How Do We Enable Chatbots To Succeed

In last week’s article on “What is required to make chatbots succeed?,  I highlighted several implications and recommendations as crucial to how we create a solution that addresses these points.

Key Implications:

  • Chatbots need a clearly articulated purpose; “Micro-target” based on customer behavior and intent
  • Businesses need to not only measure chatbots results, but augment with customer data from other digital marketing and CRM solutions to deliver persuasive interactions with customers
  • Chatbot solutions need to understand the natural ways in which humans converse dynamically
  • Customers know they’re engaging with a chatbot, yet still want a human-like experience, requiring companies to design chatbots in a way that they pass requests to a human agent to continue the conversation
  • Customers expect consistency, regardless of touchpoint, and the chatbot solution must provide this.

To enable chatbots to succeed, there are several foundational pieces that are required. At the core of the chatbot solution are chatbots themselves, so XCentium works with our client teams to understand requirements.  Then optimizing the chatbots by utilizing customer data to train and maintain them to pulling this all together. To do this, we leverage the latest developments in digital marketing, Artificial Intelligence (AI) and Natural Language Processing (NLP) to create chatbots. We focus on integrating Sitecore based solutions with their strategic technology partners to develop these enterprise-class chatbots.

One alternative is leveraging the Microsoft Bot Framework together with Sitecore to provide a solid chatbot solution. The Microsoft Bot Framework is a library released by Microsoft that enables you to easily create chatbots using C#/JS and host them on Azure. Microsoft even provides a tool to build such bots without coding using markdown syntax dialogue definitions. That said, I am going to focus on an enterprise-grade chatbot solution, centered around Sitecore integrated with Cognigy and Coveo. This solution, combined with the appropriate digital marketing approach, checks all the boxes on the chatbot recommendations summarized above.

The following high-level diagram, illustrates this comprehensive chatbot solution:

How do we enable chatbots to succeed_
Example of Sitecore-based Chatbot Solutions_

On the top of the diagram above are the messaging platforms such as Slack, WeChat, Facebook Messenger and the like. These are channels integrated into the conversational AI, powered by Cognigy AI, allowing for the construction of self-learning AI-driven chatbots and cognitive abilities resulting in a bidirectional customer dialogue. In this setting, we are utilizing integration and database capabilities from Sitecore to help close the loop. Cognigy AI uses information stored in Sitecore xDB via Sitecore xConnect to address the customer’s intended use of the chatbot. Information retrieved from xDB about the customer is used by the chatbot to deliver a personalized experience. During the conversation, the solution can send back information on the customer’s intent to Sitecore xDB through xConnect. This information can then be used on the website as well where we can use this information from xDB to personalize the web experience. At that point, the online behavior that the customer exhibits on the site can then be pushed back into xDB, which in turn will personalize the interactions the user has on the conversational Message Platforms.

To augment the scenario just described, Coveo’s search capabilities are an integral component of the solution above. Coveo is using artificial intelligence and machine learning to study behavior patterns to determine what content will provide the best outcome for the customer. Both Cognigy and Sitecore augment their capabilities with Coveo’s intelligent search capabilities to enhance the prediction of the customer’s intent. Coveo applies machine learning leveraging contextual browsing data and information from other systems stored in xDB. Coveo’s strategic relationships with both Sitecore and Cognigy provide critical consistency in this solution.  More details on both these relationships with Coveo can be found in these video links below:

The results of this best-in-class chatbot collaboration between Sitecore, Cognigy, and Coveo eliminates the noise that was generated early in the evolution of chatbots and addresses all the recommendations needed to silence the critics.  With the chatbot’s business objective defined, combined with constant learning, this Sitecore-based solution delivers human-like conversational experiences making customers much happier and creating solid brand experience for the business.