The healthcare AI sector is currently experiencing a reality check, as investors are now looking for startups to demonstrate clear financial and operational benefits. During last week’s ViVE conference in Los Angeles, several investors shared their thoughts on the return on investment (ROI) for healthcare AI solutions.
Larry Cohen, CEO of Health2047, the venture studio of the American Medical Association, highlighted that the strained finances of health systems mean AI acquisitions increasingly require the approval of Chief Financial Officers (CFOs). This shift necessitates a robust justification for financial returns. "There's a notable change in how products are being pitched; the conversation has moved beyond just ease of use and satisfaction towards illustrating financial benefits," Cohen stated. He emphasized that AI solutions must now demonstrate their capacity to be self-sustaining financially.
Uma Veerappan, a vice president at Flare Capital Partners, pointed out that many healthcare AI startups are still in the early stages of validating their financial returns, especially for tools designed for clinical applications. In her view, the initial focus may need to be on user adoption rather than immediate ROI. “Given the nascent stage of many of these companies, it's not critical for ROI to materialize within the first six months. What's pivotal is showcasing product stickiness. Over a longer timeline, financial and clinical ROI will become essential benchmarks,” Veerappan noted.
Rachel Feinman, managing director at Tampa General Hospital Ventures, explained that ROI can vary greatly depending on the type of AI tool in question. For instance, tools like ambient clinical documentation scribe systems tend to provide indirect ROI; healthcare organizations may invest in them to enhance their appeal to prospective physicians, even if they don’t contribute to immediate revenue increases. "Healthcare systems are beginning to promote their use of ambient listening technology for providers, making it a top consideration for recruitment," Feinman said. On the other hand, AI systems designed to automate administrative tasks—such as revenue cycle management and data processing—often deliver clearer and more rapid financial returns.
Vig Chandramouli, a partner at Oak HC/FT, observed a trend where more AI startups are moving away from conventional SaaS pricing models toward outcome-based pricing, such as charges per transaction or completed task. This approach allows clients to feel they are remunerating solely for measurable outcomes. "Businesses are being compelled to adopt pricing that reflects three to five times ROI. Value capture is central to nearly every business pitch," he remarked. However, Chandramouli cautioned that outcome-based pricing carries hidden risks, particularly if a startup miscalculates the costs associated with achieving successful outcomes, which could undermine unit economics. He stressed the importance for startups to meticulously define contract terms to clarify what constitutes a successful action, distinguish between fixed and variable fees, and assess the realism of projected revenues.