Theranos: what went wrong, from a decision engine's view
The essentials in 30 seconds
- What went wrong at Theranos can be read in 4 layers: evidence, governance, secrecy, and patient safety.
- The EQOS Asia back-test returns φ(S) = 0.085, a readability of 0.012, and a projected utility of 0.007 (EQOS Asia Theranos case page, as of October 2026).
- The study records $700M raised and a ratio of 19 to 1 between notoriety and transparency.
- The ethical safeguard blocks both investment and deployment.
- Each layer converts into a gate that a team can apply before a commitment.
Theranos is often told as a story about one person. A decision engine reads it as a stack of 4 layers that failed together, and each layer can be checked in a new case. This page uses the EQOS Asia Theranos study, as of October 2026, to explain what went wrong, and it builds on the Theranos warning signals guide.
On this page
Compare the EQOS Asia offers and request a 30-minute pre-diagnostic.
What went wrong at Theranos when a decision engine reads the case?
The EQOS Asia study describes $700M raised on a technology that did not exist, 0 publications in 10 years, and patients who received false medical results (EQOS Asia Theranos case page, as of October 2026). A decision engine splits that outcome into 4 layers: the evidence behind the claim, the governance around it, the openness of the company, and the safety of the people served. Each layer has its own measure.
The split matters because a single story invites a single remedy. When 4 layers fail together, a remedy that fixes one of them leaves the other three in place. Reading the case by layer gives a decision maker a short list of gates, and each gate has a yes or no answer that a colleague can verify.
Where did Theranos governance go wrong in a decision view?
The study lists four board members drawn from diplomacy, the military, and politics, and finds no medical competence among them for a blood-diagnostics company (EQOS Asia Theranos case page, as of October 2026). The engine reads the gap between board expertise and company mission as alignment close to zero, and it scores integrity at 0.03.
A board does two jobs: it opens doors and it challenges claims. The study suggests that this board served the first job. The decision gate that follows is concrete: ask who on the board can read the technical evidence, and require at least 2 independent experts when the product depends on science. The answer takes one email to obtain.
Why did secrecy about the Theranos device distort the decision?
The study records no public demonstration of the device, non-disclosure agreements for everyone, compartmentalised employees, and escorted visitors. It sets the resulting gap at 19 to 1 between notoriety and real transparency, against a normal alert threshold below 3 to 1 (EQOS Asia Theranos case page, as of October 2026).
Secrecy distorts a decision because the reader receives reputation in place of evidence. Investors, partners, and a pharmacy chain acted on the public image. A gate for this layer asks for one independent test of the product under conditions the buyer controls. A company with a working product can usually offer such a test within weeks.
What do the three Theranos outputs say about the decision?
The engine returns three outputs for the case. The score φ(S) = 0.085 measures overall coherence. The readability ratio X(S) = 0.012 measures how much usable information exists compared with the complexity of the claim. The projected utility of 0.007 measures the living value the system can produce (EQOS Asia Theranos case page, as of October 2026).
The three numbers tell a consistent story. A low score, a near-zero readability, and a projected utility close to zero point in one direction. The method places its STOP zone below 0.618 (EQOS Asia method page, as of October 2026), so a result of 0.085 sits far inside it. The scale in the figure shows the distance.
Why did the ethical safeguard double-block the Theranos decision?
The safeguard asks whether a system harms the people it claims to serve. The study documents false results that put patients at risk, scores care at 0.04, and concludes that the system is in ethical failure (EQOS Asia Theranos case page, as of October 2026). The result is a double block, one on investment and one on deployment.
A double block differs from a low score. A low score invites repair, while a block stops the decision until the condition is met. For a manager the practical meaning is a clear order of questions: confirm that the product does no harm first, and discuss returns second. That order keeps a strong growth story from outranking a safety question.
What did the Theranos failure cost, according to the study?
The study lists $700M raised, $9B of value destroyed, and patients who received false HIV and cancer results. It records prison sentences of 11 years for the founder and 13 years for the former president (EQOS Asia Theranos case page, as of October 2026). Courts determined those outcomes, and the page reports them as part of the toll.
The comparison the study draws is between that toll and the cost of a structural diagnostic. The cost of the checks was small and the cost of the failure was large. For a reader the useful part is the ratio: a few days of verification against years of exposure. The pricing page lists the current offers, with their dates.
How does a Theranos review change your own decision process?
Turn each layer into a gate. For evidence, ask for one independent publication or test. For governance, ask for 2 independent experts. For openness, ask for a demonstration the buyer controls. For safety, ask for the regulator’s approval before any user relies on the product. Four gates fit on one page and can be reviewed in a single meeting.
Then record the answers and score them. The free simulator estimates φ(S) from ratings on 9 dimensions (EQOS Asia simulator page, as of October 2026). The Theranos study is a retrospective exercise on public data. This page is general information, not legal, tax or investment advice.
See also in this guide
Frequently asked questions
What went wrong at Theranos in one sentence?
Four layers failed together: scientific evidence was missing, the board lacked medical expertise, the company stayed closed to independent testing, and patients received false results. The EQOS Asia study scores each layer and shows that a gate existed for every one (EQOS Asia Theranos case page, as of October 2026).
What does a projected utility of 0.007 mean?
Projected utility measures the living value a system can produce over time. A value of 0.007 is close to zero, which the study reads as a system that generates almost no viable long-term value. The figure comes from the proprietary engine, and its weighting is not published.
Can these gates apply outside medicine?
Yes. The four gates ask for independent evidence, independent experts, a buyer-controlled test, and regulator approval where one exists. A supplier choice, a partnership, or an acquisition target can pass through the same list, and the answers are easy to document and share.
Where can a reader learn how the score is built?
The method page explains the 3 pillars, the 9 dimensions, and the deterministic core, and states that the full equation is proprietary and shared under NDA (EQOS Asia method page, as of October 2026). The simulator gives a first score on a real case.
Test your knowledge
1. How many layers does this page use to read the Theranos failure?
- 2
- 4
- 9
Show the answer
4.
2. What readability ratio does the back-test record?
- 0.012
- 0.085
- 0.618
Show the answer
0.012.
3. What gap does the study compute between notoriety and transparency?
- 3 to 1
- 9 to 1
- 19 to 1
Show the answer
19 to 1.
4. How many independent experts does the governance gate request?
- At least 2
- Exactly 9
- At least 40
Show the answer
At least 2.
5. Which question comes first after a double block?
- How large are the returns
- Does the product do harm
- Who sits on the board
Show the answer
Does the product do harm.
About the author
Stéphane Jambu
President, EQOS Forecast Asia
Stéphane Jambu leads EQOS Forecast Asia and works as a consultant in e-commerce, search visibility, and AI. LinkedIn profile
Sources: EQOS Asia Theranos back-test page (eqos-asia.com/case-studies/theranos.html), method page (eqos-asia.com/method.html), pricing page (eqos-asia.com/pricing.html), and simulator page (eqos-asia.com/simulator.html), consulted on 22 October 2026. Page updated on 22 October 2026.