Starbucks Is Getting Into Delivery. Can the Advantages of Its "Third Place" Move Online?

Finally.

Update

At 9:30 am on August 2, Starbucks announced an exclusive strategic partnership with Alibaba Group, on a three-year contract.

Starbucks' delivery will be handled by Ele.me. The pilot starts in mid-September at 300 Starbucks stores in the main commercial districts of Shanghai and Beijing, and the plan is to cover more than 2,000 stores in 30 cities by the end of the year.

Delivery won't come only from the stores. The two companies will open the first branded delivery kitchen inside a Freshippo store — the Starbucks "Delivery Star Kitchen." Freshippo will give Starbucks a dedicated delivery channel, Starbucks-only delivery boxes, and cup lids and packaging designed for the trip. Hot drinks are said to lose no more than 10 degrees Celsius by the time they arrive, cold drinks no more than 5. On timing, Freshippo has always promised delivery in as little as 30 minutes; an order sent from a store through Ele.me also takes 30 minutes (8 to make, 22 to deliver).

This will be the first time Starbucks makes its products anywhere but its own stores. Freshippo will take part in the production process, though the coffee-making team stays Starbucks'. Freshippo customers who order through the Delivery Star Kitchen will earn Starbucks Rewards stars too.

Alibaba and Starbucks will also build "smart stores" that use algorithms to split orders and speed up delivery. Once a customer opens the "Starbucks New Retail Smart Store" section on Ele.me, Freshippo, the Taobao app, or Tmall, they pick a product and nothing else — no choosing which store to order from. An algorithm weighs distance against how busy each store is and assigns the order to a nearby Freshippo or Starbucks.

The two companies will link their membership systems, so anything bought through the digital stores also earns Starbucks Rewards stars. Starbucks has 7 million members in China.

The original article follows

Starbucks is finally doing delivery.

It had hinted at this many times. On the company's earnings call in July, CEO Kevin Johnson said: "Our China team is in discussions with a major technology company about a delivery partnership, and we expect it to be finalized by the end of the year."

It came sooner than that. At 9:30 am on August 2, Starbucks and Alibaba, which owns Ele.me, will hold separate press conferences at the St. Regis Shanghai Jing'an and the InterContinental Shanghai Jing'an.

Both companies have kept the contents secret, but the run-up gave most of it away. Howard Schultz, Starbucks' founder and its recently retired CEO, has been trading friendly public gestures with Jack Ma since late last year.

According to Wang Jingying, CEO of Starbucks China, who spoke at the event, Starbucks will pilot delivery in Beijing and Shanghai this fall and take it nationwide in 2019.

Plugging into a third-party delivery platform is a huge shift for Starbucks. At home in the US, the coffee giant has insisted on owning its channel, running mobile payment and order-ahead pickup entirely inside its own app. Today 40% of mobile payments in the US happen in the Starbucks app, more than through Apple's or Google's payment systems.

In China, Starbucks has accepted WeChat and Alipay, but on terms that kept the data in its own hands. Joining a delivery platform and competing with other brands on the same shelf is the biggest concession Starbucks has ever made. It made it for growth.

Opening stores is no longer enough to keep growth going in China

Starbucks is moving into delivery in the worst growth quarter its China market has ever had. In the second quarter of this year, same-store sales at Starbucks China fell 2%, against 7% growth a year earlier. It was the first same-store decline in nine years, and it made China the company's worst-performing market in the world.

Analysts at the investment bank Sanford C. Bernstein believe the drop will last, and blame strong local rivals such as Luckin Coffee. In the first half of the year Luckin spent 1 billion yuan subsidizing users and buying ads, opened 809 stores, and closed a $200 million funding round in July.

But the threat from Luckin may be smaller than it looks. Customers bought with subsidies don't stay bought.

Data that Aurora Mobile provided to QDaily shows that by late June only 40,000 to 50,000 people were opening the Luckin app on a workday, falling below 30,000 on holidays. That is the ceiling on its daily paying users.

To push adoption, Luckin makes users order in its app — even when they are standing in the store. Opening the app doesn't mean you buy that day, but not opening it means you certainly don't. A user base this size cannot be the main reason for Starbucks' same-store decline.

The same figure peaked at 75,000 in mid-May, when the subsidies were heaviest. The subsidies are still running; simply easing them has already cost Luckin a good share of its users.

