Health systems have spent the last two years building the vocabulary of AI strategy — governance frameworks, steering committees, technology roadmaps. Mark Townsend, MD, chief clinical digital ventures officer at Cincinnati-based Bon Secours Mercy Health, thinks one common term deserves closer scrutiny.
“I often say that I take issue with the term pilot because it denotes with the term that there may be some threshold whereby we’ll accept failure,” Dr. Townsend said during a recording of the “Becker’s Healthcare Podcast” at Becker’s 11th Annual Health IT + Digital Health + RCM Conference. “I like to refer to projects rather than pilots as they path forward.”
The word “pilot” arrives carrying an implicit permission structure, a threshold beneath which a team can walk away and declare the experiment inconclusive. That’s not good enough for most IT executives, and can lead health systems down a disastrous path.
At Bon Secours Mercy Health, every AI initiative launches from a defined problem statement, a clear articulation of the value it will create, and a plan that specifies how that value will be measured and when.
“As we launch them, we launch them fully intending to succeed,” Dr. Townsend said. “We also are pretty intentional about putting together milestones. And so 18 to 24 months into any given project, we then double down and say, okay. Is this project actually meeting the deliverable per the business plan, per the expectations, per the problem statement at the inception of the project.”
That 18-to-24-month review is where the operational discipline lives. If a project is meeting its milestones, it scales. If it is not, the team re-evaluates.
“At that point, we then have a go, no go conversation, and we scale,” Dr. Townsend said. “It’s got to be able to scale such that it is sustainable. So the cost is actually absorbed in the projects, and the value that is being created does set it up for sustainable success.”
The sustainability test matters as much as the accuracy test. A tool that works in a controlled environment but cannot operate at the cost structure of a 47-hospital system is not a solution. Scaling without a viable cost structure doesn’t fix the problem; it inflates it.
Bon Secours Mercy Health runs this same discipline through Accrete Health Partners, its digital venture arm. The system invests in startups only when it already has a clear path to becoming a meaningful client, and the investment relationship is designed to keep operational accountability at the center even after financial valuations move favorably.
“Just because you’re in our ecosystem, just because you’re a partner doesn’t mean that you can take for granted the work that is happening or the work that is being done,” Dr. Townsend said. “Just because the valuation of the company is performing very well relative to the market and returning value even through our investment arm does not mean that you can take for granted the operational engagement. You’ve got to continue to massage that, curate that, and really focus on that relationship with the operational teams.”
