In a recent episode of GTM in the Tofu Bowl, Elaine Zerbe sat down with David Yockelson, Distinguished VP Analyst and Fellow at Gartner, to explore how artificial intelligence is reshaping the world of go-to-market strategy. Yockelson—who has spent decades at the intersection of marketing, product strategy, and technology—shared his perspective on AI’s role in account-based marketing, the evolution of agentic workflows, and what organizations get right (and wrong) when deploying AI in their go-to-market motion.

From Marketing to Go-to-Market: A Natural Evolution

Yockelson’s career has spanned both sides of the table—practitioner and analyst. His shift to focusing on go-to-market was natural:

• Alignment matters: Sales, marketing, and customer success often operate on different data sets, creating fragmentation. Go-to-market, he argues, is fundamentally about aligning these functions to tell the right stories and deliver the right products.
• Account-based approaches: What some still narrowly view as “ABM” should be understood more broadly as account-based go-to-market. This requires closing the gap between siloed functions and uniting around a shared ICP (ideal customer profile) definition.

AI’s Role in Account Selection and Precision

One of the clearest opportunities for AI lies in account selection:

• Sharper ICPs: AI can refine and validate ICPs, often surfacing multiple high-probability cohorts where a company has outsized odds of winning.
• Persona precision: Within those accounts, AI enables a more nuanced understanding of buying centers, ensuring that demos, case studies, and content align to the needs of specific decision-makers.
• Closing messy human gaps: As Elaine noted, humans are messy; AI, less so. Aligning on shared datasets can reduce friction and bring a more data-driven approach to campaign foundations.

Agents vs. Assistants: Where We Really Are

A major focus of the conversation was agentic workflows. Yockelson drew a clear line:

• Assistants: Transactional, conversational tools (e.g., co-pilots) that provide outputs when prompted.
• Agents: Autonomous or semi-autonomous entities that execute tasks toward defined goals, with guardrails in place.

While there’s hype around “agents,” Yockelson estimated that over 75%—closer to 90%—of so-called agent deployments today aren’t truly agents, but rather automations bolted onto LLMs. Marketers are still in phase one of adoption: building trust. Phase two will involve “trust but verify” workflows with humans in the loop, and only then will we move toward true autonomous agentic systems.

Real-World Deployments: Signals and Lenovo’s AI Studio

Despite the noise, there are promising implementations:

• Lenovo’s AI Studio: An agent-based system that assembles language-, geography-, and channel-specific content packages in hours instead of weeks.
• Signals > intent data: Yockelson and Zerbe agreed that intent data has been a “black box.” Signals, by contrast, are specific, actionable, and tied to prescribed plays. With historical CRM data and behavioral cues, signals can fuel next-best-action recommendations for agents.

Humans, Orchestrators, and the Path Forward

Will orchestration belong to humans or master-agent systems? Yockelson’s answer: both. Strategic jobs remain human-led, but orchestration of task-level agents will increasingly be automated.

• What most get wrong: Companies stick to “low-hanging fruit” like content generation and image creation, avoiding the harder (but more valuable) work of transforming execution and process.
• Process-first thinking: Organizations must identify bottlenecks in product launches, campaign orchestration, and sales cycles—and then decide what should be human-led, augmented, or automated.
• Cultural readiness: Many marketers resist automation out of caution, but experimentation is non-negotiable. Hackathons, carve-outs for testing, and “fail-fast” spaces are ways leading companies are pushing adoption forward.

Looking Ahead: Agents, Outcomes, and Pricing Models

Yockelson’s ongoing research points toward a shift in how agent-based products will be marketed and sold:

• Outcomes over features: Buyers don’t want to hear about “feeds and speeds.” They want clarity on process improvements and business outcomes.
• Outcome-based pricing: As agents become outcome-delivery machines, pricing and packaging will need to shift to reflect that value.
• Experimentation imperative: His advice is simple—if you’re not experimenting with AI in your GTM motion today, you’re already behind.

Final Takeaway

The conversation underscored both optimism and realism. AI is already helping refine account focus, generate signals, and speed content workflows. But most organizations are still at the beginning: experimenting, testing, and learning to trust. The leap to fully autonomous agents won’t happen overnight—it will require cultural readiness, process clarity, and a willingness to put strategy before technology.

For now, Yockelson’s call to action is clear: don’t wait for perfect. Start experimenting, carve out space for learning, and focus on outcomes—not outputs.

‍