Fixing Shadow AI and Tool Sprawl in Enterprise Marketing - with Gillian Hinkle of Salesforce

**Matthew DeMello** (0:19)
Welcome everyone to the AI in Business Podcast. I'm Matthew DeMello, editorial director here at Emerge AI Research. Today's guest is Gillian Hinkle, Senior Director of Growth and Digital Marketing for Heroku at Salesforce. Gillian joins us on today's show to unpack how enterprise marketing leaders can cut through operational complexity, distinguishing automation from true AI and design practical human in the loop approaches for deploying generative and agentic systems inside real workflows. Our conversation also examines how teams can reduce tool creep, improve data hygiene and governance, and narrow AI initiatives to high-impact use cases, such as lead qualification and customer service handoffs, driving measurable gains in efficiency, employee engagement, and decision quality without risking compliance or brand risk. But first, are you driving AI transformation at your organization, or maybe you're guiding critical decisions on AI investments, strategy, or deployment? If so, the AI in Business Podcast wants to hear from you. Each year, Emerge AI Research features hundreds of executive thought leaders, everyone from the CIO of Goldman Sachs to the head of AI at Raytheon, and AI pioneers like Yoshua Bengio. With nearly a million annual listeners, AI in Business is the go-to destination for enterprise leaders navigating real-world AI adoption. You don't need to be an engineer or a technical expert to be on the show. If you're involved in AI implementation decision-making or strategy within your company, this is your opportunity to share your insights with a global audience of your peers. If you believe you can help other leaders move the needle on AI ROI, visit emerj.com and fill out our thought leader submission form. That's emerj.com and click on be an expert. You can also click the link in the description of today's show on your preferred podcast platform. That's emerj.com/expertone. Again, that's emerj.com/expertone. Without further ado, here's our conversation with Gillian.
Gillian, welcome to the program, it's a great pleasure having you.

**Gillian Hinkle** (2:32)
Thank you, great to be on the program.

**Matthew DeMello** (2:34)
Absolutely, we're hearing all the time from enterprise marketing teams that they are confronting unprecedented operational complexity. You know, their tool stacks are expanding faster than leaders can really govern them, AI features appear across every platform, there's very little standardization, and frontline teams are adopting their own solutions to fill workflow gaps. These pressures make it difficult to maintain coherence, measure impact and demonstrate value quickly. With these changes in mind, we just want to start at the highest level, especially since I think maybe rivaling agentic AI for AI word of the year might be shadow AI is maybe one of the more esoteric phrases that have come up on the show within the last year. Just from your vantage point, what do you see is the most important challenge enterprise marketing leaders face today when it comes to that operational complexity and tool creep?

**Gillian Hinkle** (3:27)
Yeah, I think it's around where do you start? Where do you, where do you, and how do you choose the correct projects? So it's really easy because it's an exciting time. There's been so much change to want to do it all. And that, in my opinion, is always the wrong answer. So instead, when people ask me about this, I ask them about their pain points and the problems they're trying to solve. And then, how does it align with potential AI tools? And is there anything that exists in their existing systems that are AI tools instead of going elsewhere? And when I think about how I break that down, I tend to put it into three parts. I tend to think about tool evaluation. I think about what is AI and what is automation, because sometimes people confuse those two things. They sort of conflate the two.
Or additionally, within AI, sometimes they confuse things like generative AI with agentic solutions. And then the other sort of pillar of those, of my three hot topics, is the data. AI needs data. And it has to be good, clean, compliant data. So you have to evaluate those three things.

**Matthew DeMello** (4:45)
Right, right. That sort of tool automation versus AI. We'll get into that difference in a moment. And really the capabilities we're seeing, where they really blur the lines between generative and agentic.
And then data hygiene, which I think we've just as a practical manner on the show, I think we've been talking about even before kind of the ChatGPT explosion. That discipline has kind of always been there. I want to start just with tool automation versus AI. Because I think one of the big gripes, especially from the research side of AI, is that these things, just for how this was rolled out to the public, and even going back to the real 60 year, 70 year history of artificial intelligence, going back to the 50s, that it always had this blanket term of artificial intelligence, but the capabilities that are all under those umbrella are very, very different things. You're talking about a machine gun and umbrella, you know, like, you know, they serve extremely different functions, especially when you get in even for deterministic capabilities, machine learning versus predictive analytics are just two completely different animals, and it'll all get called AI. And when someone expresses maybe in kind of the mainstream media conversation we have about, you know, this perception of zero some job losses, it's like, oh, all AI is bad. And it's like, well, man, actually, you know, the machine learning deterministic stuff wouldn't necessarily take your job and has actually been making your job easier for the last 10 years, whether you know it or not. But that is extremely different than generative AI, agentic AI, that has these properties that you will use as digital FTs. Let's get into that difference. Dividing tool automation from AI, how do you look at the difference in terms of training your own employees?