BACKGROUND IMAGE: iSTOCK/GETTY IMAGES
Cognitive computing is fundamentally changing how we interact with information -- just consider IBM's Watson. In part because of the Jeopardy-winning contestant, cognitive computing has become a central theme in designing and presenting Web-based content.
As IBM describes them, these systems "learn and interact naturally with people to extend what either humans or machines [can] do on their own" and help solve problems. What the marketing hype doesn't capture is that a lot of hard work goes into setting up the knowledge domain in the first place.
Certainly the idea of a "thinking machine" is as old as the computing age itself. While many companies may not have the dollars to invest in a complex cognitive computing system, they are exploring affordable practices that leverage currently available technologies. For example, company websites can get smarter to more effectively engage and win customers. One of the most important aspects of an engaging website is the way Web content management (WCM) systems categorize content behind the scenes. Content should be structured with customers' needs and search habits in mind.
If you are responsible for developing and maintaining a website, there are three essential steps that can smarten up your content:
- Know your target audience.
- Define your taxonomies of relevant content categories.
- Manage your content with meaningful tags and metadata.
Define your target audience's needs
Your website should target at least one constituency, but figuring out how to reach that group is the challenging part. Once you've identified your customers, determine what information they are looking for and the ways they want to access it.
Consider the business case for your website. Perhaps you are publishing support information so that customers can find answers to their questions. Maybe you are trying to engage customers in an interactive experience so that prospects return to get information and share experiences with one another. If you are trying to sell goods and services, you need to include a catalog of offerings, and develop strategies for up- and cross-selling.
Define the steps that customers go through to solve key problems. Often it's helpful to map out these stories in a series of scenarios; identify the key decisions that audience members are going to make along the way and make sure you understand what content they are going to need at these steps. This information may be stored in a database or file system, or accessed as a service from a social network.
Provide relevant taxonomies
These stories and scenarios will come together by identifying the relevant content categories for how things fit together. You want to be able to answer a few simple questions. What are your customers thinking about? What are the meaningful categories that they are using?
"The best practice that is still emerging is to start with a product catalog that is based on your business," said Paul Wlodarczyk, vice president for industrial solutions at Earley & Associates and an expert in taxonomy development. "It's important to have clean categories and not group together dissimilar things."
Moreover, design several taxonomies that lead to one another. Once you have mastered your product data, Wlodarczyk advised that you turn to your product technical information. A third facet might be your brand characteristics or the relevant marketing categories that you want your customers to notice. "You're going to need to meet with people and talk about the categories that make sense," Wlodarczyk said. Compiled into multiple taxonomies, these categories are going to help your customers describe their stories.
Manage content with meaningful tags and metadata
Once you have invested time and effort into developing smarter taxonomies, manage them in a similarly intelligent way, with a framework for categorizing the content on your website. This is the role of a modern WCM system.
People creating content need a simple and transparent authoring and editing environment. All they need to do is pick the right terms from a list. Behind the scenes, this list adapts to the "content type." And when content is produced and distributed, these terms need to be automatically surfaced as tags and metadata, attached to the content components themselves and useful for search engine optimization, assigning meaning, and incorporating into smart applications. Dynamic tagging is an essential aspect for defining a knowledge domain that can be incorporated into a cognitive computing application (perhaps even powered by a Watson-like computer).
Remember, smartening up your content is an ongoing journey, not a destination. Have the map of where you want to go and gather the stories that you need to hear. You can get started with a few modest steps: collecting the stories, developing the maps, and when you have time and resources, continuing to invest in modern technologies for managing content through applying meaningful tags and metadata.
Customer data quality is key in customer experience
Can crowdsourcing improve customer experience?
CEM lessons from Breaking Bad