U.K.-based online grocer Ocado Group PLC carries 50,000 SKUs. It has more than 645,000 customers and processes 280,000 orders per week. Its average order value is nearly $140. That’s a lot of bread, milk and butter.
Such a vast base of sales, orders and customers means that even the smallest efficiency-oriented tweaks can have a great influence on Ocado’s bottom line, says Greg Cempla, general manager of Ocado Technology, a unit of Ocado that develops robotics, machine learning, simulation, data science, forecasting and routing systems.
“It’s the economy of scale,” Cempla says. “A tiny savings on the cost of processing an order can have a significant impact on your bottom line when you have hundreds of millions in sales. We are always looking for the biggest bang we can get for our buck.”
And that’s what the Ocado Technology team of roughly 1,000 employees has been plugging away at for more than nine years—small improvements that make a big impact.
Ocado has plenty of company among merchants investing in technology. 78% of retailers are planning to increase their spending on technology this year, according to an Internet Retailer survey of 77 online retailers conducted in October 2017. Retailers like Ocado are using technology in a variety of ways to boost their bottom lines, drive efficiencies and better appeal to consumers. They’re implementing technologies in their back-end systems to improve operations and save time and money. Others are offering customer-facing tools to make shopping easier. Some choose to buy technology talent by acquiring vendors or use service providers rather than develop complex systems themselves. There are many paths to using technology to improve ROI and many retailers report finding successful routes.
We are always looking for the biggest bang we can make for our buck.
For instance, Ocado’s most recent program, launched last month, is a fraud detection and prevention machine-learning algorithm built using Google’s TensorFlow, an open-source software library for building machine-learning frameworks.
Ocado collects data. Petabytes of it. It captures millions of events every minute as customers navigate its website and apps, add items to their carts, choose delivery times and check out.
The new fraud model collects data from Ocado’s order management system, payments, customer-relationship management system and e-commerce teams to predict if an order is fraudulent or not. Its algorithm also adapts over time based on the data it gathers, Cempla says.
For example, the machine-learning model consumes data from the retailer’s contact center using tools developed by its data engineering, data platform and machine-learning services teams. If an order is incorrectly flagged as fraudulent and the customer calls to complain that it was legitimate, that complaint and the details of the order are noted and used to improve fraud forecasting.
It also helps in other ways. If a customer emails or calls the call center to say she will not be home to receive her order and wants to cancel it, the machine-learning system can take that information and automatically cancel the order, making that cheese or eggs in the order available for other shoppers to buy and saving a warehouse worker time picking and packing.
Ocado tested the program against the manual fraud detection program it had been using and found it to be 15 times more accurate at correctly spotting fraud than an employee. “Analyzing orders for fraud can be very tiring for an employee,” Cempla says. “Doing it for a long time increases the likelihood of a mistake. Machine learning guarantees the same level accuracy applies to all orders.”
Like Ocado, Boxed Wholesale employs an in-house technology team. The web-only merchant of household products has been busy developing and implementing consumer-facing technology to make shopping more convenient, says Will Fong, chief technology officer for the retailer.
It recently implemented several new customer tools, including Smart Stockup, launched in August, and a group-ordering feature, which it rolled out in December.
Smart Stockup predicts when shoppers will most likely run low on particular household products and then sends them an email reminder about the items they are likely to need. Boxed uses machine learning to scan customer data and predict what shoppers are going to need and built its algorithm for the service over three months.
Shoppers who use the feature spend up to 15% more per order on average than shoppers who don’t use it, Fong says. Additionally, Boxed attributes “a few percentage points” of revenue to the feature by tracking orders from consumers who click though a Smart Stockup email reminder or place an order after tapping the Smart Stockup reminder that appears when a shopper opens the Boxed mobile app.
Not all innovation is practical. Not all innovation is useful. What’s most important is to think of what your customers would want.
Consumers like the idea of such helpful technologies. A January 2017 survey of 709 consumers from Oracle Corp. found that 58% had a positive attitude toward having their grocer suggest a shopping list for their approval based on their purchase history, as well as social and environmental data.
The Boxed group-ordering feature enables consumers to use a shareable link to build their cart with others, even if they’re not Boxed customers. The feature is designed for roommates, colleagues, families or anyone organizing a group trip, Boxed says. Shoppers can see who’s buying what items and split the bill. Fong says that the feature has proven popular with managers of small offices with around 50 employees or fewer, which is a growing customer base for Boxed. Fong says B2B sales account for about 20% of its revenue and about 50% of its new business.
