Florida Keys Tourism CEO on Leading AI Innovation

Kara Franker, president and CEO of The Florida Keys & Key West tourism organization, was among the first DMO leaders to embrace AI platforms and processes in 2023 when she was chief executive at Visit Estes Park in Colorado.

In May 2024, Franker published two papers on AI strategy and policy:

• From the Base of the Rocky Mountains to the World: A case study through Visit Estes Park’s journey into Generative AI solutions — and how your DMO can carve its own AI path

• Navigating AI and the Law in the Travel Industry: An Introduction to Legal Implications for DMOs

The first paper highlights how Visit Estes Park integrated GuideGeek AI chat into their website and three Meta platforms: WhatsApp, Facebook Messenger and Instagram. The DMO also branded the chat Rocky Mountain Roamer to give it an identity and bit of personality.

The second paper focused on AI policy and data privacy, security and compliance. Franker, who is an attorney, co-wrote that with Roxanne Steinhoff, also an attorney at Steinhoff Law, who leads the firm’s destination marketing and sports tourism practice.

I caught up with Franker to see how her perspectives on AI strategy and policy are evolving almost a year later.

Greg Oates: You've been leading AI strategy in tourism for almost two years beginning with your work at Visit Estes Park. However, there’s still a lack of AI training today specific for DMO CEOs, and many leaders are presently unsure how to proceed with AI. As a CEO, what’s your take on that? 

Kara Franker: I'm not surprised that there's a lack of training because there's just so much for everybody to wrap their heads around. I think some people are trying to watch and wait and see. My philosophy has been more of a fall-forward, fail-forward kind of approach. That's why I put the white papers out. Anybody could learn from what we had done, and do it better. You know, a case study is sometimes the best way to figure this stuff out, so that's why we did it.

The second paper on AI policy was with Roxanne Steinhoff on the legal and ethical implications of AI. Because when I would go and speak in front of groups, those were always the big questions from the leadership. They don't know how far is too far once you open this can of worms. That's why we got together and started to put guardrails in place that we thought might make sense for people to study, just as a first stab at it. That all launched last summer, so I do think that there's a little bit of education out there but we could definitely use more. 

GO: What do DMOs need to know about AI policy at a very high level in terms of privacy and security? 

KF: Employee and client information should be kept private and not uploaded into any AI tools. That could be a big deal, thinking of all the different rules and regulations about collecting information, especially how it's more stringent in Europe than in the United States. That can cause potential problems if you're uploading, say, a client list that has personal cell phones, email addresses or other information that people just don't want out there. So, it's really more about information that's attached to a person.

And then, is there competitive data that you don't necessarily want others to have? Possibly, but so much of what we do in the DMO space is public anyways so the lines for that are pretty blurry. I’m trying to imagine a scenario where a DMO is, like, basically hacking somebody else's information through AI. I can’t imagine that’s worth the time and effort, or even possible. 

GO: A few months ago we were discussing what AI tools might be available for DMO departments like HR and accounting to help with privacy, compliance, efficiency, etc. But everything available seems to be enterprise grade and out of reach for DMOs. So what’s the answer for scaling AI across an organization?

KF: I think this is a huge need. I hired somebody from Cisco who’s my new VP of technology and partnerships. She's a total AI nerd like I am, and way more qualified to dive into all this. She and I have been looking at different options to put together, almost like little plugins. We really, really need this. For example, how do we take an AI solution from the legal world to analyze all our contracts so that we don't have to spend as many hours with attorneys, because we really just need a template, right?

And then for HR, same thing. How do we organize all our employee files into a much more streamlined and effective system? And then there’s our policies. We're a nonprofit but everything we do is so heavily influenced by County government. We have so many compliance issues, so I want an AI to learn every single piece of County policies so we’re able to compare and contrast what we're doing with those policies.

I haven't been able to find anything to help with these things. We've looked at a lot of different options but it's like what you're saying. They're these big enterprise pieces built for law firms or financial people, and they’re all really expensive. None of them fit the mold for improving different operations at a DMO.

In my mind, the marketing side is covered, especially with GuideGeek, and there are so many great options now for everything involved with marketing. So I feel like that's been tackled and done really well. But for the operations, we need more. I mean, we’ve just got to keep digging deeper, I think. 

GO: When you and your technology person from Cisco figure all this out, will you share what you find with the rest of us?

KF: Of course, always. 

GO: Backing up a bit, you were arguably the very first DMO CEO to lean into AI, at least in North America, when you were with Visit Estes Park. How did you convince your board that investing in AI was important at the time, and what were some of your processes with bringing along staff? 

KF: We were fortunate in the sense that the board was all in and the staff were all in. And I don't mean to make age a part of it, but the staff was a little on the young side and they were all very ambitious. Everybody wanted to try to figure it out. So we made the argument to the board, if we could figure out some of what AI could do, it would make the community stronger and they could take pieces of it back to their businesses. There wasn't really anybody who was against experimenting and trying it out. We made some mistakes along the way and we put that in the white paper.

Also, when we first rolled out GuideGeek, which scrapes the website, we found a few holes in our content. There were particular board members who didn't have very robust information about their business, so they got a little bit mad, but not in a, like, burn it down way. It was more of, "Oh, okay, we have some work to do." So part of the AI process has to actually be community outreach. It can't just be about the technology. We have to talk to all our partners and get really robust information from them, so when the AI crawls the site it gets the best possible information about our partners. 

GO: One thing you did at Visit Estes Park was brand your GuideGeek AI as Rocky Mountain Roamer on your website and three Meta platforms. We use that as a case study a lot, and emphasize that a DMO should promote their AI just like any other channel, right?

KF: Definitely. We knew we had to put money behind marketing the tool if we wanted people to interact with it. Traditionally with a DMO, one of your calls to action is to send people to the website. Our new call to action for a lot of our campaigns back then was to send people to the AI, so we called it Rocky Mountain Roamer to help promote and drive traffic to it.

And then in terms of KPIs, we wanted to track all the interactions you can see in the backend, but not just how many people used it. We wanted to know how long people interacted with the AI, what their journey was, and what they were asking about because that’s where things get really interesting.

GO: After the initial thrill of launching GuideGeek, how did you manage how staff used AI internally? Were there any incentives involved to help spur adoption and experimentation?

KF: The incentive for my team was, well, it wasn't really an incentive. It was a requirement. They had to write weekly reports for me about how they tested something with AI, whatever it was. It could apply to their job or not, I didn't really care. So I almost forced them to try AI.

But then people started really getting into it. They were getting more and more user friendly with different AI tools, and they would be like, “Oh, I like Anthropic because of this or that,” because they were branching out from just using ChatGPT.

So, yeah, there has to be almost like a culture of AI in your DMO for it to really work, and it's on the CEO to create and foster that.