A sports team works from a playbook that provides strategies to create or respond to different circumstances on the playing field, and guides each team member in how to perform their part. Similarly, contact centers have playbooks to address customer issues and opportunities. With the expansion of self-service and the introduction of AI, the scope of the playbook needs to expand.
The most common elements of the playbook in the contact center today are the script and the knowledge base, which may be built and maintained by supervisors or by special knowledge managers. While the data analyst turns data into insights, the knowledge manager gathers and curates these insights, deciding how and where to best use them, and builds assets that make the insights available to others. Common assets include:
- FAQs to enable customer self-service
- canned and suggested responses for agents
- full-on scripts that guide agents step-by-step through a call
- knowledge bases for agents to search
With the rise of big data analytics and virtual agents, the knowledge manager’s job is in for big changes. Let’s take a look at some of the changes:
- Canned and suggested responses can now be enhanced by AI to provide just-in-time suggestions to agents based on real-time analysis of the conversational context.
- Data scientists will look to the knowledge managers for raw data for their analyses, as well as stores of insights to use as training materials for bots.
- Bots may augment some existing assets, such as a searchable knowledge base, or a company may decide to replace an asset with a bot.
- The data scientists will provide a wealth of new insights, which the knowledge manager can deploy in remaining scripts, knowledge bases, and training modules.
As we move forward, the knowledge manager is becoming increasingly important. Contact centers need to ensure that the right staff are in these positions, provide the needed training, and develop a good process for capturing knowledge and incorporating it into the playbook, so as to put it to work to improve customer service and reduce costs.
Teaching the Bots to Talk – The Conversational Designers
In the world of virtual agents, the script becomes not just a conversation guide, but literally the conversation itself. What works in a script for a live agent does not necessarily work for a bot, as humans are much better than AI at deciphering ambiguity and context (though AI keeps getting better!). Conversational designers not only put words in the mouth of the bot, but imbue it with personality.
The conversation designer needs to bridge together three worlds:
- the problem domain
- how customers think and react
- how machines “think” and react
- They use their communications skills to create natural and effective conversation flows. A great conversational designer will build a bot that employs the context of a conversation and the emotion of the customer to tailor the bot’s flow and language. A great bot will employ empathy with an upset customer, and use persuasion with an uncertain customer. And it will do this all within the context of a consistent persona that fits the brand’s image.