As AI implementation accelerates across hospitals and health systems, leaders are recognizing that successful initiatives hinge less on the sophistication or “flashiness” of the technology itself. Instead, success increasingly depends on solutions that are trusted, embedded and measured — and whether stakeholders, processes and change management are fully aligned.
To learn more about what it takes to effectively deploy AI across healthcare organizations of all sizes, Becker’s Healthcare spoke with Alison Thebado, vice president of customer success at Luma Health.
Question: What are the first signals that a health system’s AI rollout in patient access will go well, or that it’s going to struggle?
Alison Thebado: Health systems that succeed with AI don’t always have the biggest budgets or the most sophisticated IT teams. They treat AI implementations as organizational change projects that happen to have a technology component.
If organizations don’t standardize their data and processes before they automate, that creates significant challenges. The best implementations clean up their data and processes prior to implementing new technology.
Also, for AI adoption to happen, staff need to be onboard. Positioning technology as a cost-cutting measure is often a roadblock. It’s better to leverage AI as a way for staff to shed manual work they don’t enjoy.
Another problem is when outcomes aren’t clearly defined prior to an initiative’s launch. Organizations must determine their baseline and align on where they want to go. In addition, poorly thought-out patient communication strategies can be a pitfall. Robust patient engagement depends on conversations about the patient experience.
Q: Staff often hear AI will make their job easier, and then experience workflow disruption. How do you help customers manage the gap between the promise and the reality?
AT: The gap is real. In reality, AI doesn’t necessarily make jobs easier right away. It makes jobs different. As a partner, our job is managing that gap and helping clients achieve their desired outcomes.
We believe the fastest path to AI adoption is identifying and addressing the manual processes that staff members hate the most. If AI visibly eliminates pain points, organizations get believers, and believers become internal champions.
Time and again, I’ve seen staff who are reluctant and fearful about AI. Once we reframe it as a way to eliminate tasks they dislike, they champion AI. We encourage clients to track how staff spend their time before and after AI. If a front desk coordinator is no longer making 40 outbound calls a day and instead is handling more complex scheduling exceptions, that’s a meaningful narrative for staff.
Q: AI features are changing fast. How do you help customers decide which updates are worth the operational disruption of rolling out and which ones they can skip?
AT: There’s so much noise in the AI space that health system leaders often don’t know where to focus. As a partner, one of the most valuable things we can do is to help customers cut through the noise and make confident decisions about where to spend their energy.
We start with the desired outcome, not with the functionality or capability. Before discussing features or tools, we anchor conversations in the outcomes the organization is trying to address. It could be no-show rates, staff capacity or patient response rates. We tie capabilities into those priorities.
We also build confidence through proof points. The most effective path to broader adoption is creating visibility into successes within a single department.
Lastly, we show customers what’s possible using what they already have. Often, health systems look outward for new solutions before they fully activate what’s already in their hands. Our job is to help close that gap and maximize the impact.
Q: When it comes to proving the value of AI, which metrics have lost their persuasive power with health system executives? What’s replacing them?
AT: There’s a massive shift away from measuring AI adoption to measuring AI operationalization. Broad adoption statistics and activity metrics, like numbers of clicks, chatbot interactions or messages sent, aren’t persuasive.
What executives really want are operational and financial outcome metrics tied directly to their priorities. These include metrics that demonstrate how AI is improving access, reducing labor dependency, decreasing denial rates or protecting margins.
Q: What does a mature AI customer look like? What are they doing that newer adopters aren’t yet?
AT: Mature health systems think about AI in terms of operational capability, rather than a set of tools. While early adopters often focus on pilots and point solutions, more mature organizations don’t think about those things at all. They embed AI into core workflows with clear ownership.
At the Becker's 11th Annual IT + Revenue Cycle Conference: The Future of AI & Digital Health, taking place September 14–17 in Chicago, healthcare executives and digital leaders from across the country will come together to explore how AI, interoperability, cybersecurity, and revenue cycle innovation are transforming care delivery, strengthening financial performance, and driving the next era of digital health. Apply for complimentary registration now.