Selling to a Bot: AI Buying Agents Are Reshaping B2B Sales

Agentic AI is poised to advance beyond serving as a digital assistant to becoming an independent purchasing agent for businesses.

According to audit, tax, and advisory services firm PwC, 79% of companies are already adopting AI agents to help with tasks like procurement and vendor sourcing. AI agents also help sales teams craft email openers and improve forecasting while giving buyers deeper engagement as they vet and interact with vendors.

Oren Blank, VP of product at Walnut, a B2B interactive demo developer, sees AI already influencing how buyers engage during the SaaS sales process. He offers insight about what a fully agentic AI buyer may look like in the future.

Adobe and Honeywell, among others, utilize the Walnut software sales platform to address tomorrow’s biggest sales challenge — charming non-human sales prospects. It creates personalized interactive product demos to help businesses move away from generic presentations and provide a tailored experience for each potential client.

“We’re seeing this shift accelerate the need for comprehensive product-led platforms that provide deep self-assessment and self-exploration for buyers,” Blank told CRM Buyer.

Sellers now leverage this previously hidden process with analytics across the entire buyers’ journey — spanning demo creation, content curation, delivery, and deal intelligence. For company reps, it is pushing the first touchpoint much later in the sales funnel where buyers already have context, data, and a recommendation from an AI assistant, he added.

Adapting CRM for AI Buying Agents

Sales teams may no longer reap the benefits of personable, human-led introductions that can shape early-stage thinking. AI agents can handle pre-screening vendors.

“Sales teams may soon need to win over a bot just as much as they would a person. Learning how to aggregate insights like product data, competitor comparisons, pricing, and security credentials to be machine-readable will be key for continuing to secure sales,” Blank said.

The evolution of AI has shifted from prompt engineering to context engineering. The new process aggregates and indexes all relevant information into a single, cohesive environment, where AI can provide thoughtful analysis rather than surface-level responses, he explained.

Context engineering transforms deal management from reactive relationship management to proactive intelligence, anticipating needs and surfacing insights that drive informed decisions. Platforms like Walnut develop context hubs for deals that automatically capture and organize all stakeholder interactions, content engagement, demo analytics, and deal progression data in one place.

Fixing CRM Data Gaps

Traditional CRM systems struggle with data hygiene and incomplete capture. Deal rooms automatically index every interaction, track stakeholder engagement patterns, and maintain complete context across the entire buying journey, according to Blank.

“This isn’t just about housing demos and documents. It’s about creating rich, contextual intelligence that both AI systems and human decision-makers can leverage,” he explained.

The new approach provides both AI agents and human buyers with access to comprehensive, structured data, rather than fragmented information scattered across multiple systems.

Retail’s Influence on B2B Sales

Blank identifies the most significant e-commerce trend in B2B circles as the expectation for immediate, self-service evaluation. This “try before you buy” mentality is becoming standard.

B2B buyers now expect the same frictionless experience they get as consumers: research independently, evaluate asynchronously, and make decisions without gatekeepers. Even enterprise solutions that historically used content gates and qualification funnels are now trying to become “the Amazon of their industry” — making products immediately accessible for exploration.

This shift is fueling the rise of AI-powered personal shopping assistants in B2B, where AI agents help buyers navigate complex solution landscapes, compare options, and manage stakeholder evaluation processes completely asynchronously.

“The buyers control the timeline and process, while AI provides intelligent guidance throughout their journey,” Blank said.

As buyers gain more control of the process, sellers are moving from information gatekeepers to strategic advisors. The discovery and initial evaluation phases become increasingly automated and buyer-led.

Evolving Sales Relationships

The traditional SaaS sales process relies on building relationships and understanding a buyer’s pain points. With AI agents becoming the primary point of contact, trust will shift from relationship-based to outcome-based, Blank noted.

“This will elevate the salesperson’s role. While AI agents excel at data processing and initial evaluations, they can’t understand the deeper business context that drives real decisions,” he added.

As a result, salespeople transform from feature explainers to strategic advisors. Instead of walking through product capabilities, they become consultants who understand organizational dynamics, unspoken stakeholder concerns, and the change management required for successful implementation.

“The salesperson’s value becomes providing nuance that AI can’t capture,” Blank explained. “Trust isn’t built through product demonstrations anymore. It’s earned by demonstrating a deep understanding of business outcomes and the complex dynamics that make or break success.”

The Human Role in AI-Driven Sales

Blank sees a significant evolution in the role of salespeople from context readers to strategic advisors. While AI handles the product education and technical evaluation, humans become essential for navigating the organizational dynamics, change management, and business transformation aspects that determine real success.

In a future with a fully agentic AI buyer, the biggest sales challenges involve how a salesperson negotiates with a machine that is optimizing purely for objective criteria. The winners, Blank expects, will be platforms that seamlessly serve both AI and human buyers, providing machine-readable data for initial screening while enabling rich human interaction for final decisions.

However, he warns about a deeper challenge. By making software development easier, AI is also commoditizing many technical differentiators.

“In a world where AI amplifies software commoditization, the winners will be those who can be most relevant to specific pain points, needs, and use cases rather than offering generic solutions,” he explained.

Redefining the Sales Champion Role

Blank shared his vision of the sales champion concept evolving into that of a context engineer. This human salesmanship will ensure that AI systems understand an organization’s unique constraints, politics, and success criteria, which determine real implementation success.

Selling to AI agents will require sellers to be incredibly specific and personalized rather than relying on broad value propositions. Generic messaging often gets lost in the noise when AI systems evaluate hundreds of similar solutions.

“The challenge isn’t just optimizing for machine-readability. It’s standing out with precise relevance to the buyer’s exact situation,” he said.

Blank explained how the champion becomes more valuable by being the bridge between generic AI evaluation and organizational reality. The task helps AI surface decision-making reasoning that resonates with the buying committee’s specific context.

“This makes them indispensable to both the AI evaluation process and building internal confidence in the decision,” he noted of the human salespersons’ new mission.

Guardrails for AI Transactions

According to Blank, AI-driven B2B transactions will require human-in-the-loop verification, robust security frameworks, and what he called “cyber consciousness,” that is, awareness that somebody can game these systems.

“Bad actors, or even competitors, can now potentially manipulate AI evaluation processes at scale rather than targeting individual deals one by one. This creates systemic risks we’ve never faced before,” he said.

The CRM vendors who succeed will build in these safeguards from the start: audit trails, security measures that prevent systematic manipulation, and transparency so buyers understand how AI reached its recommendations, he concluded.