Theranos

Elizabeth Holmes founded the company in 2003 at age 19, shortly after dropping out of Stanford's chemical engineering program, initially under the name Real-Time Cures before renaming it Theranos. Her pitch was simple to state and enormously hard to build: a device that could run comprehensive panels of laboratory blood tests — for cholesterol, cancer markers, infectious disease, and more — from just a few drops of blood collected with a finger prick, instead of the multiple vials a conventional venous draw requires. If it worked, it promised cheaper, faster, less painful diagnostics that could be run in a retail pharmacy rather than a centralized lab.

That promise was compelling enough, delivered with enough confidence, that Holmes raised early funding from prominent investors including media mogul Rupert Murdoch and members of the DeVos and Walton families years before the technology was independently verified by anyone outside the company.

As Theranos grew, it adopted a culture of extreme internal secrecy uncommon even by Silicon Valley standards: employees worked in siloed teams with limited visibility into other departments, departing staff signed aggressive non-disclosure agreements, and the underlying lab science was treated as a closely guarded trade secret rather than something to be published or peer-reviewed. Rather than assembling a board with deep expertise in clinical laboratory science or diagnostics, Holmes recruited an unusually prestigious board of statesmen — including former Secretaries of State Henry Kissinger and George Shultz, and former Secretary of Defense James Mattis — whose gravitas lent Theranos credibility with investors, regulators, and the press, even though almost none of them were positioned to evaluate whether the blood-testing technology actually worked.

This combination — a famous, telegenic founder frequently compared to Steve Jobs in profile pieces, a board of statesmen rather than scientists, and a company culture built around concealment — meant that for over a decade, almost no one outside Theranos's own walls had the access or expertise needed to check whether its central technical claim was true.

The company's proprietary device, known internally as "Edison," could in fact run some blood tests from small sample volumes. What it could not do was reliably run the full menu of more than 200 tests Theranos marketed to patients and physicians. According to reporting and subsequent legal proceedings, for the large majority of those tests, Theranos secretly diluted the small finger-prick blood samples and ran them on modified, commercially available lab analyzers from vendors like Siemens — the exact kind of conventional lab equipment the company had built its entire brand around replacing — while representing to patients, doctors, and investors that the results came from its own proprietary technology.

This wasn't a private R&D shortfall corrected before shipping. Real patients received real medical test results generated by an unvalidated, undisclosed process, and some of those results were later found to be inaccurate in ways that could plausibly have affected real clinical decisions — misdiagnosed thyroid conditions, missed indicators of miscarriage risk, and unreliable results for tests as consequential as those used to guide blood-thinning medication.

In 2013, Theranos signed a partnership with Walgreens to open "Theranos Wellness Centers" inside its retail pharmacies, starting in Palo Alto and expanding to dozens of locations across Arizona and California. This was the moment the company's unproven technology stopped being an internal R&D question and started directly touching real patients making real healthcare decisions based on the results they received. Walgreens, a large and reputable retail partner, effectively vouched for a product it had not independently verified — a partnership Holmes had helped secure in part by demonstrating the devices in ways that later testimony suggested did not reflect how the tests were actually being run for real patients.

Across its life, Theranos raised roughly $945 million, pushing its valuation to a peak of about $9 billion by 2014 — briefly making Holmes, on paper, one of the wealthiest self-made women in the country. The investor roster reads like a cross-section of American wealth and influence rather than typical healthcare or biotech specialists: the family of Betsy DeVos and the Cox family each invested around $100 million, the Walton family (of Walmart) invested $150 million, Rupert Murdoch put in $125 million, and Mexican billionaire Carlos Slim invested $30 million, among many others. Very few of these investors had the specialized clinical-diagnostics expertise to independently evaluate Theranos's technical claims, and the company's culture of secrecy meant even sophisticated investors were often kept from the kind of technical due diligence that would normally accompany a bet of this size on unproven medical technology.

The company's account of itself began to unravel in October 2015, when Wall Street Journal reporter John Carreyrou published the first of a series of investigative articles reporting that Theranos ran the vast majority of its tests on modified conventional machines rather than its own proprietary devices, and that the accuracy of its results was in serious doubt. Carreyrou's reporting relied heavily on internal whistleblowers, most notably Tyler Shultz — grandson of board member George Shultz — and lab associate Erika Cheung, both of whom had raised internal concerns about test accuracy and been rebuffed or threatened by the company before going to the press and regulators.

The Journal's reporting triggered a cascade of regulatory scrutiny that the company could not survive. In 2016, the Centers for Medicare and Medicaid Services inspected Theranos's Newark, California lab, found deficiencies serious enough to pose "immediate jeopardy to patient health and safety," and ultimately revoked its CLIA certificate — the certification required to legally operate a clinical laboratory — while also banning Holmes from owning or operating a lab for two years. Walgreens terminated its partnership in June 2016 and later sued Theranos over the relationship. In March 2018, the SEC charged Holmes and Balwani with massive fraud, and in June 2018 the U.S. Attorney's office filed criminal wire fraud and conspiracy charges against both. Theranos, facing a cash crisis and unable to find a buyer despite hiring an investment bank to run a sale process, laid off most of its remaining staff and shrank from a peak of roughly 800 employees to a skeleton crew.

Unable to secure new financing or a buyer and in default to its creditor Fortress Investment Group, Theranos informed shareholders in September 2018 that it would formally dissolve, assigning its patents to Fortress in partial satisfaction of a $65 million debt and distributing what little cash remained to unsecured creditors — an unusually quiet ending for a company that had once been valued at $9 billion. Holmes and Balwani were tried separately; in January 2022, a jury convicted Holmes on four of eleven federal fraud and conspiracy counts related to defrauding investors (she was acquitted on the counts related to defrauding patients). In November 2022 she was sentenced to just over 11 years in federal prison. Balwani was convicted on all charges against him, including counts related to defrauding both investors and patients, and received a longer sentence.

Theranos is a different category of failure from most startups in this database: it did not fail because the market rejected a real product, or because unit economics didn't work, or because a competitor out-executed it. It failed because the core product claim was never true, and the company spent a decade concealing that fact behind secrecy, a prestigious but technically unqualified board, and an extraordinarily persuasive founder narrative. The case is a reminder that investor prestige and board credentials are not a substitute for independent technical verification, especially in a regulated field like clinical diagnostics where a product's failure mode is not "customers don't want it" but "patients receive wrong answers about their own health." It's also a reminder that internal whistleblowers and independent journalism, not the company's own investors or board, were what actually surfaced the fraud — the checks that were supposed to catch this earlier didn't.