Bad Blood | Shortform

1-Page Summary

Bad Blood: Secrets and Lies in a Silicon Valley Startup is written by John Carreyrou, the very Wall Street Journal reporter who first exposed Theranos. It covers all known history of how Theranos started, maintained its lies, and fell.

The story is an incredible demonstration of the weaknesses of human psychology. Even very sophisticated investors, whose job is to sniff out these exact situations, fell completely for the fraud until it was exposed.

Beyond just detailing the chronology of the story, our Bad Blood summary covers three major questions:

How did the deception begin?
How was the deception allowed to continue?
What finally led to Theranos’s downfall?

We’ll focus on the themes of why so many sane, professional people were completely taken in and ignored the warning signs until it was too late.

How Did the Deception Begin?

It has to start with Elizabeth Holmes, who dreamed big. She clearly had large ambitions. When asked what she wanted to be at 10 year old, she answered “a billionaire.”

Theranos had a vision people wanted to believe in - “detect diseases early so no one has to die unnecessarily.” All types of stakeholders saw what they wanted to see in it, allowing large suspension of disbelief.

Investors saw a huge financial opportunity. The medical testing market is huge - allowing patients to test at home, and using the resulting data to inform medical decisions, makes it even bigger.
Patients and their families saw better decisionmaking and less pain.
The public saw a thrilling female founder who could be the next Steve Jobs. They wanted to celebrate a brilliant female innovator in an age of female empowerment.
Partners like Walgreens and Safeway saw a way to compete against competitors like CVS, and to rejuvenate their financials. So they partnered with Theranos to roll out

Elizabeth Holmes practiced charismatic techniques to win over supporters.

Elizabeth spoke sincerely and enthusiastically about the mission. It gave the impression that there was no way someone with this sincerity could be beguiling.

Stakeholders used pattern matching to jump to wrong conclusions.

On paper, the technology fit the pattern of disruption - dramatically lower cost replacing the dinosaur incumbents, leading to massive adoption and market expansion. The story was intoxicating and irresistible.
Holmes looked like the stereotypical genius dropout founder. She even deliberately cast a resemblance to Steve Jobs with black turtlenecks.

How Did the Deception Continue?

To hide the fact that their proprietary machine that didn’t work, Theranos crossed the limits of scientific legitimacy. These were non-standard distortions of scientific practice.

In coefficient of variation studies (a measure of precision and repeatability), deviant results were repeated until they got satisfactory results - basically retaking a test until you get the right answers. Furthermore, they used only the median values of 6 replicates, guaranteeing a tighter CV.
In quality control checks, outliers were inappropriately thrown out - basically changing the answer to get what you want.

When scientific manipulation wasn’t enough, Theranos actively deceived people and lied by omission.

To press and investors, Elizabeth embellished claims level of accuracy and number of assays that could be run on their proprietary machine.
Theranos was scheduled to run demonstrations for pharmaceutical company partners. When their devices didn’t work live, they ran fake demonstrations instead.

When company insiders doubted the company’s strategy, Theranos actively quashed dissent.

Non-believers were quickly fired for being disruptive, then bound by an NDA so they couldn’t leak details to press. Instead, sycophants and order-followers were promoted.
Former employees who were suspected of leaking bad information were threatened legally and monitored by private investigators.

Mental biases like fear of missing out and sunk cost fallacy prevented partners from looking deeper.

Partner company Walgreens couldn’t pull out despite mounting doubts about Theranos. If the innovation were real, it couldn’t risk CVS taking it. Furthermore, it had already shouldered the money of the building out new clinics, and once it got deep enough, pulling out would have been intensely embarrassing.
Because of all the hype, the investing rounds in Theranos were hot. Therefore, investors didn’t have the time to pause for deep due diligence. If they did express doubt and ask for more time, they’d likely be excluded from the deal.

How Did the Deception End?

Theranos overreached in ambition, causing them to require delivering at some point.

They promised too much to Walgreens and had to deliver. This prompted them to cheat.

Theranos stepped too far down the slippery slope of deception, until even their strong-arm tactics couldn’t hold back a conscience.

Embellishing results in prototyping phase was somewhat acceptable. Lying about real patient data, which affected clinical decision making, crossed the line for many employees.
Many employees left; most stayed quiet due to fears of reprisal; some brave whistleblowers let conscience override fear, and they reported to regulators and reporters.

A journalist fought for the story.