Boxed, which launched in 2013 and generated more than $100 million in sales in 2017, now has nearly 600 employees. As it has grown, it’s invested heavily in technology. Late last year, it launched an augmented reality tool that lets shoppers envision how that pack of 20 paper towel rolls will fit in their pantry, as well as a Facebook chatbot appropriately named Bulky.
Other recent technology-focused initiatives include investing tens of millions of dollars on a new warehouse automation system. Four miles of automated conveyor belts, which were integrated in April 2017 at its New Jersey facility, deliver goods to warehouse workers for packing. Automation has led to a 600% fulfillment productivity gain and a 350% total productivity gain, which measures labor cost per shipment, the retailer says.
In October, Boxed went one step further, unveiling another tool designed to get orders out the door quickly—self-driving carts that navigate through its Dallas warehouse and pick products. The carts travel the picking route and human warehouse workers pick the items and punch into a computer screen that the item has been selected, and then the cart delivers the finished order to a packer. The vehicles are forecasted to increase picks per hour by 80%, Fong says. Boxed also plans to use the vehicles for replenishment and transportation of supplies.
Of course, before a retailer can implement new technology, it first has to conjure up ways technology can improve their businesses.
Boxed’s group ordering and self-driving carts stemmed from the two companywide hackathons the retailer holds each year, Fong says. “Anyone can participate,” Fong says. “It’s a space where we let our team unleash their creative juices. We aren’t just a bunch of executives in a room sitting around thinking about we are going to do next.” Some teams create a PowerPoint presentation outlining their ideas and others develop full-scale prototypes. The program enables employees who typically don’t get to work together to collaborate and brainstorm. Employees receive a small prize if their idea gets chosen to pursue further.
Ocado hosts a similar program called hackdays, where data scientists and software engineers explore new features, test new models and analyze and visualize new data.
While Ocado and Boxed are developing technologies in house, others are looking outside their offices for technical expertise. For example, department store chain Nordstrom Inc. in March acquired two e-commerce technology vendors, BevyUp and MessageYes. MessageYes is a conversational commerce software platform and BevyUp is a digital selling system that Nordstrom plans to integrate into its employee-facing app in the coming year. In 2012, Nordstrom distributed 6,000 mobile devices to its stores so employees can check customers out in the aisles and email them receipts.
Nordstrom isn’t alone: 51% of retailers planning to add a new technology in the next year intended to buy that technology from a vendor rather than develop it themselves, according to a February 2017 Internet Retailer survey of 81 online merchants.
It’s no wonder some retailers decide to buy rather than build. Building takes a lot of work. For example, Ocado spent six months building a system using machine learning to better manage emails sent to its contact center. The system takes email as it comes into the center and determines whether the email has a positive or negative sentiment and tags the message with a description of its content, such as a request for website help, a complaint about delivery, a product-related issue or a cancel-order request. Based on that data, each email is immediately prioritized on how quickly it should be read and answered.
Before implementing the system, the retailer’s customer service staff had to scan and sort each email that came in. Moreover, the previous system didn’t have a way to prioritize emails, which meant that messages were handled in the order they were received. And so an urgent message such as “I can’t place my order” might be read after 50 messages asking the retailer to add a new product line.
The new email system saves Ocado 100,000 pounds ($144,233) a year, Ocado says. And, after using the system, Ocado found that 7% of emails don’t even need to be answered. Additionally, response times for non-urgent emails decreased from 19 hours in 2016 to nine hours in 2017. For urgent emails, the average response time is now two hours. Additionally with the new system, Ocado now answers 95% of emails in 24 hours, an increase from 74% in 2016, the retailer says.
But developing such helpful, cost-saving technology can be tedious.B For example, in building its email system Ocado employees used Google’s TensorFlow, an open-source software library for building machine-learning frameworks, to train the computer system using a backlog of three years of customer service emails the retailer had saved. Ocado then looked at which tags the system applied and examined how closely the assigned tags described the email’s content and the corresponding decision made by the company’s customer service employees.
It took Ocado’s in-house team of four data scientists and two software developers six months just to get to the testing phase.
Other noteworthy internal projects include a skill for Amazon’s voice-activated Alexa software that can determine, based on a shopper’s order history, when a customer says “Alexa, add milk to my Ocado order,” she probably means add two liters of semi-skim Alpo-brand milk. Ocado, which uses both machines and humans to pack and ship out items from its warehouses, also has developed systems to optimize the routes machines take in its fulfillment centers.
Ocado’s technology initiatives have been so effective that Ocado now sells its Ocado Smart Platform technology to other online grocers including U.K.-based Morrisons, Canada’s Sobey’s Inc. and Groupe Casino in France. The retailer plans to grow its Ocado technology team to 1,200 over the next year, which would mean the unit would account for nearly 10% of the retailer’s total workforce of 13,000, Cempla says.
