AI is reshaping how marketing and sales teams work together, but the real opportunity lies in how we build and experiment with these tools before moving to full execution.

This is the perspective of Agnes O’Connell, who describes herself as being “in builder mode right now, not in execution mode quite yet.” For Agnes, the focus has been on exploring what’s possible on the back end of today’s AI platforms—work that many marketers, in her view, aren’t spending nearly enough time on.

Agnes outlines three key directions for applying AI in marketing and sales:

Building Agents to Reduce the Heavy Lift

Agnes emphasizes that new AI tools can take on the majority of the grunt work that once required big investments in software or services. “You can build tools that take 90% of the lift,” she explains. What once cost companies hundreds of thousands of dollars—like personalized gifting platforms—can now be replicated by combining agents with integrations to platforms like Amazon.

This shift makes it possible to create solutions that get teams “50% of the way there” at a fraction of the cost, fundamentally changing the economics of personalization.

Making Marketing Content Actionable for Sales

Agnes sees a major opportunity in bridging the long-standing gap between marketing and sales. Too often, sales teams never find or use the wealth of content marketing has produced. AI offers a lightweight way to fix this

Her idea: create systems that can pull the top three most relevant assets from a company’s content library for a given persona or situation—and supplement that with one or two external thought-leadership pieces that buyers will find interesting. As she puts it, “It’s another way to leverage AI… you could offer an insight in a trusted advisor capacity that’s not your content, but would be valuable for the problem your prospect is trying to solve.”

Lowering Barriers to Experimentation

For many, the hardest part of adopting AI isn’t the strategy—it’s simply getting started. Agnes recalls that even downloading Python and figuring out folder taxonomies was intimidating at first. But once that foundation was in place, building agents became much easier.

Her advice: “Don’t be intimidated by the learning process.” Starting small with manageable projects helps teams gain confidence and see value quickly, rather than waiting until everything feels perfect.

Agnes’ approach reflects both curiosity and pragmatism. By experimenting in builder mode, reducing the lift of execution, and aligning marketing and sales through actionable content, she shows how AI can create value today. And by encouraging others not to be intimidated, she reminds us that the biggest breakthroughs often start with simply taking the first step.