Despite continuing threats of litigation, harassment of his sources, and personal surveillance by private investigators, John Carreyrou pressed forward with his story, knowing it would be big. Arguably, without his story, Theranos would have been allowed to persist for years longer, harming many more patients.

In the downfall, momentum and vicious cycles are as punishing as they are powerful on the way up.

Once the evidence of Theranos’s failure became clear, the flywheel effect worked in the opposite direction. Walgreens dropped the partnership; regulators fixated their attention on Theranos; the public lambasted her and the gullibility of Theranos’s supporters; sources readily reached out to reporters to share useful information; investors sued Holmes for fraud.

Part 1: How Did the Deception Begin?

It has to start with Elizabeth Holmes, who dreamed big.

She clearly had large ambitions. When asked what she wanted to be at 10 year old, she answered “a billionaire.”
When her dad suggested she get a PhD, she declined. Instead, she said, “I want to make money.”
It might be reasonable to assume Holmes started with good intentions. After a summer in Asia during the SARS period, she researched the patent literature and proposed a way to test blood with small amounts of blood. Her adviser was impressed and encouraged her to start a company.
But this same ambition brought her down a slippery slope of deception.

Investors saw a huge financial opportunity. The medical testing market is huge - allowing patients to test at home, and using the resulting data to inform medical decisions, makes it even bigger.
Patients and their families saw better decisionmaking and less pain.
The public saw a thrilling female founder who could be the next Steve Jobs. They wanted to celebrate a brilliant female innovator in an age of female empowerment.
Partners like Walgreens and Safeway saw a way to compete against competitors like CVS, and to rejuvenate their financials. So they partnered with Theranos to roll out blood-testing clinics at their locations.
Senior mentors like George Schultz, General Mattis saw a granddaughter-like figure to mentor. So they lent their name to Theranos as advisors.
Anyone with equity (employees, investors, and advisory board included) had a huge incentive to believe the company would work.

Elizabeth Holmes practiced charismatic techniques to win over supporters.

Elizabeth spoke sincerely and enthusiastically about the mission. It gave the impression that there was no way someone with this sincerity could be beguiling.
To the Walgreens CFO, Elizabeth gave a gift of an American flag “flown over Afghanistan” with a dedication to Walgreens written on it.
Many people repeated: “she had this intense way of looking at you while she spoke that made you believe in her and want to follow her.” Here’s an example of Holmes speaking in a Theranos promotional video.
She apparently affected her voice to be a strikingly low baritone, possibly convinced it would lead to better results in a male-dominated world.

Stakeholders used pattern matching to jump to wrong conclusions.

On paper, the technology fit the pattern of disruption - dramatically lower cost replacing the dinosaur incumbents, leading to massive adoption and market expansion. The story was intoxicating and irresistible.
Holmes looked like the stereotypical genius dropout founder. She even deliberately cast a resemblance to Steve Jobs with black turtlenecks.
They crafted a cohesive narrative. From childhood, Elizabeth had a phobia of needles that led to the breakthrough technology.

Part 2: How Did the Deception Continue?

It’s common for early stage startups to raise money without a product. But typically, when they don’t show meaningful progress, their investors back out and they shut down.

How could Theranos continue its operations for over a decade, when its product did virtually nothing it claimed?

In summary, it was a combination of active deception by Theranos management, combined with psychological biases preventing outsiders from pushing further for the truth.

To hide the fact that their proprietary machine that didn’t work, Theranos crossed the limits of scientific legitimacy. These were non-standard distortions of scientific practice.

Theranos’s pitch was they could diagnose diseases on far less blood than normal blood draws. But it didn’t actually work. To process microliters of blood on standard third-party machines, they had to dilute the blood sample to have enough sample to read. However, this lowered the analyte concentrations below what the machines were qualified to handle, below their sensitivity and causing distorted results.
In coefficient of variation studies (a measure of precision and repeatability), deviant results were repeated until they got satisfactory results - basically retaking a test until you get the right answers. Furthermore, they used only the median values of 6 replicates, guaranteeing a tighter CV.
In quality control checks, outliers were inappropriately thrown out - basically changing the answer to get what you want.
In sensitivity tests for syphilis, results that fell into an “equivocal zone” weren’t included in the calculations, letting Theranos widen the zone until sensitivity reached whatever number it wanted.
When running proficiency testing for clinical lab accreditation, they ran the samples on third-party analyzers instead of their proprietary Edison machines. In other words, they were marketing the importance of their Edison machines to the public, while they actually ran samples on third-party machines. Their distorted logic to justify was that Edison’s technology was unique and had no peer group, so its results couldn’t be compared to the accrediting body’s machines.

