The Agentic Evolution according to Salesforce’s CMO

AI agents fail because companies lack proper data context and change management. Ariel Kelman is President and Chief Marketing Officer at Salesforce, leading their global marketing organization and AgentForce platform development. He discusses Salesforce's trust-first approach using their Data360 customer data platform to provide AI agents with complete customer context, implementing two-way email campaigns that allow interactive customer engagement, and deploying lead qualification agents that generated $27 million in incremental pipeline by processing 200,000 previously unworked leads.
About the speaker

Ariel Kelman

Salesforce

- Salesforce

Ariel Kelman is President and Chief Marketing Officer at Salesforce

Episode Chapters

  • 01:38: Salesforce's AgentForce Platform

    Discussion of Salesforce's approach to agentic AI through AgentForce and the industry-wide 95% failure rate of generative AI pilots.

  • 02:03: Context and Data Requirements

    Explanation of why business AI requires enterprise data context at runtime, unlike consumer AI applications that rely on pre-trained models.

  • 04:45: Native Integration vs MCP

    Comparison between Salesforce's native data integration approach and Model Context Protocol solutions for co ecting enterprise data to AI systems.

  • 06:55: Common AI Implementation Failures

    Analysis of technical and cultural barriers preventing successful AI deployment, emphasizing the importance of change management and leadership modeling.

  • 10:06: Leadership-Driven AI Adoption

    Examples of how executives at Salesforce personally experiment with AI tools to model behavior and drive organizational change.

  • 13:53: Salesforce's AI ROI Results

    Specific metrics from Salesforce's AI implementations, including 77% case resolution rates, $100M+ savings, and 20% pipeline increases.

  • 17:09: Quality Over Quantity Strategy

    How AI agents can reduce lead volume while improving conversion rates by better qualifying prospects and directing them appropriately.

  • 20:13: Restructuring for Agentic Workflows

    Discussion of how marketing teams need to adapt their processes and metrics to accommodate AI-driven customer interactions.

  • 21:48: Human-Centered Content Strategy

    How organizing website content for AI consumption actually improves human readability and customer-centric communication.

  • 24:36: Evolution Over Revolution

    Advice for marketers to gradually integrate AI capabilities into existing workflows rather than completely replacing current systems.

  • 26:09: SaaS Platform Transformation

    Analysis of how successful SaaS companies are evolving to include agentic workflows while maintaining data security and deterministic processes.

  • 27:51: CEO Leadership Traits

    Common characteristics of successful tech CEOs, focusing on intellectual curiosity and adaptability to new technologies.

  • 29:58: Task Automation vs Human Replacement

    Clarification that AI should automate specific tasks rather than entire job roles, increasing productivity while maintaining human oversight.

  • 33:10: Human Oversight Requirements

    Discussion of which marketing activities require human review, particularly messaging and creative content that represents brand voice.

  • 34:35: Future of Video Production

    Prediction that traditional filmed TV advertisements will become obsolete as AI video generation tools advance rapidly.

  • 37:18: AI Agent Readiness Indicators

    Key signals that indicate when companies are prepared to implement AI agents, focusing on data foundation and process understanding.

Episode Summary

  • The Agentic Evolution according to Salesforce's CMO

    Introduction

    Ariel Kelman, President and Chief Marketing Officer at Salesforce, reveals why 95% of generative AI pilots fail to deliver measurable business impact—and what separates the 5% that actually work. Leading marketing for the world's largest cloud-based CRM, Kelman oversees the rollout of AgentForce, Salesforce's platform for building AI-powered workflows that has already powered over 2.5 million customer support conversations. His insights on building trust-first AI implementations offer a roadmap for marketers struggling to move beyond AI experiments to becoming truly agentic organizations.

  • The Context Problem: Why Most AI Agents Fail

    The fundamental difference between consumer AI success and business AI failure comes down to context and data. While ChatGPT can easily provide chocolate chip cookie recipes from thousands of training examples, business AI requires real-time access to specific company data that isn't in any model's training set. "Understanding the context of the customer is absolutely required if you're going to get any useful information back out of an AI model," Kelman explains. Most companies fail because they focus on the AI functionality without building the necessary data infrastructure—attempting to construct the house before laying the foundation.

