Schrodinger (NASDAQ:SDGR) executives outlined a strategy focused on expanding adoption of its computational chemistry platform, accelerating a shift to hosted software delivery, and partnering out internal clinical-stage assets during a virtual fireside chat at an annual healthcare forum.
Platform thesis: “replicate experiment” at scale
CEO Ramy Farid said the company’s mission is to build a computational platform that enables life sciences and materials science researchers to “design better molecules more rapidly, more efficiently.” Farid described traditional drug discovery as a trial-and-error process in which researchers synthesize molecules, test them experimentally, and iteratively optimize—a workflow he said is time-consuming and prone to high failure rates.
Farid also said the company’s software is broadly used across the pharmaceutical industry and cited “100%” customer retention as validation of the platform’s impact.
Business structure changes: partner clinical programs, emphasize platform synergies
CFO Rishi Jain said the company has taken steps to “simplify and clarify” its business structure. He said Schrodinger had been running some clinical programs internally but is now seeking to partner those programs. Jain said the clinical-stage programs are advancing “in the hands of our partners,” with Schrodinger retaining downstream milestone payments and royalties.
Farid framed the pivot as a way to reinforce the company’s core advantage: the synergy between licensing software to a wide user base and using it internally in discovery. That internal usage, he said, helps the company learn what works, improve the product in real time, and provide proof points that encourage an industry-wide shift toward computational methods.
He also said the cost and complexity of running clinical trials have increased versus expectations due to changes in the broader landscape, citing factors such as evolving oncology dynamics, higher U.S. trial costs, “Project Optimus,” and the cost advantages of running trials in China. Farid said the company is prioritizing capital toward advancing the platform and focusing therapeutics efforts on discovery rather than clinical execution.
Hosted transition and profitability target
Jain reiterated a three-year goal of achieving adjusted EBITDA profitability, which he said would be supported by growing both software and drug discovery businesses while maintaining expense discipline.
He also detailed a strategic shift toward hosted software contracts rather than on-premise deployments. Jain said hosted contracts represent about 25% of the mix today, with a goal to reach roughly 75% hosted over about three years. He described the shift as common across software companies moving toward hosted/SaaS delivery, but noted it will affect reported revenue in the near term as revenue recognition becomes more ratable.
To help investors track underlying business momentum during the transition, Jain said Schrodinger is emphasizing annual contract value (ACV) as an operating metric. He stated that at the end of 2025, ACV and revenue were “in lockstep,” citing software revenue of $200 million and software ACV of $198 million. Over time, he said, ACV and revenue should converge, but during the conversion period the company expects revenue to decline as contracts move from upfront on-premise recognition (which he characterized as “almost 80%-90%” recognized in the booking quarter) to ratable recognition under hosted arrangements.
Jain added that the transition is “cash flow neutral,” saying cash flow from operations at the end of 2026 would be the same regardless of the accelerated move to hosted, because the change is primarily in revenue timing and deferred revenue accounting. He also said the company has already converted at least one pharma customer from on-prem to hosted and converted one multi-year on-prem deal to hosted ahead of its renewal date.
New products and the Predictive Tox opportunity
Management said product innovation remains central to the growth strategy. Farid highlighted Predictive Tox as a recent release aimed at a major drug discovery challenge: predicting toxicity related to off-target binding. He argued that existing machine learning approaches often require large training datasets, which means toxicity modeling may occur late in a program after years of work and substantial spending—potentially forcing teams back to redesign molecules when issues appear.
Farid said Schrodinger’s approach is physics-based rather than solely machine-learning-driven, enabling use on novel molecules earlier in discovery. He also said the tool provides insight into why a molecule may be toxic by showing structural binding interactions, which he said can help scientists modify compounds as part of multi-parameter optimization.
On product development, Farid said breakthrough innovation does not come solely from asking customers what they want. He said Predictive Tox was motivated by Schrodinger’s own internal projects and collaborations, while customer feedback helped refine the concept. He noted the company has a regular cadence of product work, with “multiple new product releases, or major enhancements to existing products every year,” and said Schrodinger has “four releases every year.” He added that the company discussed Predictive Tox’s beta publicly in part because of an associated grant and heightened FDA attention to computational tox prediction as a way to reduce animal testing.
Agentic AI, throughput-based licensing, and pharma interest
Farid also discussed “agentic AI,” describing it as a way to automate workflows and augment scientists rather than replace them—helping scale expert use of advanced computational tools. He cautioned that building reliable agents is difficult and may take longer than some expect, noting the complexity of drug design.
Jain said the company’s commercial model is positioned to capture incremental demand because most products are licensed on a throughput/utilization basis, with customers buying additional licenses and tokens rather than seats as usage expands. He also said agentic tools could broaden the user base by helping chemists trained on traditional methods adopt computational workflows.
Farid said the broader excitement around AI is increasing interest in computational approaches among pharmaceutical customers. He argued that AI models depend on training data and that experimental data alone is insufficient, requiring simulated data generated by platforms like Schrodinger’s. Farid said he views the increased attention to AI as a tailwind for the company’s business.
About Schrodinger (NASDAQ:SDGR)
Schrödinger, Inc is a life sciences and materials discovery company that specializes in the application of physics-based computational platforms to accelerate drug discovery and advanced materials design. Founded in 1990 by Professor Richard A. Friesner, Schrödinger has developed a suite of proprietary software tools—such as Maestro for molecular modeling, Glide for molecular docking and Jaguar for quantum chemistry calculations—that enable scientists to predict molecular behavior with high accuracy.
