CVS Health x Simile: Simulations for faster, safer decisions

Simulation is evolving from faster surveys and concept tests to a full-stack capability: modeling individuals, then journeys over time, then interactions, and ultimately entire markets. The frontier is no longer predicting what a customer might say at a moment in time, but understanding how people update beliefs, influence one another, and respond to incentives inside designed systems.

The next inning extends beyond end consumers to multi-agent market simulations, where customers, competitors, partners, and policies interact and second-order effects can be stress-tested before capital is committed.

Organizations like CVS Health, which began by validating the individual agent as the quantum unit and scaling stepwise from static to dynamic and multi-agent use cases, are well positioned to lead this shift.

CVS Health has used generative agent simulations to help guide decision-making with the support of Simile, the first company to build and bring this technology to market.

Built on 2.9 million consented responses from more than 400,000 participants across 200+ behavioral scenarios, these agents are modeled on data from real people — including interview responses and past choices — so they act as safe, accurate stand-ins for the individuals they represent.

Over the past year, CVS Health partnered with the Simile team to build a significant simulation capability. That foundation created a persistent, high‑fidelity lens on customers and competitors without repeatedly fielding new studies.

Together, these early stages delivered four tangible advantages:

Faster validation of known insights. Simulations have reproduced conclusions from prior research quickly, letting teams pre‑screen ideas and hypotheses, and reserve fieldwork for the most promising directions.

Sharper NPS and experience drivers. By testing end‑to‑end journeys–including access to pharmacists, wait times, and communication clarity–CVS Health has isolated which levers matter most for satisfaction and where improvements will have the most impact across segments. This proved especially valuable for hard to-reach populations such as patients with chronic conditions who are slow, expensive, or unevenly represented in traditional surveys and panels, yet disproportionately influence health outcomes.

Adherence and behavior change under real constraints. Dynamic agents enabled teams to test reminder cadences, education content, and benefit designs to see which combinations increase intent to refill or adopt clinical services–before running costly pilots. Critically, this made it possible to explore questions involving sensitive health behaviors, privacy-constrained data, and second order effects that are difficult or sensitive to study directly in the wild.

Differentiation and competitive positioning. By benchmarking perceptions against grocery and mass competitors in simulations, CVS Health identified which experience elements most clearly differentiate, and where investment would actually move the needle. These simulated results were then used to prioritize and validate downstream pilots and experiments, tightening the feedback loop between strategy, experimentation, and real world outcomes.

Sign up for a demo here