On Implementing AI, Internal Pushback, and the Cost of Waiting: Lessons from State of FM Ep. 4

AI adoption inside facilities management organizations rarely stall because of the technology itself. It stalls because someone has to put their name on it, walk it past legal and IT, and explain it to a room of people who are already stretched thin. On episode four of State of FM, Fexa’s Director of Customer and Product Marketing Beth Mooney sat down with Chief Customer Officer Maranda Dziekonski and VP of Product Michelle Klaer to talk through what’s actually holding operators back, and what it looks like to move past it.

What follows is a cleaned-up transcript of their conversation, edited for length and clarity.

What Actually Stops Operators From Saying Yes to AI?

Beth Mooney: In your experience, what actually gets in the way of someone saying yes to AI? What’s the real hesitation?

Maranda Dziekonski: Rarely is it the technology itself. It’s usually that nobody wants to be the person who signed off on something they can’t explain to a room of skeptical people, or the champion is carrying some career risk on the topic. They’ll call it timing, but it’s really “am I willing to put my neck out for this?” How will I get this funded? What am I going to have to do to get through security? Do I want to deal with my legal team on this? There are just so many unknowns.

Michelle Klaer: Maranda’s right, and I’d add that there’s a confidence piece too. Do I have the right ROI story? Do I have a clear vision of the outcomes we’ll achieve, and can I present that to my leadership in a way that lets me move forward feeling empowered about the conversation?

Beth: People often feel like they need to become an AI expert before they can even start that conversation. What have you seen that debunks that?

Michelle: It really comes down to knowing how the service works on a basic level and being able to talk about it in simple terms. You don’t need to know how it’s engineered. You need to know that it’s inside your system, working with your configurations and your defaults, and helping you move more efficiently. You need the use cases and the flows, not the architecture.

Maranda: The questions I hear tend to fall into three buckets. First, where does our data go? People hear a lot in the news about models being trained on data that was uploaded somewhere it shouldn’t have been, so there are real concerns there. Second, will my team actually use this, or will it just become one more tab open in their browser? What problem is it actually solving? And third, am I about to add more work to people who are already stretched thin?

What Makes AI Adoption Succeed: The Technology or the People?

Beth: This is starting to sound less like a technology decision and more like a people decision.

Michelle: This is less about AI itself and more about the technology partner and the relationship behind it. Yes, the efficiency gains matter, but what matters more is knowing your vendor is there to partner for the long haul, to hear you, to understand your workflows, and to build alongside you. That relationship, on both the customer success side and the product side, makes a real difference, because not every company partners the same way.

Maranda: I agree, and I’d say this isn’t specific to Fexa or even to AI. If you’re evaluating any vendor to solve a problem inside your organization, you have to look past the technology to the people part. What does support look like after you sign? What does implementation and change management look like? What does the partnership look like as you move from “I agree, I have this problem” to “help me build the case I need to sell this internally to the people who hold the purse strings”? With so many options out there, it stops being about whether you use a given technology and becomes about the full package.

Maranda: One more thing on that, because I want to make sure I say it: this isn’t going away. It’s revolutionary, and it’s here to stay. Either you find the right partner to help you move that needle, or somebody inside your organization makes the move without you, and sometimes they’ll pick something you’re not going to be happy with.

Michelle: AI is software now. That’s just part of the roadmap for any vendor going forward. Thinking about it that way changes the question from “should I use AI” to “what makes a good vendor,” full stop.

How Do You Stop Sitting on the Fence?

Beth: For someone sitting on the fence right now, unsure whether to move or wait, what would you tell them?

Maranda: Stop deciding about AI in the abstract. Decide about one workflow, one problem, that if you fixed it would make your team’s lives infinitely better or make you look like an operational genius. Instead of asking “should I use AI or not,” look at the problems you’re actually trying to solve, and plug the tools into those.

Michelle: I’d reframe it even further to “what could I automate?” What could I set and forget that would make my day-to-day more efficient, and how much time would that give back to me and my team? AI has a scary connotation for a lot of people. But if we break it down to “we can automate this so you don’t have to,” it’s a much easier door to walk through, and it lets us have a real conversation about the use cases we offer.

Maranda: Automation is only one piece of it. If I were a director of facilities management looking at my operations, I’d pick out the two or three problems that are continuously consuming my budget, whether that’s false truck rolls or my team chasing down information on work orders. I’d look at the ten things I could solve for and pick the two or three that would free up the most budget from reactive management, so I could shift it toward proactive management.

