When data scientists — especially those involved in marketing analytics — talk about cognitive computing, they describe a technology that holds great promise but may not be quite ready for prime time.
“We’ve been looking at it for a while, but it’s still a little too black-boxy,” said Ben Yurchak, president of KnowClick, an analytics consulting firm in Bryn Mawr, Pa. “But we like it as marketers, and see a lot of promise in applying artificial intelligence to marketing challenges.”
Yurchak’s opinion is shared by many technologists and data specialists: Cognitive computing is a powerful tool but still in its early stages. However, “early stages” is too often interpreted to mean something’s still in the laboratory. In truth, cognitive solutions have already entered the marketplace, allowing more companies to manage and make sense of increasingly complex data sets.
Learn more about what cognitive can do at IBM Amplify 2017. Watch it live here March 20-22.
What We Mean by ‘Cognitive’
Because definitions vary, let’s be clear on what we mean by “cognitive computing.” It’s essentially a mashup of cognitive science and computer science that processes and analyzes large quantities of data and puts it in a form that people can use in their decision-making. Think of something like Apple’s Siri that sorts through its own data and comes to its own conclusions rather than relies on a pre-programmed set of possible answers.
While they’re doing that, cognitive systems “self-learn” by using data mining, pattern recognition and natural language processing to “mimic the way the human brain works,” writes TechTarget’s Margaret Rouse. The goal, she says, “is to create automated IT systems that are capable of solving problems without requiring human assistance.”
Such capabilities hold great potential for industries where recognizing and understanding data patterns is critical. For example, Deloitte reports that a number of enterprises are already using cognitive systems to sort through data sets in search of risk indicators that even today’s advanced analytics packages might miss. Faced with surging volumes of information and increasing demands for faster transaction speeds, the financial and healthcare industries especially are looking to cognitive computing to address everything from equities trading strategies to personalizing healthcare.
Simplify The User’s Life
In the meantime, other businesses are putting cognitive systems to work to both streamline user experiences and learn more about their customers. To end users, a cognitive computing system’s major benefit is often seen as simplicity and personalization. Tasks that once required telephone calls, drilling down through website searches or visiting a retail store can now be completed by clicking or even speaking.
For instance, the office supply giant Staples harnesses a cognitive system to power a pilot program called “Easy Button,” which allows customers to place orders by voice commands transmitted through a dedicated smartphone app, text or actual Easy buttons placed throughout the office.
It’s ironic that easy-to-use technologies require complex systems to make them work. Easy Button is no exception. Its cognitive capabilities join with Staples’ internal personalization APIs to learn more about each customer’s preferences over time.
For example, when a user orders “blue pens,” the system will come to understand that they want Bic medium ballpoints. A chat feature allows customers to get quick answers to questions about common topics such as delivery status or product availability, and soon Easy Button will be able to recommend specific products and services based on how a customer has addressed similar needs in the past.
The Easy Button pilot was rolled out to about 100 customers in New York and Austin, Texas in late 2016. Meanwhile, the product team is researching new applications for the technology, including the development of cognitive APIs that could help customers book a flight or even allow the system to communicate with other devices. That way, a printer could warn Easy Button when it’s getting low on toner, so an order for new cartridges can be made before the existing ones run dry.
Create More Powerful Environments
Successful brands are always looking for ways to improve their customer experience and build loyalty by treating users as individuals rather than anonymous shoppers. That’s why Staples developed Easy Button, and why 1-800-FLOWERS.COM used the need to combine almost a dozen separate order-management systems as an opportunity to improve its efficiency on the back end in a way that increased customer satisfaction on the front end.
1-800-FLOWERS.COM is among the world’s leading online retailers of floral arrangements, gourmet foods and gifts. Over the years, it has grown dramatically by acquiring a number of competitive or complementary businesses. While that’s proven a sound component of its development strategy, acquisitions nearly always come with behind-the-scenes operational headaches.
“When you grow through acquisition, you often adopt home-grown tools from the acquired brands,” observed Leslie Leifer, the company’s vice president of e-commerce and product development, in a statement. In its case, 1-800-FLOWERS.COM had accumulated an array of order-management systems, each running on different platforms and in different environments. When your sales volume runs to more than $1.17 billion annually, disparate technologies only get in the way. So the business decided to consolidate its order-fulfillment processes on a single, integrated cloud platform supported by cognitive tools.
This was no small undertaking. Not only did customers need a common experience whether they ordered through the web, an app, the telephone or another channel, the system needed to remember their preferences and provide order-status information from any touch point. Behind the scenes, it needed to keep track of inventory, offer flexible options for fulfillment and identify the optimal delivery time of each order based on its own unique dynamics.
The cognitive foundation of 1-800-FLOWERS.COM’s system is sophisticated enough to do all that. Once completed, it will track both customer and inventory data along with order status and the progress of each delivery. Besides enhancing the customer experience, that will boost efficiency and drive down costs.
“Our service teams do a lot of heavy lifting so that each customer has a great experience, regardless of how many systems are involved,” Leifer said. “Soon they’ll be able to provide that experience more quickly and easily.”
User, Meet Your Personal Experience
Despite his hesitance over how today’s cognitive systems operate, Yurchak sees an illustration of their promise on the website of outdoor apparel maker The North Face. Through cognitive technology, customers find products by answering a brief series of questions in natural language to describe how, where and in what conditions they expect to use the product.
So, for example, a user tells the system they want a coat for hiking along Maine’s coast in August, whether it’s for a man or a woman, and the type of fit they prefer (“relaxed,” “active,” or “not important”). After asking a few more questions about style details and colors, the site presents its recommendations. The user can then examine and purchase a product or revise their search if they like.
Yurchak particularly likes The North Face’s approach because it helps the company understand its users. “It’s not just learning about my usage, it’s learning about what I think,” he said.
As a marketer, Yurchak appreciates what today’s cognitive systems can do but anticipates the even greater value they’ll offer in the future.
“I see a lot of promise to applying AI to marketing challenges,” he said. “If you can run campaigns and directly tie them to purchases and to a customer’s lifetime value, that’s tremendously valuable. When we’re at the point where you can mix in external factors from the weather to the economy, that’s even better.”
Eager for more fabulous insights just like these? IBM Amplify 2017 is the premier conference for all things cognitive computing. Watch it live here March 20-22.
Mark Feffer is a writer and editor who focuses on topics related to technology, analytics, technology, and workforce development. His most recent work on technology has been copywriting for the website of services provider INSYS Group (www.insys.com) and stories on the use of IT in recruiting and workforce management for SHRM Online and Dice Insights. Mark is a paid contributor to THINK Marketing.