By ZHAO Xiaojuan

Seven months after being laid off from the convenience store chain Bianlifeng, WANG?Huan has finally received his severance pay. In August last year when Wang went to work as usual, he couldn’t log in. More than two-thirds of his department were similarly locked out of the system.

Not dead, only resting

Wang’s department was in charge of opening new stores, and, since then, Bianlifeng has gone through three rounds of layoffs that affected everyone from cashiers to data scientists.

Like many consumer brands founded during the post-2015 start-up euphoria, Bianlifeng has moved quickly from aggressive expansion to mass closure.

With a claimed US$1.5 billion (10.35 billion yuan) of funding, Bianlifeng went from barely 600 stores in 2019 to an estimated 2800 in 2021. But just a few months later, in March 2022, COO WANG?Zi told reporters that Bianlifeng had voluntarily “gone into hibernation.” At least 1,000 stores are no longer hibernating. They’ve been put to sleep for good.

Like other brief candles, now flickering, Bianlifeng blames the same old culprits – rising labor costs, rising rents, and the pandemic. Blaming the pandemic now longer impresses investors. Everyone faced the same pandemic.

Data will set you free

Wage bills and rents have risen, sure, but not beyond reasonable expectations. By some estimates, salaries and rent make up 60 percent of costs for convenience stores. Bianlifeng had no hesitation in offering high pay and few qualms about signing up expensive locations – street corners were most favored – which led to high penalties when the store went out of business.

But Bianlifeng has mistakes that are unique to its own. First and foremost was a blind belief in data, which was relied upon for everything from choosing new locations to forecasting sales.

The problem – or rather the excuse - was that when operations are spread across thirty cities, the sample size for each is simply not large enough. In the first few months when there are not many stores to begin with, it’s nonsensical to use data at all.

The second mistake was Sleepless Sea, Bianlifeng’s in-store cafe.

Waking up to the age of machines

Introduced in March 2021, when venture-capital-fueled third-wave coffee chains were opening hundreds of stores every week, Sleepless Sea appeared in more than 500 stores. Instead of using typical convenience store push-button coffee machines, Sleepless Sea boasts La Marzocco, the most sought-after espresso machine among China’s newly minted coffee connoisseurs.

The problem is not just that each La Marzocco costs 100,000 yuan. The problem is that this is a convenience store.

A 20-yuan hand-made cappuccino just doesn’t fit among hot dogs and instant noodles. And the Seven-Eleven next door has decent coffee for 10 yuan. Only about 100 Sleepless Sea are left. Some good news for Bianlifeng’s creditors is that the highly desirable La Marzocco retains excellent resale value.

Bianlifeng’s goal for this year is to be profitable. To achieve this, it has turned, where else, but to data. Data and robots.

Meet the new boss

The robots are not meant to completely replace human workers. While humans spend their time on complicated tasks such as dealing with other humans, the robot looks after inventory and prices balances the books, and schedules the next shift, using “data insights accumulated through years of operations.”

This, theoretically, makes training less costly. A robot-assisted store clerk needs only 5 days of training; a store manager, 45 days, instead of the industry average of two years. It sounds very little like a recipe for success, in fact, the comedy potential is perhaps the plan’s most attractive feature.

As expected by perhaps everyone except Bianlifeng’s board, introducing a posse of robots to save a failing business has not gone well. The robots have only added to the human workload.

Smiling for the robot

A store clerk in Shanghai posted on social media about her day working with a robot. Not only does she need to finish the assigned tasks in whatever time the robot thinks fit, a task is not considered complete without photos of herself doing it.

And a photo is not accepted unless it’s taken how and when the robot wants it taken. To convince the robot that she really cleaned the countertop, for example, she has to pose in front of the camera with one hand holding a kitchen cloth.

“I spend two hours a day taking photos for literally everything, throwing away the trash, putting things in order, you name it,” she wrote.

Bianlifeng told Jiemin News that robots were introduced last year to address worker shortages. The employee's complaint has its merit, they admit, but as robots collect more data, there will be less waste and Bianlifeng will be able to cut costs even further.

According to our data…

Other convenience store chains control costs either by franchising (Lawson, 4,466 stores) or licensing to big local retailers (Family 2,902 stores).

Bianlifeng owns and operates its stores, which entails a large upfront investment for each location. The company says it constantly assesses the performance and potential of each location and may reopen some closed ones in the near future.

In other words, the money spent opening them is not completely lost. To improve customer experience, the interior of stores will be renovated, according to data insights.