At the Dreamforce Service Cloud keynote, Service Cloud GM Kishan Chetan contrasted how much things have changed in the last year - with AI moving from theory to reality.
He challenged customer service leaders to be the leaders of change in their organizations, and embrace AI agents to move from "good" benefits to exponential ones. He also highlighted how AI agents can solve top customer service requests from tech support, scheduling, and general questions - but also more complex needs such as de-escalation, billing inquiries, as well as cross and upselling.
In reality, Valoir has found that many Service Cloud customers are in the early days of adopting AI, and generative AI in particular. Although it's been accelerating in recent months, most customers today are only achieving incremental benefits driving an increased productivity for individuals, not the kind of exponential returns highlighted in Dreamforce keynotes. In order to move from incremental to cumulative to exponential returns, customers will need help moving from simple automation and summarization to AI-supported transformation with Agentforce. To that end, Kishan and the Service Cloud team outlined four steps to transform experiences with Agentforce.
Agentforce Service Agent is not a chatbot, but an autonomous AI agent that can handle both simple and complex requests and understand not just text but video and audio. Throughout Dreamforce - from Marc's keynote on down - customers were encouraged to go to an area where they could build their own Service Agents (and, allegedly, thousands of customers did). It's not surprising that a lot of these agents would be service-related agents, as Valoir's research has found Service Cloud customers, on average, are the readiest for AI because of the volume, quality, and consistency of customer data they have in their customer relationship management (CRM) systems.
Customer Experience Intelligence (planned for general availability in October 2024) provides an omnichannel supervisor Wall Board that enables a supervisor to view conversations in real time with sentiment scores, with metrics organized by topics and geographies. When a supervisor identifies an issue, it can ask Service Agent to drill down on details and root causes, suggest a proactive message to send to customers, and potentially propose a discount offer.
To us, this looks like the next evolution of Service Intelligence, which brings together Data Cloud, Tableau, and Einstein Conversation Mining to help service supervisors understand what's happening in real time. It also looks a lot like the speech analytics solutions from contact center vendors like Verint that bring interaction, sentiment, and other data together in real time - highlighting the continued blurring lines of the traditional contact center and Service Cloud-based service organizations that have been traditionally segregated by channel.
Salesforce is still tiptoeing on the human versus digital agent replacement story, particularly when it comes to Service Cloud - but the key here is the intelligent transfer of customer profile and interaction data when a digital agent (or a human customer) escalates to a human. Based on this data, Service Planner gives the agent step-by-step guidance on how to resolve the customer's issue based on Unified Knowledge. The keynote demo also showed how Service Agent can bring Commerce and Service together, by recommending the agent offer the customer a complimentary item from their shopping cart.
Salesforce also announced advances in field service capabilities to enable both dispatchers and field service agents to be more proactive and productive. Agentforce for Dispatchers can help dispatchers more quickly address urgent appointments and changes.
Asset Service Prediction uses AI to predict potential upcoming asset failures and service appointments, and AI-generated prework briefs techs can listen to on the way to the job, see an asset health score, and be more prepared to service customers.
Laying out a roadmap for adoption of Agentforce for these four specific goals is a good first step toward helping customers digest Agentforce and plan for leveraging it for more than incremental improvements in their service experiences. Also key will be helping customers map out a data strategy that leverages Data Cloud and Salesforce's ecosystem partners to bring in critical data that will be needed for true data-driven service, and more investments in capabilities like My Service Journeys that help customers frame and articulate not just what they can do with Agentforce and AI but which capabilities are likely to deliver the greatest returns in their own environment.
