Betting on technology

January 16, 2017 02:28 PM

When John Armstrong, North America leader for IBM Interactive Experience, was planning a trip to Iceland with his son a few years ago, he wanted nothing more than to speak to one of the numerous websites he used for researching the excursion what he wanted.

“I wanted to say, ‘Look, in the next six months I’m thinking of taking a trip to Iceland with my son. Here is approximately what I want to spend. Here are the types of places I want to stay at,’” Armstrong says. “We’ve been chasing personalization forever in this space. We take a bunch of data and then we try to figure out what the shopper wants. Wouldn’t that be great to just be able to ask? To just talk? And if the system doesn’t quite get it right the first time, to say ‘Well, sort of, but let me tell you a little bit more.’” 

In short, Armstrong wanted what IBM Corp. is seeking to deliver with IBM Watson. IBM thinks there is big business in “conversational commerce”—using technology to enable a more natural back-and-forth in facilitating transactions—and it has invested heavily in developing it over the past three years with its Watson supercomputer. For example, a center IBM opened last year in Munich seeks to bring Watson to the growing internet of things, says Niklaus Waser, vice president of Watson IoT Europe and the Watson IoT Center in Munich. The center, which opened in December 2015, is part of IBM’s broader $3 billion investment to expand the reach of Watson’s machine learning technology. IBM is investing $200 million in the Munich center alone, Waser said during a recent visit arranged as part of a media tour by Germany Trade & Invest, Germany’s economic development agency.

Watson uses an algorithm that learns when it’s exposed to new data instead of being explicitly programmed so that it can engage in dialogue with people, learning over time how to improve its answers and its conversation skills. In e-retail Watson enables shoppers to describe what they want, or need help with, and can provide informed suggestions as opposed to a retailer guessing—often incorrectly—that because a shopper bought a slow cooker around Mother’s Day last year, she must be seeking culinary tools all year long.

About a dozen global e-retailers are using Watson technology, Armstrong says. Only two will go on the record—outdoor gear retailer The North Face and Inc. [See 1-800-Flowers’ CEO’s take on conversational commerce on page 23.]

On the, shoppers access Watson via an Expert Personal Shopper link on the home page. The link takes consumers to a black screen with large white lettering that says, “Where and when will you be using this jacket?”

One could type, for example, “I’m going to Maine in December.” The site then asks, “Are you looking for a men’s or women’s jacket?” Consumers then view six jackets and either click on one of them to get more details on why the jacket is well suited to the Maine climate that time of year, or drill down by answering additional questions like, “Do you expect it to rain or snow?”

The Watson-powered tool on The North Face site, called XPS or Expert Personal Shopper, was developed by digital agency Fluid Inc. The North Face says shoppers who use the tool have a 43% lower cart abandonment rate than those who don’t and consumers view twice as many products online with XPS.

Technology buzz words like machine learning and artificial intelligence are attracting real investments from merchants and e-retail service providers like IBM that want to get in now on the future of online retail technology. And many companies are choosing to buy already existing specialized technology to stay at the forefront of e-commerce.

In fact, IBM bought Fluid’s XPS division this fall. IBM had already invested in XPS in 2014 as part of a larger pool of $100 million in funding it earmarked for direct investment in technology using Watson.

IBM has been on a spending spree as of late, snapping up technology firms for its Interactive Experience division, which houses Watson. In the past year it bought Resource/Ammirati, a digital marketing and creative agency; Aperto and, two Germany-based digital agencies; and Bluewolf, a web consulting and an implementation partner for customer relationship management system Inc. None of the purchase prices were disclosed.

But IBM isn’t the only established player investing in forward-looking technology. Online marketplace and technology company eBay Inc. in October said it will buy Corrigon Ltd., a visual search technology company, for an undisclosed sum.

“As shoppers continue to shift from desktop to mobile devices, it’s really difficult to use your keyboard,” says Mrinalini Loew, eBay’s head of mobile product experience. “If you like your friend’s boots, do you want to type in the entire brand name on your phone or just snap a picture?”

EBay is buying up mobile technology companies because that’s increasingly how consumers are shopping. Consumers use mobile at some point in their shopping journey for 58% of eBay purchases, eBay says. Loew says eBay, which posted third quarter global mobile gross merchandise volume of $94 billion, is seeking to go further with mobile image recognition by, for example, showing the shopper pants or other items that would also go with that picture of boots. Loew says she also sees Corrigon technology being useful for eBay in selling auto parts. For example, a consumer might snap a picture of his vehicle and the technology would match it to the part he needs that will fit the make and model of his automobile, saving him from having to manually sort through eBay’s massive catalog of parts. In the United States, eBay hosts about 75 million auto parts and accessories and three auto parts or accessories are sold on eBay in the United States every second, eBay says.

Corrigon is eBay’s third acquisition in 2016 related to artificial intelligence, machine learning and data science. Others include Malmo, Sweden-based Expertmaker, which focuses on artificial intelligence, machine learning and big data analytics, and SalesPredict, an Israel-based company that uses analytics to predict customer buying behavior and sales conversion.

EBay is also experimenting with augmented reality and virtual reality. For example, consumers in Australia can now shop from department store Myer on eBay using virtual reality that works with eBay’s virtual reality app, which also uses machine learning. When a consumer first uses the app, she selects clusters of products. From there she keeps narrowing down what she is searching for. The next time she shops, the eBay virtual reality headset will have begun to learn the types of products she likes and will put those items in front of her immediately.

EBay CEO Devin Wenig is looking for ways to apply that technology to other goods sold on eBay.

For example, he has formed a research team called “N” that focuses on exploratory projects for the future. The team is called “N” because it focuses on new product developments. This past year for example, the N team worked on eBay ShopBot beta, eBay’s shopping chatbot for Facebook Messenger.

E-retailers are also buying in to new technology. For example, home décor retailer Wayfair Inc. recently bought the Boston-based startup Trumpit messaging app that enables shoppers to chat with Wayfair on a product detail page rather than sending an email or calling customer service.

Wayfair also is also investing in virtual reality and augmented reality in a big way, says co-founder Steve Conine. Just last month it unveiled View in Room, a feature in its mobile app which lets shoppers see furniture and décor in their homes before they buy. The new feature places two-dimensional product images in any room through the camera of a mobile device. 

“Augmented reality and virtual reality technology is closing the gap for shopping online for furniture,” Conine says. “We want people to get to the point where they think it’s easier to shop for furniture for their home while they are in their home.”

Wayfair has also created about 10,000 3-D modeled products from its catalog of 7 million items that can be placed in photos of rooms both on Wayfair’s own site or on outside sites and apps. Conine says the cost for rendering a 3-D model of a typical piece of furniture ranges between $150 and $300 compared with $1,000 to $3,000 for high-quality product photos of the same product produced in a photo studio. Between six and a couple dozen of Wayfair’ 800-person engineering team are working on 3-D modeling at any given time, Conine says.

“3-D modeling and rendering dramatically changes the cost for us,” Conine says. “And the model library is paying off.  We see an improved yield when we have improved photography.”

Making it easy to shop online is no easy feat. Luckily new technology is there to help. It isn’t cheap, but many retailers and vendors are betting their investments will pay off with increased sales.

And for some, like Wayfair, they already are.




Top Solution Providers