Microsoft has made three more Copilot features generally available for its CCaaS platform: Dynamics 365 Customer Service.
These follow the first capability – auto-summarizing customer cases and conversations – which the tech giant released last month.
Microsoft’s own contact center agents have trialed these use cases, with the vendor claiming that they helped agents troubleshoot customer queries quicker and more accurately.
However, supervisors may also benefit from Copilot – as evident in the third use case below.
Jeff Comstock, Corporate Vice President of Dynamics 365 Customer Service, noted this when announcing the news on Microsoft’s blog.
In addition, Comstock reassured potential users of the solution’s safety. He wrote:
Copilot uses knowledge and web sources that your organization specifies, and your organizational and customer data are never used to train public models.
He also stressed how the CCaaS platform enables data encryption, secure storage, and data access controls after introducing the following three use cases in greater detail.
Use Case 1: Answering Employee Questions
In the pane of their workspaces, multichannel agents can utilize Copilot to ask questions and receive answers without opening various applications.
The virtual agent does so by parsing resources – including the knowledge base, web links, product manuals, etc. – that the contact center allows it to access.
Moreover, Copilot will cite where it has gathered the information so that agents can dig deeper if they choose to do so.
It also allows agents to rate responses as “helpful” or “unhelpful”, and that information passes through to managers, who can then address knowledge gaps.
When Microsoft’s agents trialed this use case, they reportedly appreciated the opportunity to quickly gain answers to common customer requests.
In addition, Comstock suggested that they could now avoid lengthy investigations, scouring through previously documented procedures, which add to handling times.
As a result, one team of agents cut their “time to close cases” by 22 percent.
Use Case 2: Drafting Email Responses
Copilot can auto-generate customer responses over email, automatically understanding customer intent and scouring knowledge sources for answers to inform the reply.
Agents may then review, edit, and send to balance efficiency and effectiveness.
To draft such replies, reps can simply press a pre-defined prompt. These could include prompts such as: “Resolve the customer’s problem”, “Empathize with feedback”, and/or “Request more info”.
Yet, they may also write out custom prompts for more complex issues.
Finally, if Copilot can drink from case notes and encrypted data stores, it may produce personalized responses, facilitating a more natural, authentic conversation.
Use Case 3: Copilot Reporting
Like any addition to the contact center tech stack, leaders must measure Copilot’s impact on customer, agent, and operational outcomes.
In doing so, they should also consider how to optimize its use.
Gartner recently emphasized the significance of doing so after finding that 45 percent of contact center agents avoid adopting new technologies.
To equip leaders with the tools to do so, Dynamics 365 Customer Service offers a “comprehensive” view of metrics and insights for Copilot.
Via these new Copilot metrics, managers can uncover:
- How often agents use Copilot when replying to customers
- The number of customer conversations that involved Copilot
- The duration of interactions where Copilot plays a role
- The percentage of cases agents resolved with Copilot’s support
There are several ways in which managers may use this data. For instance, they can compare how handling times and customer satisfaction (CSAT) rates correlate with Copilot usage. Doing so will help build a case for the broader adoption of the virtual assistant.
Moreover, they could analyze the behaviors of agents that use Copilot often and consistently score high CSAT scores. That information could inform an invaluable set of best practices.
Lastly, with this data, leaders may spot non-engaged agents, talk to them about their hesitations in adopting the tech, and help ease any concerns.
Where next for Copilot in the contact center? Other GenAI use cases across the space perhaps hold the answer. Check these out here: 20 Use Cases for Generative AI In Customer Service