Such internal projects are luxuries many retailers can’t afford. Many merchants don’t have the deep pockets, manpower and technical knowhow in house as Ocado does.
A February Adobe survey of 13,000 marketing, creative and technology professionals found that organizations using artificial intelligence tools to create personalized experiences for their customers were 50% more likely to significantly exceed their business goals and 46% expected to adopt AI to support customer experience by the end of 2018. However, 40% lacked the required knowledge and resources to implement it.
And it’s not just nascent technology that retailers seek help with. For example, Hobo, a midsized manufacturer and retailer of leather handbags, wallets and accessories, recently sought a more straightforward technology—an easy-to-customize e-commerce platform—says Jane Scott, senior director of direct to consumer for the retailer, which sells online at HoboBags.com.
Hobo sells more than 500 SKUs online and its web sales have grown more than 100% over the past two years. However, Hobo needed help to manage and continue such rapid growth.The retailer employs seven people who each spend part of their time supporting e-commerce initiatives, Scott says. These employees generate creative for the site, write front-end code and work on product merchandising, including launching new products and copy. They also prioritize site updates and enhancements and manage digital marketing, customer service and analytics and planning—a lengthy list of tasks for a handful of employees.
“We are not the huge brand in the handbag space,” Scott says. “Since we are a midsized player, it is important that we make use of the competitive advantage of our size and our ability to be dynamic.”
Scott says Hobo has a loyal base of shoppers. Additionally, Hobo’s average units per transaction grew more than 5% last year.
To keep up with its swift pace of online growth, Hobo moved to e-commerce platform firm Kibo late last year. Hobo began working with Kibo in February 2017 and launched on the new platform in early October to be live in time for National Handbag Day on Oct. 10.
“We were seeking a company that would provide the technology and support required for the growth that our emerging direct-to-consumer brand is experiencing,” Scott says.
Hobo says the platform gives the retailer creative and coding control over the site so that Hobo’s staff can quickly and easily make changes.“[We have] complete control over the look and feel of our home page, our branded pages and many other pages throughout the site, in addition to the header and footer, all of which we can update any time with our own in-house front-end developer,” Scott says. For example, staff can write custom code into a blank HTML field to add a feature to the site, Scott says, adding that the platform is user friendly for entry-level employees to highly seasoned front-end developers.
The ability to quickly respond to user behavior is paying off in visitor activity. Time on Hobo’s home page has increased 5% so far this year over last year, and bounce and exit rates have each dropped by 4.5%, Scott says.
“Our users are staying on site and clicking through to learn more and shop,” Scott says. For example, shoppers are spending 24% longer in Hobo’s lookbook than they did before Hobo switched to Kibo. The majority of visitors who come to the lookbook spend at least one minute there, and the bounce rate from the lookbook has dropped by more than 16% year over year, Scott says.
With the new site, Hobo also is able to use the same catalog, marketing and merchandising tools on its just-launched B2B site as its B2C site.
“Our B2C and B2B sites can share attributes, which allow for one product upload to cover both sites, which saves us from having to create and maintain two complete lists,” she says. And with the ability to access and manipulate both sites, staff can easily copy and paste any look-and-feel changes across both sites.
Our users are staying on site and clicking through to learn more and shop.
In addition to creative control, the new site enables Hobo to easily offer complex discounts, update orders with notes from customer service, maintain accurate inventory, load new products to the site and plug in new technology—such as fraud protection or international checkout—Hobo says.
Since moving to Kibo, Hobo’s sales are up 62% in 2018, and units per transaction are up 12%, Scott says.
Not all technology projects produce such favorable results, says Boxed’s Fong. Case in point: Boxed recently tested offering shopper bundles—discounts on very specific groups of products that would fit snugly into a single package and, therefore, would ship at a very low cost to Boxed.
“The logic was sound,” Fong says. “We thought that if we could bundle products that ship well together, we could pass the discount on to the consumer.” Boxed tested it as a feature in its iOS app for a few months. But the feature was too rigid for consumers who frequently wanted to tweak their orders by adding or taking away an item or two, Fong says. “It was too inflexible.”
Indeed, not all technology projects are successful or beneficial and no two retailers have the same needs and resources when it comes approaching technology. But retailers can nearly always find solutions and improvements with the right technology—if they look hard enough.
“Not all innovation is practical. Not all innovation is useful,” Fong says. “What’s most important is to think of what your customers would want.” And, of course, what makes your business run smoother.