Beth: What do you tell someone who’s worried about what happens to the time or budget they free up?

Maranda: I’d ask: do you have enough time right now to finish everything on your plate every week? Are you sitting there wondering what to work on next? Probably not. Peeling off the work that doesn’t need human judgment frees up your time and your team’s time for the things that actually move your operations from good to great.

Michelle: There are always wishlist projects that never get touched because the team is heads down on the checklist. If you can automate a piece of that checklist and get the time back, you can finally dig into the strategic work you didn’t have room for because you were firefighting.

Maranda: One thing I’d flag for anyone managing people through this: there’s a real dopamine hit that comes from checking something off a list. When AI takes over the busy work, that hit doesn’t disappear, it just needs a new place to land. That’s worth thinking through as part of change management. Where do those hours go, and how does your team get that same sense of progress in a different way?

How Do You Bring a Hesitant Team Along?

Beth: What does it look like when part of a team is on board and part of it isn’t?

Michelle: Hesitancy usually comes from a gap in understanding what exists and how it can help. It takes conversations, demonstrations, and data to move someone from “I was kind of hesitant” to “that’s actually going to help me.” Maranda and I spend a lot of time together in those exact conversations with customers.

Maranda: I see Michelle more than I see my husband at this point. My advice is to go back to the problem statements first and make sure everyone agrees the problem is real. Then get creative: what would this person need to see to believe the technology solves it? What proof points would help, and how do you gather them together, maybe through a pilot? Once you’ve got buy-in that it’s the right problem to solve, you can get creative about how you solve it and how you measure progress.

Michelle: The key is that it’s not one-size-fits-all. We partner with each customer to figure out what that looks like specifically for them.

What Do IT, Legal, and Security Actually Need to Know?

Beth: Let’s talk about IT, legal, and security, since those conversations can get complicated fast. What do people assume they have to figure out alone that’s already been figured out for them?

Maranda: The good news is these teams are predictable. IT usually asks about data isolation, SSO, MFA, and where the model actually runs. Legal usually asks about data retention, whether their data trains anything, and indemnification. Finance wants to know where the money comes from and what the run rate looks like. Because these questions are so predictable, we can put together a packet that helps our partners walk into those conversations prepared. In 27 years of doing this, I’ve never seen a security question that hasn’t already been asked by somebody else’s IT or legal team.

Michelle: These aren’t new revelations. We have pre-built questionnaires customers can request, and a lot of the foundational questions get answered through the reviews customers already go through when they come on board with Fexa. One of the most common questions is whether we’re using customer data to train models or sharing it across environments, and the answer is no.

Maranda: One hurdle I’ve seen is how mature a company’s IT and legal teams are with procuring software versus services. Sometimes I’ll get an MSA written for someone providing services in person, not for software. We can partner through that, but it’s a reminder that the vendor you choose needs the know-how to help you build the right internal talk track, not just the product.

What Does It Actually Cost to Wait Too Long?

Beth: What’s the actual cost of waiting too long to move on this?

Michelle: The risk isn’t loud. It’s the silent effect of not becoming more efficient and not transforming your processes. You stay stuck in “good” because you’re still doing things the old way while the technology to break through those barriers already exists. And if you’re leading people, there’s a change management cost to bringing your team along later rather than now.

Maranda: I’ll say something a little provocative. I think we’re heading toward a point where every leader will be judged on how well they leverage AI to run their organization efficiently and use their people where people actually add the most value. That point is probably sooner than later. The real cost of waiting is the quiet compounding of not changing, and then having to play catch-up once everyone else has already adopted the use cases that matter.

Where Should You Actually Start?

Beth: For someone who’s nervous but ready to start the conversation, what’s your closing advice?

Maranda: Narrow your scope. Capture your baseline measurements before you change anything. Be crisp about what change management looks like in your organization, what your problem statements are, and how you’ll measure progress. Treat it like any other technology adoption. AI carries a scary reputation for some people, but it’s just another tool in the toolkit, and if you measure it like one, you’ll be ahead of a lot of your peers.

Michelle: If that still feels daunting, reach out. We’d love to help you build the frameworks to move forward and answer questions along the way. You don’t have to go at it alone.

Maranda: You’re not alone. Lean on your partners.

State of FM is available on YouTube and Spotify. New episodes drop regularly, covering the real conversations happening across facilities management today.