CVS Spent $69 Billion Trying to Fix Healthcare. Now It's Trying Again.

"If you want a complete picture of your health, you have to manually gather it, then you have to reconcile it."

That is how Tony Ambrozie, CVS Health's chief digital and technology officer, describes the everyday reality of navigating American healthcare.

The consequences of that fragmentation have monetary costs.

A 2025 preprint study of 256 million U.S. adults found that annual healthcare costs ranged from $909 for patients with little fragmentation to $34,956 for those with high fragmentation, a 38-fold difference. More than half of American adults fall into the higher end of that scale.

CVS Health says it has an answer. The company announced this week a strategic partnership with Google Cloud to launch Health100, a free AI-powered platform that will aggregate data across pharmacies, insurers, primary care providers and pharmacy benefit managers, acting as what Ambrozie calls an "always-on personal health care partner."

It is an ambitious idea. It is also a familiar one.

The Weight of History

In 2018, CVS Health completed its $69 billion acquisition of Aetna, with CEO Larry Merlo framing the combined company as "the front door to quality health care."

The idea was to combine a pharmacy giant with a major insurer, add a network of MinuteClinics, and create the integrated health experience that consumers had never been able to find. The execution proved more complicated.

CVS's stock fell roughly 43% in 2024 alone, driven in part by mounting losses in Aetna's Medicare Advantage business. By June 2025, the company had announced a $20 billion technology commitment, a recognition that the acquisition alone had not delivered what Merlo promised.

CVS is not the first company to discover how hard this problem is. Google launched Google Health in 2008 with the goal of giving consumers a unified view of their medical records, but struggled to move beyond tech-savvy early users and announced the shutdown in 2011, closing the service in January 2012.

Patients, it turned out, did not proactively engage with their own health data outside of acute care.

"It is different from selling products, like toothpaste and breakfast cereal," said Robert Field, a professor of health management and policy at Drexel University, after Walmart closed all 51 of its health clinics in 2024. "It requires different kinds of expertise and management."

So What Is Different This Time?

CVS is aware of this history, and its executives argue that several things have changed enough to make Health100 a different proposition.

The most substantive is TEFCA, the Trusted Exchange Framework and Common Agreement, a federal data-sharing infrastructure that did not exist when previous tech companies built their health platforms.

CVS also enters this effort with assets that pure technology companies never had. Its network of roughly 9,000 retail pharmacies and more than 1,000 walk-in and primary care clinics gives Health100 a physical presence in communities across the country. Pharmacists, Ambrozie argues, are already a trusted clinical touchpoint for millions of Americans, a foundation that Google Health could never replicate from a browser.

The company's approach to data sharing is a departure from the walled-garden model that hobbled previous attempts.

"We need to compete with service, not on hoarding data," Ambrozie said, adding that CVS intends to work with anyone who wants to participate, including competitors. CVS says it has already built a working proof of concept connecting its Caremark, Aetna and pharmacy entities.

Still, the landscape offers reasons for caution. Amazon spent $3.9 billion, including debt, to acquire primary care company One Medical in 2023 and has since watched a succession of senior health executives depart. Money and infrastructure have never been the missing ingredients. Even Google, now CVS's technology partner, shut down its own consumer health records platform in 2012 after failing to move beyond early adopters.

Health100 is built on the premise that fragmentation is fundamentally a coordination and data problem, one that agentic AI, federal interoperability frameworks and a nationwide pharmacy network can finally solve.

What remains an open question is whether fragmentation persists because the data is disconnected, or because the financial incentives of American healthcare do not reward connection. The platform launches later this year, and the harder test begins after that.