When scientific manipulation wasn’t enough, Theranos actively deceived people and lied by omission.

To press and investors, Elizabeth embellished claims level of accuracy and number of assays that could be run on their proprietary machine.
- In truth, its accuracy was poor and it could run much fewer tests than claimed. They relied on third-party devices to run most of their assays, meaning Theranos was far less innovative than believed.
Then, when regulators came to audit their operations, Theranos employees blocked investigators from accessing the room containing their new machines, misleading them into believing their third-party machines were the only ones operating. In fact, their faulty proprietary machines were inappropriately processing samples, when they weren’t improved to do so.
Theranos was scheduled to run demonstrations for pharmaceutical company partners. When their devices didn’t work live, they ran fake demonstrations instead.
Requests from a Walgreens employee to run validation tests on their machines were dismissed.
- The senior Walgreens management didn’t push back, because once they were committed to building out Theranos facilities, they desperately needed Theranos to work.
Revenue projections to investors were vastly embellished by >5x.
Outsiders had the impression that Theranos was flying high - that in a military partnership, Theranos devices were in the back of Humvees; that Theranos had plenty of secure partnerships with big name firms; that Theranos was cash flow positive. Naturally, Theranos didn’t care to correct them

When company insiders doubted the company’s strategy, Theranos actively quashed dissent.

Elizabeth said to employees, multiple times: “This is the most important thing humanity has ever built. If you don’t believe this is the case, you should leave now.”
Non-believers were quickly fired for being disruptive, then bound by an NDA so they couldn’t leak details to press. Instead, sycophants and order-followers were promoted.
Theranos retained a powerful legal team that dissuaded dissent, and made a public show of punishing former employees. For unhappy employees, it was easier not to fight the moral fight and just leave quietly.
Former employees who were suspected of leaking bad information were threatened legally and monitored by private investigators.
People in partnered companies with negative opinions (like a technology auditor at Walgreens) were removed from update calls for being disruptive to progress, further insulating executives at the partnering company from the truth.

Theranos actively kept information opaque, preventing the flow of information.

Teams at Theranos were siloed in the name of security. Communication was stifled, and internet access was monitored.
New devices and lab spaces were kept under heavy security. Many people never had a chance to look at the device’s innards.
Departing employees were bound by NDAs. Visitors to the building were bound by NDAs.
Only the best news were shared with team. Negative results were stifled.

Theranos explained away odd practices within the company.

The heightened security and siloing were justified by the paranoia that Quest Diagnostics and Laboratory Corporation of America were actively undermining Theranos; by their supposed military work; and suggested that there was in fact something valuable to protect.
Any negative press could be blamed on the incumbents trying to block their disruptor.
People easily form their own conclusions, especially in support of their preexisting beliefs.

With all the above techniques suppressing doubt from leaking, Elizabeth continued her charisma and PR campaigns.

In 2008, the board decided she was in over her head and wanted to replace her as CEO. She gave mea culpas, saying she promised to change and persuading them to step back. This never seemed to come up again.
Her strength of conviction and sincerity likely made many people overlook shortcuts in pursuit of the larger goal.

Mental biases like fear of missing out and sunk cost fallacy prevented partners from looking deeper.

Partner company Walgreens couldn’t pull out despite mounting doubts about Theranos. If the innovation were real, it couldn’t risk CVS taking it. Furthermore, it had already shouldered the money of the building out new clinics, and once it got deep enough, pulling out would have been intensely embarrassing.
Because of all the hype, the investing rounds in Theranos were hot. Therefore, investors didn’t have the time to pause for deep due diligence. If they did express doubt and ask for more time, they’d likely be excluded from the deal.

Once something gains enough momentum, virtuous cycles can perpetuate the fraud for a long time.

Once you have a critical mass of investors, press, and big-name supporters, it becomes easier to attract more of all, leading to further entrenchment of opinion
- First her professor Channing Robertson vouched for her, which got her early credibility. Then she got investor Donald Lucas on board; then she got statesman George Schultz on board. Each of these people opened doors to their own supporters, who heavily trusted their judgment. “If person X likes Theranos, then they’ve surely done their research, and I can trust their judgment.”
- Likewise, Rupert Murdoch invested big money after seeing Cox Enterprises and Walmart’s Waltons invest as well.