Luckin is far from the only brand competing with Starbucks through delivery. There are the old rivals, Wagas and Costa; there are KFC, McDonald's and the convenience stores now pushing coffee; and there is Coffee Box, an "internet" coffee brand cut from the same cloth as Luckin. Coffee Box is a Meituan partner, and it is four years old, though it prefers to say three, because it spent its first year essentially reselling Starbucks orders.

The competition isn't even limited to coffee. At this year's annual shareholder meeting, Kevin Johnson boasted: "There's probably no foreign brand more deeply embedded in and connected to the rapid rise of China's middle class than Starbucks."

But not everyone in the middle class drinks coffee. Starbucks' own research shows that the average Chinese person drinks 0.4 cups of coffee a year. In the US the figure is 300.

Look only at the market for delivered drinks and milk tea may be the bigger threat. It has a wider audience than coffee, and nearly every office building with a Starbucks nearby is also ringed by Yidiandian and T Plus. Milk tea carries caffeine too, and plenty of office workers who have had one won't bother with coffee afterward.

The rise of delivery is itself bad news for Starbucks. When the people waiting at the office-building elevators at noon stop being employees and become riders in red and blue uniforms, fewer people go downstairs for lunch and pick up a Starbucks on the way.

Add all this up and opening stores no longer works as well as a growth strategy for Starbucks in China.

In the year to July 1, 2018, Starbucks opened 60% more new stores outside East China than in the year before, and new stores eating into old stores' sales — the price of fast expansion — was one reason for the same-store decline this quarter. Wang Jingying explained it as an investment in the long-term development of the Chinese market and in growing a coffee culture.

When new stores pushing into second- and third-tier cities can't bring in enough revenue in the short run, the growth delivery offers starts to look necessary.

Unofficial Starbucks resale shops, in fact, are already all over the delivery platforms.

We looked at the Starbucks resale shops around six major commercial districts in Beijing and 11 in Shanghai, on Ele.me and Meituan Waimai. Their order counts show that online demand for resold Starbucks in the two cities runs to more than 130,000 orders a month, and since the minimum order is more than one cup, the number of cups is higher still. That is with no promotion at all, and without Starbucks even letting these shops use the mermaid logo.

Now Starbucks is finally doing delivery itself.

Starbucks' "third place" probably can't move online, but its stores are still its advantage in delivery

Starbucks has made plenty of digital moves before. Mobile payment made paying in the store easier; the "Say It With Starbucks" mini program used social sharing to send traffic to the store. Both served the store. The point was always to get you into what Starbucks calls the "third place," the space that is neither home nor office, where you don't only drink coffee but wait for someone, chat, work, close a deal. Take a selfie, if you feel like it.

We have written before about why Teavana, Starbucks' tea sub-brand, never worked in China: it never built a "third place" of its own and ended up as two lines on the menu board, Shaken Peach Green Tea and Shaken Pomelo Honey Black Tea, with almost no brand experience and almost no presence. That is exactly the problem Starbucks will meet first on a delivery platform.

A delivery platform offers no physical space in which to build a fine-grained brand experience. Everything is treated alike, one entry in a list of search results, and there price is what people look at. Luckin's billion-yuan bet on subsidies rests on that belief.

It may no longer be a "third place," but store density still matters enormously for delivery. It sets how long the rider takes to reach the customer, and it sets the distance the user sees in the app. Both shape every single order.

The 3,400 stores Starbucks has packed across China will, the moment delivery goes live, become the largest and most convenient coffee delivery chain on the market.

Starbucks plans to reach 6,000 stores in mainland China by the end of fiscal 2022, covering 230 cities, 100 of them new. You will still be able to order a Starbucks whenever you want coffee, and delivery does widen the reach of each store.

The logic looks a lot like the one Alibaba applies at Freshippo. But a coffee shop is not a supermarket, where people do one big infrequent trip; Starbucks can put five or more stores in a single commercial district. Delivery adds a new dimension to how Starbucks picks sites. If it is entering 100 new cities over the next four years, coffee demand in some of them will be nothing like Shanghai's or Beijing's — and in parts of those cities, demand can be served from two kilometers away.

Starbucks' brand should work in virtual space as well. Consumers used to seeing the green mermaid on the street may pick it when ordering coffee, or search for "Starbucks" directly. A brand strong enough can, to a degree, resist how a delivery platform hands out traffic.