  • Data Infrastructure as the Secret Sauce

    Salesforce's approach with AgentForce prioritizes building a trusted way to co ect enterprise data before focusing on agent functionality. This includes integrating customer purchase history, geographic location for regulatory compliance, previous support interactions, and all touchpoints across the customer journey. The platform's Data360 customer data platform pulls information from enterprise data warehouses and operational systems, creating a single source of truth. This comprehensive context allows AI agents to operate with the same information a human agent would have, dramatically improving their effectiveness and reliability.

  • Beyond Technology: The Human Change Management Challenge

    Technical infrastructure represents only half the equation for successful AI implementation. The people aspect proves equally critical, with many failures stemming from resistance to changing work methods. Kelman observes that employees have varying "change comfort" ratings—some eagerly experiment with new tools while others resist departing from established processes. At Salesforce, executive adoption drives cultural transformation. Their CEO actively experiments with AI tools, while the chief creative officer has tested over 100 video production tools, and the product marketing lead uses Gemini to simulate reporter interviews for media training.

  • Leadership Modeling Drives Adoption

    "If the leaders are getting their hands dirty, I think it's a lot more credible if you're asking people to do the same thing," Kelman notes. This top-down approach transforms AI from a threatening technology to an empowering tool. Leaders demonstrate practical applications—from AI-generated video production to automated interview preparation—showing teams how AI enhances rather than replaces their expertise. This visible leadership engagement proves essential for overcoming the natural resistance to workflow changes.

  • Measurable Results: From Experiments to Business Impact

    Salesforce's own AI implementations demonstrate the tangible benefits of properly executed agentic workflows. Their customer support AgentForce deployment has resolved 77% of all support cases, saving over $100 million while allowing human agents to focus on complex issues. In marketing, their website AI agent initially reduced lead volume but increased sales pipeline by 20% by qualifying prospects more effectively. Additionally, routing sales leads through AI agents first enabled them to work nearly 200,000 previously untouched leads, generating $27 million in incremental pipeline.

  • Rethinking Success Metrics

    These results required fundamental changes in how success is measured. Traditional lead volume metrics gave way to pipeline quality measurements. Kelman emphasizes the importance of metric flexibility: "You gotta be fluid with your metrics and make sure they're always aligned with the business outcomes you're trying to drive." This includes tracking new engagement types, like Slack notifications to sales teams when existing customers visit pricing pages—interactions that don't generate traditional leads but drive significant business value.

  • The Future of Marketing Work: Evolution, Not Revolution

    Contrary to fears about AI replacing marketers, Kelman advocates focusing on task automation rather than job replacement. The goal isn't eliminating positions but increasing productivity—allowing teams to say yes more often, meet aggressive deadlines without burnout, and focus on higher-value creative work. "You're not going to be replaced by AI. You're going to be replaced by someone who's great at using AI," he emphasizes. This evolution requires marketers to embrace AI as a productivity multiplier rather than a threat.

  • Conclusion

    The path from AI experimentation to agentic marketing success requires three essential elements: robust data infrastructure providing comprehensive context, leadership-driven change management, and flexible metrics aligned with business outcomes. Salesforce's experience demonstrates that the 5% of successful AI implementations share these characteristics, transforming AI from an experimental technology into a business transformation engine. For marketers ready to make this transition, the message is clear: focus on automating repetitive tasks with good data foundations, lead by example, and measure what truly matters to your business.
About the speaker

Ariel Kelman

Salesforce

- Salesforce

Ariel Kelman is President and Chief Marketing Officer at Salesforce

Related Podcasts by Category

B2B

See More

AI, Email Marketing

See More

SAAS, Marketing Consultancy

See More

Artificial Intelligence, Data-driven products, Email Marketing

See More

Up Next: