Rafiqul Alam pauses for a split second as a bright red plastic box clatters toward him. He quickly grabs a package of paper towels and puts it into one of the three plastic bags neatly tucked inside the box, all the while keeping an eye on a computer screen that tells him what to pack next and where to put it in the box. He adds a pack of bottled water before the box automatically races off to another workstation down the line.
All around Mr. Alam, thousands of similar boxes are wending their way along the conveyor belts – 31 kilometres in total – that snake through this sprawling warehouse north of London, each carrying dozens of different items, from fresh fruit and sliced meat to cereal, batteries, shampoo and just about anything else you might find in a supermarket. Every box is directed by a computer system that knows its precise location, contents and destination.
This is just the beginning of how London-based Ocado Group PLC sees the future of grocery shopping, where everything is computerized and no one bothers to drive to a store. Since it launched in 2000, Ocado has set its sights on transforming grocery shopping in Britain, where it has become the largest online-only grocer in the country, with 645,000 customers, 49,000 products on offer and £1.46-billion in annual sales ($2.61-billion). Two more warehouses – or “fulfillment centres” – are in the works, and there the conveyor belts will be replaced by 4,500 robots, which can fill a customer’s weekly shopping list in five minutes and push 200,000 orders out the door every week. Soon, humans will be almost non-existent at Ocado, as the company develops driverless delivery vans, robots that fix robots and mechanical arms deft enough to pack a bag of groceries in seconds without breaking an egg or squishing a tomato.
And it’s all coming to Canada. In January, Ocado signed a deal with Sobeys Inc., Canada’s No. 2 grocer, to bring the high-tech concept to the Greater Toronto Area and eventually across the country. It’s a bold step for Sobeys, which will rely on Ocado’s robots and software to run its website and delivery system; considering the British firm has yet to consistently turn a profit and Sobeys won’t see the fruits of the partnership for at least two years, there’s a lot riding on the decision.
That hasn’t diminished the enthusiasm of Ocado’s executives, who are looking for more partners to achieve economies of scale.
“You need to do things very, very efficiently to make it work – otherwise it’s cheaper just to have the customer do their own picking, which is what a supermarket is,” said Matt Soane, general manager of technology at Ocado.
Still, analysts are divided on whether the company’s expensive automation and technology are better than the simpler, cheaper method most grocers use to fill online orders: sending staff into stores to pick goods. Ocado has yet to prove that its model “exhibits a cost advantage in fulfillment versus store-based retailers,” said Niamh McSherry, an analyst at Deutsche Bank in London. Others point to the company’s lacklustre financial results, which saw it post a pretax loss of £500,000 last year, far less than the £7-million profit analysts expected and the £12-million profit the company reported in 2016.
But there are plenty who believe the Ocado model will win out, and that’s one reason the company has seen its share price more than double in the past year to touch £6 on the London Stock Exchange. “From a pure efficiency point of view, the distribution centre model almost always wins,” said Keith Anderson, vice-president at e-commerce researcher Profitero. A Credit Suisse analysis concluded that Ocado’s centralized system delivers twice the profit margin of traditional store-based models and said it will achieve triple the margin once the latest iteration of the company’s technology, which is part of the Sobeys deal, comes on stream.
The urgency for grocers is evident: Last summer, Seattle-based Amazon.com Inc. acquired Whole Foods Market for US$13.7-billion, signalling the e-commerce behemoth’s appetite for selling groceries online. Amazon had deep pockets to invest in robotic distribution centres. The acquisition has forced rivals such as Walmart Inc. and Loblaw Cos. Ltd. to closely examine whether they should continue doing what they are doing – generally picking and packing online orders from stores – or pour money into setting up Ocado-style distribution centres.
Sobeys has placed its bet on Ocado, investing $70-million over two years to get the first of what could be four fulfillment centres up and running. For a retailer that is still recovering from its botched takeover of Safeway Canada in 2013, it’s a risky venture to support what is still a tiny part of its overall business. (Grocery e-commerce, which represents less than 0.8 per cent of the estimated $120-billion Canadian grocery market, will rise to 2.5 per cent in four years and 5 per cent in 2025, Sobeys research predicts.)
