7 Ways To Get Your Agentic AI Pilot Unstuck

AI — and agentic AI, in particular — is a top priority for Salesforce, and I’ve been fortunate to help lead that charge. My background in organizational design, change management, and strategy has never been more relevant. As companies move from AI experimentation to real-world deployment, one truth stands out: Organizational readiness is key to your success.

Don’t get swept up in the AI hype. One of the biggest blockers to adoption is applying the wrong tech to the wrong problem. Agentic AI is powerful, but it needs a focused use case with a clear goal and well-defined tasks.

Once you’ve chosen that use case, design the agent experience thoughtfully. The right user experience can mean the difference between full adoption and shelfware. To succeed, keep these points in mind:

• Use the right tech for the right job
• Build trust in the AI agent
• Make the experience better than what you provide today
• Set up your organization to own and scale agentic solutions

In this newsletter and on our new agentic AI resource page, we’ll help you build the bridge from AI pilot to full-scale deployment. We’ll explore why pilots get stuck, and how to avoid common pitfalls. We’ll examine whether a formal AI committee can help govern your initiatives, and offer tips on how to improve your AI agent’s performance, based on what we've learned ourselves. Consider this your guide to building an AI strategy that starts strong — and lasts.

-Irina Gutman, Regional Vice President, Agentforce Engineering

Listen to my podcast on YouTube, Apple, or Spotify

Why AI pilots get stuck, and how to unstick them 

Companies are launching agentic AI pilots at a record pace, yet the number of projects that reach full deployment remains stubbornly low. Why do AI projects go off the rails, and how can you get back on track? We break down the seven reasons so many AI pilots fail to launch, and share strategies to break the cycle. Here’s a teaser: You’re focusing too much on the technology. 

Do you need an AI committee?

Your company doesn’t need more red tape — but it does need a cross-functional AI committee. This isn’t about bureaucracy: It’s about aligning strategy, speeding execution, and making sure AI delivers real impact across your business. So, what exactly does this team do, who belongs on it, and why is it so critical? We delve into the specifics, and share five key reasons every company needs one.

3 things we learned after launching an AI agent on our support site

Our customers told us our help site was bogged down with too much information, and getting answers took too long. So, last October, we launched Agentforce, our platform for building and deploying AI agents, to deliver smarter, faster support. But like any new employee, our agent had a learning curve — with some lessons more challenging than others. In this article, we share what we got right, what we had to rethink, and the key takeaways from our effort. This may provide a blueprint for your own AI agent projects. 

From hair dryers to air fryers, SharkNinja makes customer service more human

SharkNinja makes smart, fun-to-use household appliances that millions of people use every day. Learn how the company employs Salesforce tools, including Agentforce, to make its customer interactions more personalized, integrated, and dynamic. One example: If a customer asks how to use their new blender, Agentforce provides not only tips, but a smoothie recipe. This type of authentic interaction, says SharkNinja CIO Velia Carboni, offers a new level of service. Get the whole story.