Selling through delivery, Starbucks can no longer count on its offline partners

Faced with the rise of specialty coffee, Starbucks chose to build flagship stores with richer experiences and turn the "third place" into a phenomenon. When the Shanghai Roastery opened, the company's own word for it was a "coffee fantasy land," a two-story hall where you can watch green beans become coffee, and see brewing methods of every kind, including a Teavana bar of its own.

Black-apron baristas came from stores all over the country to help receive the crowds, and the same happened when the Reserve flagship at Beijing's Qianmen opened in July. Their job is to talk to customers, explain what is behind a particular drink, or recommend a loaf from Princi downstairs.

The other important asset of the "third place" is Starbucks' baristas and store staff, whom the company calls "partners." Espresso at Starbucks is now made entirely by automatic machines; staff refill the beans on schedule, press a button and, at most, steam some milk. Their more important job is to grow sales in the store: taking your order while you queue and slipping in a high-margin seasonal drink, asking at the register whether you want a membership card or a chocolate muffin, writing your name on the cup so it feels personal and you come back.

Starbucks' coffee carries a gross margin of about 80%, food only about 50%. But selling food with staff you are already paying lifts store revenue. Since 2016, Starbucks China has leaned harder on lunch; food also offsets the cost of the low table turnover that comes with selling coffee alone.

Employees are a critical part of the Starbucks brand experience, and the company spends hard on them. At a "Starbucks Partner and Family Forum" in Beijing in 2017, Wang Jingying announced an insurance plan for employees' parents. In 2018, after an employee at a US store discriminated against Black customers and the company faced an online boycott, Starbucks closed every store in the United States for three hours, at an estimated cost of $16.7 million.

We noted before, when Starbucks' store count was about to pass McDonald's, that on the figures in their annual reports a Starbucks employee generates roughly four times the revenue a McDonald's employee does — which may make it the most labor-efficient chain in the restaurant business.

All of this hard-won service vanishes online, where the person who reaches the customer is the delivery rider.

Online marketing and platform power — things Starbucks never had to worry about — are now in play

In a digital channel, the sales lift that staff provide can only be replaced by advertising and coupons, and marketing has always been Starbucks' weak point. Schultz once said: "I admit Starbucks' marketing has been poor. Our brand image has been built almost entirely through our physical stores and word of mouth."

Taking a market without advertising used to be one of Starbucks' great strengths. Its stores alone saved it enormous ad spending, unlike KFC and McDonald's, which spend hundreds of millions a year. But inside an app, what matters is who appears in front of the user first.

For a platform, the order of the delivery search results is the most direct advertising revenue there is. Meituan and Ele.me spent years using capital to buy users; around 2015 both began thinking about making money, and opened up paid search ranking and feed ads. In 2017 several reports said merchants on Ele.me who hadn't bought a ranking slot went from selling a dozen-odd orders a day to single digits, because their listing had dropped past the third page.

When big brands like McDonald's and KFC first joined Ele.me and Meituan, the platforms put them in prominent positions as showpieces, proof that delivery platforms weren't all unlicensed backstreet kitchens.

Over time, they became just two more merchants on Ele.me and Meituan, competing with thousands of other food brands like everyone else.

As more merchants started paying for position, you now often have to search by name to find a well-known fast-food chain. A platform's power always exceeds a merchant's, and the savage competition for online traffic and the craft of picking offline store sites are two completely different games.

The other Chinese coffee brands, the ones built on delivery from the start, are good at buying traffic with advertising and more willing to do it. Luckin says that of the 1 billion yuan it has spent, 300 million went to Focus Media ads and the rest to "buying young people coffee." Coffee Box recently launched an H5 marketing campaign called "Pocket Café": it still sells its own products, but it invites users to design a virtual coffee shop of their own and share it on Moments, and if your shop's sales pass a certain level you get a reward. Behind it, that too is paying for subsidies.

And investors take a very different view of cash burn at a mature commercial giant like Starbucks than at an internet startup.

Starbucks' almost only marketing tool is its membership program: one star for every 50 yuan spent, and the more stars, the better the perks.

By the figures Starbucks has released, it has 6.7 million active members in China, and all the data this large system has accumulated sits in Starbucks' own hands. Even though it has accepted Alipay and WeChat payments in recent years, it uses a custom terminal that folds payment together with scanning the membership code, rather than the two-in-one scanner most merchants use.