“If you believe, which we do, that grocery e-commerce is going to grow, you need to get ahead of your competitors and you need to beat them,” said Sobeys chief executive Michael Medline. “There are no real shortcuts in life.”
The last frontier
Online grocery shopping is the last frontier of e-commerce. While just about everyone has bought something online, far fewer people have used a computer to buy their groceries. An online survey by Neilsen Co. last year found that almost 60 per cent of respondents had bought clothes or booked a trip online, but only 9 per cent in North America and Europe had ever bought groceries that way. Even in Britain, considered a world leader in online grocery shopping, sales represent just 7 per cent of the market.
And there are plenty of reasons why.
Selling groceries online is a complicated business. Retailers have to stock a huge range of products, from frozen foods to dry goods, fresh meat, bread, vegetables and fruit. There’s a host of shelf-life rules, health regulations and packaging requirements. Customers tend to be very particular, concerned about freshness, when buying food. They also worry that if something they want isn’t in stock, the online merchant will choose a poor substitute. And they tend to buy many more items online from a supermarket – as many as 50 for a weekly shopping, roughly 10 times the size of other online purchases.
All of which makes the logistics of online grocery shopping tricky. In basic terms, the concept is simple: Pick a basket of items and get it to the customer as cost-effectively as possible. So far, few companies, including giants such as British-based Tesco, have been able to make much money from their online offerings.
In Canada, Loblaw was one of the early major adopters of the click-and-collect method, in which customers place an order online but still have to fetch it themselves. It remains the most economical model for retailers because it avoids the extra expense of deliveries. By the end of last year, Loblaw offered it in 300 of its supermarkets and is now adding it to one additional store every day. “We are definitely scaling it up in 2018,” said CEO Galen G. Weston.
Other retailers have followed suit, including Walmart C anada and Metro Inc., the third-largest grocer after Loblaw and Sobeys.
Walmart Canada has the pick-up service at 79 of its stores and plans to expand it to another 47 this year. It is refining its pick-up strategies, moving customers in and out of its parking lots within five minutes 90 per cent of the time, said Daryl Porter, vice-president of online grocery. “Speed-to-car is a very important part of the experience.” To improve it, Walmart is piloting mobile alerts in which the store gets pinged by the customer seven minutes before they arrive so that staff are ready to pack the order into the car.
But home delivery still accounts for much of the appeal of online shopping.
We are definitely scaling [online orders] up in 2018.
— Loblaw CEO Galen G. Weston
Soon after Amazon closed its Whole Foods deal last year, Loblaw introduced deliveries in partnership with U.S.-based startup Instacart, which sends “personal shoppers” to stores to fill and deliver online orders. Mr. Weston said click-and-collect was gaining speed but the retailer had to respond to customer demand for deliveries.
Last year, Walmart started to offer deliveries with its own trucks and third-party firms and this year with Penguin PickUp, the e-commerce delivery arm of SmartCentres, Walmart’s real-estate partner.
Still, whether a customer picks up their online order or has it delivered, having personal shoppers roam the aisles filling orders has its limits. As stores get more crowded, they start to take up space on busy weekends. That’s certainly been the case in a midtown Toronto Loblaw store that recently rolled out the click-and-collect program. Its personal shoppers wheel multitiered trays around the store, sometimes making regular shoppers feel squeezed out. The parking lot is also more congested as customers drive in and out to fetch their purchases.
“At peak periods it can interfere with our customers in the store,” said Mr. Medline of the click-and-collect at Sobeys’ IGA stores in Quebec and Thrifty Foods in British Columbia. “At a certain point it will become highly disruptive.”
Grocery Gateway, one of the country’s oldest online food sellers, found that its 20 to 30 personal shoppers in each of five picking stores (it now has 32 supermarkets in all) were starting to clog the aisles, said Anthony Longo, CEO of Longo Brothers Fruit Market Inc., which acquired the online seller in 2004. “It made for a more crowded shopping experience.”
