Marc Benioff Predicts Half of Conversations Will be With AI Agents Next Year

Logan discusses Salesforce's significant backlog of over 100 million unaddressed leads due to staffing shortages, leading to the implementation of an agentic sales system that processes over 10,000 leads weekly. Mark shares that Salesforce is the first customer of their new agentic service, having conducted around 1.5 million customer conversations with satisfaction scores comparable to human agents. The efficiency of this system has allowed a reduction in support staff from 9,000 to 5,000, integrating an omnichannel supervisor to assist both human and AI agents.

Mark emphasizes that AI's primary benefits will be cost optimization and revenue uplift, transforming their sales process by ensuring every lead is followed up. The agentic capability has increased efficiency and productivity in lead generation and customer interactions, with an agent deployed on their website to enhance marketing efforts. This synergy between humans and AI is a model Salesforce is successfully proving, while other companies adopt similar strategies.

There is investor confusion about the implications of entering an "agentic era," with concerns about the potential obsolescence of traditional software. The speaker notes a personal shift towards using applications like OpenAI as primary interfaces. Salesforce's integration of agents within Slack for various tasks highlights Slack's role as a thriving ecosystem for enterprise technology. Salesforce holds a 26% share of the CRM market, while ServiceNow leads with 44% in IT service management, with speculation about new entrants in the market.

Next-generation user interfaces are evolving, with Sales Cloud and Service Cloud transitioning to agentic models. An example is provided of Eaton using an agentic field service product, where agents access customer information via an app during interactions. The integration of human agents and technology is further illustrated through collaboration with PG&E for wildfire prevention. The importance of integrating the data layer, application layer, and agentic layer is emphasized for effective functionality.

The discussion also addresses managing agents amid organizational change, highlighting the need for effective change management and accountability. The evolving role of CEOs in overseeing both human and agentic workers is stressed, with an emphasis on establishing guardrails for when to escalate issues to human agents. The maturation of large language models is recognized as more evolutionary than revolutionary.

Speaker 1 emphasizes the necessity of managing AI agents with structured guardrails for effective interactions. They note that the current balance of work is evenly split between agents and humans, likening this dynamic to a self-driving Tesla requiring human intervention in complex situations. The importance of technology and personnel in facilitating collaboration across sectors like sales, service, and marketing is highlighted.

The conversation shifts to pricing models, with insights from customer interactions revealing a growing demand for comprehensive enterprise license agreements. They discuss the transition from per seat pricing to more agentic pricing, indicating that every company is on a path to becoming an agentic enterprise. Speaker 2 announces participation in Dreamforce, where Fortune 100 companies will demonstrate how to build agentic enterprises.

Speaker 2 categorizes the market into five segments, noting that each requires tailored products and has different deployment speeds. They predict that AI will significantly impact small businesses, fostering increased entrepreneurship. The importance of industry leaders is underscored, with Pfizer recognized as a pioneer in next-generation AI deployment. Salesforce's position as the largest customer of the US federal government is highlighted, along with recent successes in securing contracts.

The concept of forward deployed engineers is introduced, focusing on building applications before deals are finalized. Speaker 1 reflects on their experiences across various company sizes and the inspiration drawn from each stage. The conversation highlights the vibrant state of innovation in Silicon Valley, emphasizing excitement surrounding new technologies and startups.

The discussion stresses the importance of becoming an "agentic enterprise" and effectively communicating this at Dreamforce. Customers are increasingly interested in real-world applications and success stories rather than just new product announcements. The operational rhythms of AI-first companies differ significantly from traditional businesses, with insights shared on how larger organizations can accelerate change to compete with agile newcomers.

Leadership and vision are underscored as critical components for success, with a focus on core values such as trust, customer success, vision, and innovation. The quality of the team and daily operations are essential for execution. Concerns about hiring practices are raised, advocating for the inclusion of recent graduates who bring fresh insights and energy. The speaker expresses skepticism about some prevailing beliefs regarding AI, including misconceptions about AGI and the end of apps, while noting the rapid changes in messaging from industry leaders.

The conversation emphasizes the superiority of human creativity and insights over AI limitations, suggesting that future advancements will require new models beyond existing capabilities. An optimistic note is struck regarding the importance of human contributions in the evolving AI landscape.