: they can sell what they like, even if it's not in the shop, to turn themselves into internet businesses. Plus we're capturing every transaction that's made in the store, to help the shop-owner. We know who they're selling to, at what time of day and in what weather. We work with owners to decide where to shelve products for maximum impact."
The Yaocun store is open even longer hours than Lou Wener's: from 6am to midnight, 365 days a year. This is a village of just 150 households, where flowers and wood have brought relative wealth: a 60" TV is visible through an open door in a house in the small market square. "There used to be three stores in this village, but the other two have closed," explains shop owner Han Guo Min, 47, who lives upstairs with his wife, mother and 21-year-old son, who is also their delivery driver. After 20 years here, he joined Ule on May 20, 2015.
"It's increased the wealth of the village and given us better-quality SKUs," Han says. "Ule has meant a 25 per cent growth in revenue, with utilities payment and China Post insurance sales bringing more people into the store. The inventory is automated. Before, I had to memorise prices: if I wasn't in the store, we couldn't sell something."
On an Android phone, Yeung scans the store's daily stats. It's 4pm and Han has taken 22 orders worth 1,500RMB, resulting in 152RMB profit. His seven online orders - including rice wine and a pillow - amount to 436RMB. Store data is updated every five minutes.
That level of near-real-time data from stores across China opens doors that western consumer-insight businesses can only dream of. By recording millions of daily purchases and linking them to individual customers via loyalty cards or phone payments, Ule is building an unprecedented view of Chinese consumer purchases.
Let's say you're a beer firm wanting to optimise distribution when demand rises on an unusually hot April day. Ule knows where to send your trucks. Or imagine you're Chanel and you want to know which 44- to 48-year-old women, in villages a few hours from the nearest city, have today bought a Dior product. Ule's data can potentially identify them, perhaps allowing you to send a Chanel discount voucher to their phone.
"What would you do if you had all the retail data in the world?" Kerry Liu, founder of a Toronto analytics company called Rubikloud, is sitting across a Hangzhou conference table explaining how he's turning village-store data into power. "First, there's retail optimisation - you can change how large mass retailers connect with customers and influence them. Retailers need to build relationships with customers in the same way Netflix, Amazon Prime or Facebook treat their customer base, constantly tuning their parameters. Facebook wouldn't say, 'You clicked on an update from your cousin, so now we will show you your cousin every time you log in.'
"Second, you can influence brand and product development - we did a pilot for a big pharmacy chain - and can influence consumer spending, say, to encourage healthier foods. And third, you can shape new product launches. A razor company wanted to launch a new product without cannibalising sales, so we found the 25,000 most likely customers from the retailer's database, scraped online price data, used reinforcement learning. The result was a 42 per cent rise in product spend."
Rubikloud launched in April 2013 with a mission "to index and predict the world's retail behaviour while turning data into revenue". So far its machine-learning PhDs and data scientists have processed $250 billion (£195bn) in transactional data, which adds up to 500TB. Its first product, largely for the north American market, took point-of-sale data, inventory data, promotional data, customer loyalty data and more to help retailers predict the behaviour of individual customers at scale. And then Liu, 31, met Solina Chau of Horizons Ventures, the Hong Kong-based fund which manages the tech investments of Li Ka-shing, one of Asia's wealthiest men.
"After five minutes of our demo, Solina said she wanted to know how we sucked data out. She asked if we'd settle for being acquired by a big company - or whether we wanted to build a proper platform on top of that first framework." Horizons Ventures quickly led a seed investment round. But Chau had bigger plans for Rubikloud. Li's internet and media company, TOM Group, embarked on a huge joint venture in 2010 with state-owned China Post, which has a million workers, to digitise commerce. They named it "Ule", translated roughly as "Happy post".
Today, Ule and TOM Group each own 7.5 per cent of Rubikloud, now at 55 people. They have also both invested in a Hong Kong-based finance-tech startup called WeLab, which uses mobile and offline analytics to determine whether a shopkeeper or a customer is a good credit risk for Ule to offer a loan. According to WeLab co-founder Simon Loong, a former banker, 64 per cent of rural Chinese have no access to banks, and store owners lack the credit history to borrow. So his business evaluates them with data from credit bureau and social apps, but also from their mobile devices.
"We've processed five million members, and haven't lost a case in fraud," Loong explains. Shop owners can take unsecured cash loans at nine per cent APR and use them to buy stock from Ule; the postal bank provides the cash. And shop customers give WeLab's WeLend business access to a vast amount of mobile data for a loan decision within five minutes.
"We look for personality traits, level of responsibility, by collecting 800 data points," Loong explains. "The model of phone, your apps, how you interact with others, the structure of your social networks, how you fill in your address. Whether you use capital letters correlates with bankruptcy - we think that's education level. Even what time you apply affects credit performance: 1am to 6am applicants are more likely to be bad customers compared with 8am to 1pm. We work with telcos to measure inbound call frequency, the longest gap and variability between calls - as very talkative customers are not good borrowers. We even look at messaging, and connections between the phone numbers of poor credit users - as they influence each other." Prospective customers are also asked to take a selfie, which is matched using face recognition to the police ID system. "We aspire to provide affordable credit to 30 per cent of China by 2018," Loong says.
And so a retail-data and logistics business is also building a data-led money-lending division to oil the wheels of commerce.