The Starbucks app was never opened much. It posts the occasional new-drink announcement and lets you look up stores, but customers still check Dianping for locations, and they still hear about new products from the sign at the door or from a barista rather than from the app. Once Starbucks is on the delivery platforms, they will have even less reason to open it.

Whether the Starbucks membership system can be plugged into Ele.me is an important open question for its delivery push. Alibaba has done this sort of thing before: its travel booking platform Fliggy linked up with Marriott's membership system and ran a series of joint campaigns.

Update after the press conference:

The two membership systems will indeed be linked, and orders placed through Freshippo stores will earn Starbucks Rewards stars too — but how it works in practice will have to wait for the service to launch.

Starbucks' biggest challenge may still be the longer in-store lines that delivery brings

Every service a store can deliver to brand standard changes once it goes online, and that is part of why Starbucks took so long.

On August 1, the day before Starbucks and Alibaba announced their partnership, Luckin held a press conference to say it was adding light meals — sandwiches and muffins, just like Starbucks — but that it would not heat them.

Starbucks will hit the same problem. A sandwich you can heat in the store: can you heat it for delivery? Does it taste worse after sitting warm for a while? Either would chip away at the experience. Earlier this year, before May, Wang Jingying told the coffee trade publication Kamen: "Many coffee shops in China offer delivery, but our first consideration is whether the emotional connection, the product experience and the consistency with the in-store experience can be matched by a delivery service."

Two months later, on an earnings call, she stressed that even with delivery, Starbucks' "direct relationships with coffee-growing regions around the world, including Yunnan, China" and its "strong brand and profitable, sustainable operating model in China" cannot be replaced by any competitor.

But the bigger challenge is what a flood of orders does to the experience.

The US market has already set a precedent for what mobile ordering does to Starbucks.

As early as 2015, 1,200 Starbucks stores in the US turned on mobile ordering — pickup, not delivery — and payment, all inside Starbucks' own app. Strictly speaking that isn't delivery, just ordering on your phone and collecting in store without queueing at peak hours. But two years in, the pains of that shift to digital are plain.

In January 2017, Kevin Johnson said in an interview that mobile orders had surged and the pickup counters were jammed. The "invisible orders" that app ordering created in the prep area, and the long lines at the pickup counter, ended up dragging down store efficiency. In its earnings report, Starbucks said the decline in store foot traffic may have been affected by mobile ordering, because lines that long were scaring off potential customers.

At the busiest Starbucks locations in Manhattan, coffee that should have taken 2 to 5 minutes was taking in-store customers more than 10, because of the rise in mobile orders, and people started complaining about it on social media.

In the end, everything about the "third place" — the flow through the store, the staff training, the process for making coffee — is designed to handle people who walk in. That process and that staffing keep the wait reasonable for a customer who comes through the door at peak hour, without leaving too many staff idle the rest of the day.

Once delivery is added, what "peak hour" means changes completely.

In China, the traffic that comes from joining a third-party delivery platform will probably grow far more, in proportion, than what Starbucks saw from its own app in the US. That is an enormous challenge for how it runs its stores.

It makes the worry about delivery ruining the taste of the coffee, or melting a Frappuccino on the way, look almost trivial.

Facing the growth in the US, Starbucks' fix has been little more than adding staff, plus a plan to add counters dedicated to mobile orders.

Facing the challenge of growth, delivery looks like something Starbucks simply has to do. And to do it well, the stores themselves will have to change.

This partnership does suggest something different: Starbucks and Freshippo will build the first branded delivery kitchen inside a Freshippo store, the Starbucks "Delivery Star Kitchen," opening a delivery-only channel outside the "third place" of the store. For the first time, Starbucks will set up a production site outside its own shops, for delivery alone.

Starbucks will also roll out packaging and cup lids designed for delivery.

All of it takes time to prepare. Starbucks' delivery, fulfilled by Ele.me, starts as a pilot at major stores in Beijing and Shanghai in September and won't reach more stores in more cities until the end of the year. A little caution is no bad thing: what has to change amounts to far more than welcoming more "customers" in rider uniforms.

Cover image by Thirteen .J and Charles Koh on Unsplash

Graphics: Feng Xiuxia