(Shortform note: To reject Theranos as a fraud meant rejecting every pillar of strength - 1) the investors, partners like Walgreens, AND its advisory board must not have done due diligence; 2) all the gushing press writers must have been hoodwinked; 3) government regulators must be ignoring the deception; 4) the data shown would have to be lies; 5) Elizabeth Holmes would have to be a psychopath

It becomes easier to believe the naysayers must be wrong than to reject this entire monumental facade.
What is less obvious to the person being deceived is that each card in the house of cards is built on each other; each pillar is not an independent vote of truth, but strongly dependent on what came before)

Part 3: How Did the Deception End?

Theranos overreached in ambition, causing them to require delivering at some point.

They promised too much to Walgreens and had to deliver. This prompted them to cheat.
- Why not wait until the miniLab was ready? Elizabeth Holmes: “When I promise something to a customer, I deliver.”
- Thus, Theranos continued lying to the public, pretending it had a working device and giving real lab results in Walgreens centers, when really the results were flawed.
Elizabeth wanted to achieve the vision of a small device that did everything (immunoassays, general chems, hematology, DNA). The form factor came before the underlying technology worked, and ultimately they couldn’t design something that had a chance of working.

Theranos stepped too far down the slippery slope of deception, until even their strong-arm tactics couldn’t hold back a conscience.

Embellishing results in prototyping phase was somewhat acceptable. Lying about real patient data, which affected clinical decision making, crossed the line for many employees.
- Many employees left; most stayed quiet due to fears of reprisal; some brave whistleblowers let conscience override fear, and they reported to regulators and reporters.
Doctors were increasingly furious about incorrect lab reports leading to inappropriate patient treatment. They cooperated with the reporting.

A journalist fought for the story.

Despite continuing threats of litigation, harassment of his sources, and personal surveillance by private investigators, John Carreyrou pressed forward with his story, knowing it would be big. Arguably, without his story, Theranos would have been allowed to persist for years longer, harming many more patients.

In the downfall, momentum and vicious cycles are as punishing as they are powerful on the way up.

Once the evidence of Theranos’s failure became clear, the flywheel effect worked in the opposite direction. Walgreens dropped the partnership; regulators fixated their attention on Theranos; the public lambasted her and the gullibility of Theranos’s supporters; sources readily reached out to reporters to share useful information; investors sued Holmes for fraud.

Shortform Conclusion: Applying Psychological Biases

Warren Buffett’s partner Charlie Munger has a list of 25 cognitive biases to be aware of in decisionmaking. As he says, when multiple biases are in play simultaneously, a “lollapalooza” happens leading to extremely distorted results. This is as true in Theranos as it is in cults.

Here’s a rundown of major biases that played into perpetuating Theranos’s fraud:

Reward tendency - anyone who had a financial stake in Theranos clearly wanted it to succeed. This included its advisory board, investors, and employees.
Liking tendency - Elizabeth Holmes was charismatic and seemed genuine. People wanted to follow her and execute her vision.
Disliking tendency - Its supporters disliked incumbents like Quest Diagnostics and wanted to depose them.
Influence from mere association - Holmes surrounded herself with powerful people, starting with Channing Robertson and going up to Rupert Murdoch and Henry Kissinger. Their prestige rubbed off on Theranos.
Social proof tendency - Likewise, seeing other people support Theranos lowered the defenses of new supporters. How could so many people be wrong?
Overoptimism tendency - Holmes and Theranos supporters were overoptimistic about the technology, underestimating the amount of work needed to execute their vision.
Availability misweighting tendency - You use what little data you have available. For Theranos, little technical data was released, meaning supporters had to draw only on the positive press that Holmes carefully cultivated.
Inconsistency avoidance tendency - Once people publicly voted for Theranos, it was hard to backpedal. If they were wrong, they’d look foolish.
Deprival superreaction tendency - Investors and partners couldn’t imagine losing their entire investment if Theranos were fraudulent. So they avoided the painful truth and doubled down, hoping for the best.
Stress influence tendency - Employees who had a crisis of conscience were threatened by Theranos’s legal team, preventing them from assessing the situation fully and choosing to adopt the less risky path of keeping quiet.

All these influences built on each other, preventing the fraud from unraveling for years, until the accumulation of fraud became too much.