"What the hell is a Toronto-based data company doing in Hangzou?" Rubikloud's Liu reflects. "You can't ignore the world's largest consumer market. A cookie company we talked to missed their forecast by $50 million in ten major cities, because they underestimated local demand for other brands, got the pricing wrong and mistargeted promotions. Today they need a Ule - competition is too high. It's very difficult to predict demand: one company lost $100 million this year because it had no visibility of demand. They need a more real-time system." That comes down to knowing who the customers are, what they are buying, and where. "Ultimately we want to sell real-time placement in the physical store," Liu says. "Nielsen, dunnhumby - they're up for disruption."
"This is innovation from China. It's China IP. Internet companies are copying us." Chen Qing, founding member and chairman of Ule, as well as China Post's general manager in the province, is explaining over a lunch that includes snake soup (like bony white fish, actually) how he's modernising a 200-year-old enterprise that stretches from a postal savings bank to insurance.
"Ule is a major weapon, a catalyst, for China Post renewal," Chen, 50, continues. "To change a culture, you need to use innovative tech and a market-driven mindset. Our parcel business has grown 450 per cent in this province because of Ule. Now I demand at least 100 per cent growth every year. I've been with China Post for 20 years. I've never failed. I won't fail this one."
Zhejiang province was chosen as the test bed for Ule because it's an established e-commerce hub: Alibaba is based in Hangzhou, as are more than a third of China's e-commerce sites, according to the Hangzhou Daily. And now, Chen says, it's ready to roll out nationally - with central- and regional-government backing. "The government is endorsing Ule, to back its rural policy, for instance, subsidising the capex of each store to upgrade computers and encouraging farmers to list produce on Ule," he explains.
"Seventy per cent of the population is rural and there are lots of gaps: how do rural people get quality goods? How do farmers sell back to cities efficiently? Then there's information asymmetry. If you harvest when everyone else is, your price can collapse. China Post is the only entity in China that has complete coverage down to the last mile. We want to use tech to solve those problems."
Plus, of course, it's very good for business to reinvent China Post as the backbone of a national retail-commerce transport network. "The main China Post business gets a significant uplift, with boosts to the financial business and the logistics business," Chen explains. "Our transactions on Ule will soon exceed 200 billion RMB. The farmers get more business, which creates more logistics volume for us and more cash on deposit for the postal bank. In 2015, the postal bank had 150 billion RMB in cash deposits. In 2016, it was 200 billion. Ule has contributed half of that growth."
In 2017, he says, the goal is to connect 500,000 rural stores. "After that, the next 500,000 will be urban. Imagine anyone in a city being able to order organic greens from farmers via Ule. We have cold storage - so we'll deliver to the neighbourhood shop in the city. And think of the benefit to the farmer." In the city today, Chen explains, ginger is selling for 6RMB per half a kilo, of which the farmer gets 1.5RMB. "We will pay the farmer 3RMB and then sell it on Ule for 4.5RMB. China Post provides the lending capital, Ule provides sales and we all share the profit."
There is the small matter of persuading postal workers to upgrade from bicycle to minivan. And also ensuring that the workers buy the vans. "We're encouraging postmen to borrow money from the postal bank to buy their vans," Chen says, enthusiastically. "China Post outsources delivery to them, even as employees, and subsidises their gas. But staff own the car. He'll take good care of his own car! They get extra income for delivering wholesale goods. No other postal service is doing this!" He grins. "Changing people is disruptive. You need to change their brain."
And if workers refuse to buy their van? "All staff are Communist Party members," he says solemnly. "We have no unions here. They know what is in the best interests of China." He smiles. "Or - I can move them to another job."
"It's how communism started. The revolution began with the farmers." Ken Yeung, brother of Samson, is explaining how Ule's particular model will solve China's "rural problem" before scaling fast to the cities. Yeung is the Hong Kong-based CEO of the TOM Group and an enthusiast of his time at Singularity University. He is walking past stacked boxes of Wahaha water, Victory Vitamin Water, El Sotillo wine, Funkid Grapefruit Juice, Red Tea and a thousand other SKUs of snacks, sauces and toiletries in a 550-square-metre former letter-sorting warehouse in Yuhang county that's about to be replaced by one five times the size. This is one of 400 China Post warehouses across China that work with Ule to stock village stores directly. Local specialities include lotus fruit, sausage and duck. Food authenticity is guaranteed.
"This morning they've processed 80 orders, which we deliver in the afternoon," says Yeung, scrolling through a PC in the order-processing room. "We analyse data patterns, and work with suppliers to get bulk discounts," he says. "The biggest retail chain in China, Sinopec, has 25,000 gas stations with convenience stores. After that there's a gap. So we're using technology to give retailers cheaper prices on cookies, shampoo, noodles ..."
The TOM Group, which owns 42 per cent of Ule, "is here to empower China Post", which owns 43.7 per cent. "We put in people with a tech background and they run ground operations," he says. "We were running eBay in China. We understand e-commerce. We're digitising retail. We get feedback from 300,000 retailers. Postmen visit 15 villages daily, so we roll out fast."
The results are demonstrable. Ule's gross merchandise value in the first half of 2016 was 28.2 billion RMB, a threefold increase on the previous year. Politicians are coming to pay their respects: Wang Yang, one of China's vice-premiers, visited JiuDu township, Sichuan, in May with Lu Jiajin, Postal Savings Bank president. "Alibaba is also trying to connect the last mile," Yeung says. "They thought they'd have 200,000 outlets after two years. They have just 17,000 after 18 months.
"But Alibaba is a transaction company. Ule is a data company."
This article was originally published by WIRED UK