(Strangely, even Amazon uses Whole Foods employees to pick items from store shelves and prepare them for delivery.)
In 2011, Grocery Gateway switched to a central fulfillment centre that it recently tripled in size to 150,000 square feet. The switch improved in-stock availability by as much as 30 per cent and meant merchandise was handled one less time, improving shelf life and cutting the “shrink” (damaged goods that get tossed) in half, Mr. Longo said. Still, the centre will take five to seven years to pay off and yield a net profit for Grocery Gateway, although its cash flow from operations is now positive, he added. “We’re taking a hit on the profitability intentionally.”
By this summer, Walmart will shift to a central fulfillment centre model in Vancouver. It is teaming with online organic grocer Spud.ca, which serves Vancouver from a distribution centre there and has three others in Victoria, Calgary and Edmonton.
The store pick-up model works well, but as demand accelerates a central fulfillment centre can deal better with rising volumes, Mr. Porter said. “That doesn’t happen overnight. … The reality is you have to find ways to scale the delivery. The way we do that is by partnerships, and what we’re seeing is, as more and more customers choose the specific channel, then costs come down accordingly.”
Ocado’s roots
Ocado’s roots aren’t in grocery shopping. In fact, its three founders knew nothing about retail when they started the company.
Tim Steiner, Jonathan Faiman and Jason Gissing knew each other from working at Goldman Sachs in the 1990s. They all quit in 1999, barely into their 30s and eager to do something entrepreneurial.
They considered several options, including movie production, before settling on online grocery shopping, something Mr. Steiner had seen in the United States. He’d been watching the fortunes of Webvan, an ill-fated online grocer in California that collapsed in 2001 after just three years in operation, and he was convinced the concept could work in Britain, where the population was much denser.
He convinced the others, and they pooled £1.5-million in seed capital and launched Ocado in 2000 – the name doesn’t mean anything, but it sounded new.
Mr. Steiner, the CEO, and his partners got off to a rough start. Raising money was difficult, and finding the right technology to power their designs next to impossible. They ended up creating their own technology, and today all of Ocado’s operations are powered by systems and software designed in house – everything from the WiFi system in the fulfillment centres to the tracking devices on delivery vans.
Investors gradually came on board, notably British department store giant John Lewis, which took a 40-per-cent stake in the company and signed Ocado to run the online shopping for its Waitrose subsidiary. Other early investors included an investment fund managed by former U.S. vice-president Al Gore and Tetra Pak billionaire Jorn Rausing.
But even then Ocado struggled. The company opened the Hatfield centre in 2002, and, while sales climbed steadily, losses piled up even faster. It took 15 years to turn a profit. When the company went public in 2010 on the London Stock Exchange at 180 pence, the share price fell 11 per cent on the first day of trading and dropped as low as 54 pence within a year. Many analysts were blistering, with one saying: “Ocado begins with an o, ends with an o and is worth zero.”
Mr. Steiner’s co-founders left, but he pressed on. Ocado continued to reinvent its operations, developing ever faster machinery and smarter robots. “Every human touch point in [the new fulfillment centres] is designed to one day be replaced by a robotic solution,” Mr. Steiner recently told analysts.
To accomplish that, the company has hired 1,600 engineers and divided them between those who design and build the robots and those who write the software to operate the machines. Each robot is roughly the size of a dishwasher and receives 10 commands per second as they travel within five millimetres of each other on giant grids, finding items and fetching them for the human pickers who put them in totes. Computer programs keep track of supplies, tell drivers where to go and even make out shopping lists based on an analysis of customers’ buying habits. Once fully operational this year, the newest warehouse in East London – like the one it will build for Sobeys – will be able to process 200,000 orders each week, about 40,000 more than the Hatfield site.
The key to the company’s success isn’t just building the retail operation, but finding ways to license its technology to others. Until a few months ago, Ocado had managed to sign up just one European retailer – a company they won’t name but identified by analysts as a small regional chain.
