Don't expect an AI agent to invent a successful sales process. First, have your human team identify and document what works—effective emails, scripts, and objection handling. Then, train the AI on this proven playbook to execute it flawlessly and at scale. The AI is a scaling tool, not a strategist from day one.

Unlike humans who respond to branding and persuasion, AI agents make decisions based on structured, machine-usable data. To win over agent customers, companies must prioritize clear documentation, defined permissions, and verifiable trust signals over traditional marketing copy and aesthetics. Your product's value must be computable.

Don't treat evals as a mere checklist. Instead, use them as a creative tool to discover opportunities. A well-designed eval can reveal that a product is underperforming for a specific user segment, pointing directly to areas for high-impact improvement that a simple "vibe check" would miss.

Moonshot AI overcomes customer skepticism in its AI recommendations by focusing on quantifiable outcomes. Instead of explaining the technology, they demonstrate value by showing clients the direct increase in revenue from the AI's optimizations. Tangible financial results become the ultimate trust-builder.

Move beyond static scripts by using AI for dynamic sales training. Feed ChatGPT your call transcripts and common objections, then ask it to act as a specific buyer persona. Practice handling its objections in a role-playing chat, and conclude by asking it to provide a score and feedback on your performance.

There are three levels of trust for customer data: CRM data (low), customer words (medium), and customer actions (high). Use AI to compile timelines of successful customer actions (e.g., product usage) to build reliable hypotheses about who to target next.

Contrary to fears of customer backlash, data from Bret Taylor's company Sierra shows that AI agents identifying themselves as AI—and even admitting they can make mistakes—builds trust. This transparency, combined with AI's patience and consistency, often results in customer satisfaction scores that are higher than those for previous human interactions.

To overcome customer trust issues with new AI features, avoid a 'big bang' rollout. Instead, launch with a pilot group. This approach allows the AI model to be trained on real-world data in a controlled environment, improving its accuracy and demonstrating value before a wider release.

Don't abandon AI after one bad result. Treat it like a new SDR and use a 'shuttle run' approach: give it a small task (find 5 accounts), review the output, provide feedback, and repeat for each step (contacts, emails). This upfront calibration is crucial for long-term success.