Theranos 2015: When the Board Didn't Ask
Situation
It is October 2015. Theranos is valued at $9 billion. Elizabeth Holmes graces the covers of Fortune, Forbes, and Time. She is compared to Steve Jobs — she wears the black turtleneck deliberately. The company claims its Edison device can run 240+ blood tests from a single fingertip prick, at a fraction of the cost of conventional lab tests.
Walgreens has deployed Theranos wellness centers in 40+ stores across Arizona and California. Safeway has spent $350 million building out in-store clinics. The US military has discussed using Theranos devices in the field. The board includes a former US Secretary of State, a former Secretary of Defense, two former US Senators, and the former chairs of Bechtel and Wells Fargo. The board has zero members with a background in diagnostics, laboratory science, or clinical medicine.
On October 15, 2015, the Wall Street Journal publishes John Carreyrou's first story: the Edison device cannot actually run most of the tests it claims. For the vast majority of its test menu, Theranos is secretly using commercial Siemens analyzers — the same machines used by conventional labs — and diluting blood samples, which introduces errors.
The company's core technical claim is false.
Holmes's response: she goes on CNBC that afternoon and calls the story "factually and scientifically wrong." She does not demonstrate the device. She does not release clinical validation data. She attacks the reporter.
The decision moment
It is October 16, 2015 — the morning after Carreyrou's story and Holmes's CNBC rebuttal. You are a board member. You have just received Holmes's prepared statement ("categorically untrue") and a request to appear on a joint board call to discuss the company's media response.
You have served on this board for three years. You have never been shown the Edison device running a live test. You have never reviewed a peer-reviewed clinical validation study. You have never met with the laboratory director. You have seen the Walgreens contract, the Safeway revenue projections, and Sequoia's term sheet — all of which you found reassuring.
Three decisions face you in the next 24 hours:
- How do you respond in the board call? Do you demand an independent technical audit before making any public statement? Do you sign onto the "categorically untrue" narrative? Do you ask to see the clinical validation data before the call?
- What is your duty of care at this moment? Theranos has revenue contracts with Walgreens. Real patients are receiving blood test results. If those results are wrong, there are medical consequences.
- Is Holmes's refusal to demonstrate the device in public a red flag you should have caught earlier? If so, what should you have done differently as a board member over the past three years?
You are a Theranos board member.
Key financial datapoints (for reference)
| Metric | Value (2015) |
|---|---|
| Company valuation | $9 billion |
| Holmes's estimated net worth | $4.5 billion (Forbes peak) |
| Total funding raised | ~$700 million |
| Walgreens store deployments | 40+ |
| Safeway clinic investment | ~$350 million |
| Board members with medical/lab credentials | 0 of 11 |
| Peer-reviewed publications of Edison technology | 0 |
| Claimed test range (from fingertip prick) | 240+ assays |
| Actual tests run on Edison at launch | ~12 assays |
| Revenue reported to investors | Not disclosed (claims of $100M+ were false) |
Frameworks invoked
- Corporate Governance / Duty of Care. Directors have a legal duty to act in the best interest of the company and its stakeholders, which requires informed decision-making. A board that never demanded a product demonstration is a board that failed its duty.
- Stakeholder Theory. Holmes framed every board interaction around investor and partner outcomes. But the actual stakeholders — patients relying on blood test accuracy — were never represented in the boardroom.
- Asymmetric Information. Holmes controlled all technical information. The board had no independent technical advisors. Every piece of data they received passed through Holmes's filter.
- Principal-Agent Problem. The board (principals) hired Holmes (agent) and gave her nearly unchecked authority. Without independent verification mechanisms, agents with misaligned incentives can sustain fraud indefinitely.
Discussion questions
- The board included Henry Kissinger, George Shultz, and James Mattis — but no oncologists, no laboratory scientists. Why did prestige substitute for expertise, and whose responsibility was that?
- Theranos had never published a peer-reviewed validation study. At what point should that have been a disqualifying fact rather than a negotiating point?
- Holmes deployed a specific strategy: surround the board with prestigious non-technical members who would be socially reluctant to challenge her. How do you design board composition to prevent this?
- Walgreens' own medical consultant raised red flags before signing the contract and was overruled. What does this tell us about how organizations process inconvenient information?
- Holmes's defense ("we're being held to a different standard as a Silicon Valley company") was a category error: FDA laboratory standards are not negotiable because they protect patients. How do you identify when "we're disrupting regulations" is a legitimate thesis vs. a fraud shield?
The real outcome (revealed at session end)
October 2015: CMS (the federal lab regulator) begins its inspection of Theranos's Newark, CA lab. Holmes's CNBC defense holds for six months.
July 2016: CMS bans Holmes from running a clinical laboratory for two years and revokes Theranos's lab certification. Walgreens terminates the contract.
2018: The SEC charges Holmes with fraud. The DOJ charges Holmes and Balwani with wire fraud. Holmes is removed as CEO.
2022: Holmes is convicted on four counts of fraud and conspiracy. She is sentenced to 11.25 years in federal prison. The jury found she had defrauded investors but acquitted her on patient fraud counts (a verdict many legal analysts found perplexing).
The board's role: Not one Theranos board member was charged with a crime. George Shultz publicly stated he was deceived. His grandson, Tyler Shultz, was an early Theranos employee who became a whistleblower — the family rift became one of the most human dimensions of the story.
The lesson: Board composition is a risk management decision. Prestige and "pattern recognition" (she looks like Steve Jobs) are not substitutes for technical due diligence. A board with no one who can evaluate the core product claim is not a board — it's a credentialing service for a narrative.
Sources
- John Carreyrou, Bad Blood: Secrets and Lies in a Silicon Valley Startup (2018).
- SEC Complaint v. Theranos, Inc. and Elizabeth Holmes (2018).
- US v. Holmes (N.D. Cal., 2022) trial record.
- CMS Survey Report, Theranos Clinical Laboratory (2016).
- HBS case: "Theranos: Blood, Deception, and the Birth of a Biotech Unicorn" (2018).