The Amazon-Whole Foods deal gave Ocado the break it needed. Suddenly, every grocer had to respond, and getting online was critical. Within a few months of the deal, Ocado announced partnerships with British retailer Morrisons, France’s Groupe Casino, which operates Monoprix stores, and Sobeys.
“Absolutely, that was the catalyst,” Ocado chief financial officer Duncan Tatton-Brown told analysts in January. “And let’s be clear: Us announcing the Groupe Casino deal was helpful in our discussions with Sobeys, as you would imagine, because it creates a sense of urgency.”
But many analysts say Ocado has a long way to go to prove itself. They note that, even in a highly developed market such as Britain, growth has been slowing. They also point out that, despite Ocado’s investments, Tesco remains the biggest player online, with 37 per cent of the market and three times the online sales. Tesco uses personal shoppers, relying on its vast network of stores for picking and getting deliveries to customers quickly.
Bruno Monteyne, a London-based analyst at Bernstein who believes the company’s shares are overvalued, says Ocado is burning through cash to keep up with its technological demands and that the Sobeys deal could simply mean an extra £100-million in extra spending over time.
Sobeys forging ahead
When Mr. Medline joined Sobeys as CEO in January, 2017, the grocer had already done research on Ocado. But by last July, after the Whole Foods takeover had been announced, he took another look at the British e-commerce firm.
Within a month, he had called Mr. Steiner, who flew to Toronto for a meeting. “When we took a hard look at the Ocado solution, it blew us away,” recalled Mr. Medline, a former CEO of Canadian Tire Corp. “Very few times in my career have I thought, ‘You have to have something.’ Usually you don’t have to have something.”
He travelled to Britain to meet with Mr. Steiner again around Halloween, this time with Jim Dickson, chairman of parent company Empire Co. Ltd. and an engineer and lawyer by training. “It got very, very serious.”
Senior teams at both companies met on both sides of the ocean in the ensuing weeks until Michael Vels, chief financial officer at Sobeys, hammered out the final details during the holiday break.
The growing array of partnerships will lead to more online business over time, said Nick Harrison, senior partner at consultancy Oliver Wyman in London. But e-commerce players such as Amazon or Ocado will quickly reshape markets they enter, he said. “It’s the classic disrupted dilemma: Do you disrupt yourself, even though it’s very painful, or do you let someone else disrupt you?”
Mr. Harrison is a voice of authority – a retail expert who is also a regular customer of Ocado. Its service is critical for feeding his busy family of two working parents and three children, aged 10, 7 and 5.
“Home delivery saves a bunch of time,” he said. “Delivery arrives on time, it’s relatively rare you get substitutes – [and] when you do get a substitute, it’s normally something sensible.”
Ocado substitutions are rare, with almost 99-per-cent accuracy compared with 90 to 95 per cent for store-picked orders, Mr. Medline said. And damaged or spoiled goods, a particularly tough challenge with fresh foods, make up just 0.7 per cent of sales in the Ocado model, compared with an industry average of about 3 per cent, his research found. A 50-item online order will take the Sobeys-Ocado system five to 10 minutes to pick and pack, compared with 40 to 60 minutes in a store-picked model, he estimated. Meanwhile, online orders, at more than $100 each, are worth about four times traditional store purchases, his data show.
Ocado has designed a sophisticated routing system that allows vans to complete an average of 182 deliveries a week – a top industry accomplishment, said Simon Mayhew, online retail insight manager at food researcher IGD in London. Ocado’s systems pick faster – more than 42-per-cent faster, at 164 items per hour, compared with a store-picked model, at about 115 items, he said. “Ocado’s central fulfillment model is more efficient and delivers better customer satisfaction in densely populated cities and, in this scenario, is industry leading,” he said.
For all its efficiency, the Ocado e-commerce “weapon” will require Sobeys to make heavy investments that won’t produce a net profit for a few years, Mr. Medline acknowledged.
But he taps into a sports analogy to back his gamble. “You can play hockey with wooden sticks or you can play hockey with composite sticks,” he said. “They both work. But no NHLer plays with a wooden stick. They all play with a composite stick. … At Sobeys, 10 years from now, someone is going to thank me – thank the team – for doing this.”