What 100 Million Missed Leads Taught Us About Selling Smarter

The transition of AI from tool to teammate is solidly, profoundly, here. Companies are deploying AI agents that book meetings and handle other tasks once reserved for humans. But this new era comes with critical questions every leader must answer: How do you become an AI-first company? How can you give AI enough autonomy to deliver real results without risking compliance, trust, and control?

The promise is enormous. When agents can reason and act independently, they create capacity that human teams cannot match. But the guardrails matter just as much as the capabilities. Give agents too much freedom and you risk regulatory nightmares; too little, and they're reduced to expensive chatbots that don’t move the needle. In this issue, we explore what it takes to get this right, from real-world wins to the frameworks that define the pioneers.

How we stopped leaving money on the table

Sales teams are limited by time, and it shows. Millions of promising leads slip through the cracks simply because there aren’t enough humans to follow up on them. Now, with agentic AI, that’s changing. At Salesforce, Agentforce autonomously reached out to prospects, qualified them, and booked 800 meetings in just a few months. The result? Dozens of new deals closed that would otherwise never have materialized. Here’s how we did it, and what we learned. 

Lucky 7: The keys to an agentic enterprise

Imagine if employees had an AI-powered teammate that listens, reasons, and takes action right alongside them. No more sifting through reports or waiting on approvals. Just humans and AI working in sync to move faster and smarter. That’s the agentic enterprise, a workplace where AI isn’t a bolted-on widget but an ingrained part of every business process. Think of it as Google Maps for work, where AI guides every decision in real time with the accuracy to take the fastest route forward. What does an agentic enterprise look like? Here are 7 key elements. 

Your AI agents need autonomy, but how much is too much? 

Companies are walking a tightrope when it comes to how much autonomy to give AI agents. On the one hand, too little autonomy could stifle innovation and hamstring productivity. On the other, too much could trigger a regulatory or reputational mess. The key is to strike the right balance between empowering agents to reason and act, and sticking to guardrails that preserve compliance, reliability, and trust. But how do you do that? Experts weigh in. 

While you were prompting … stories you might have missed from the world of AI

A recent study by Sweden’s KTH Royal Institute of Technology found that fully AI-generated memes scored higher on average for humor, creativity, and shareability than those created by humans — but the funniest individual memes were still created by humans. Bottom line? Machines win the popularity contest, but you still need a human being to make people snort-laugh.

This newsletter was curated by Lisa DiCarlo Lee, Contributing Editor, Salesforce.