Bad Blood Reveals Theranos Was Just an Extreme Startup Story

I read Bad Blood.

Bad Blood: Secrets and Lies in a Silicon Valley Startup

This is the book by the Wall Street Journal reporter who exposed the fact that Theranos, once one of the largest unicorn companies in the world, was in reality a company steeped in lies and fabrication. Looking back at the timeline, I had just moved from the US to Japan and was working on launching a business at a startup when the exposure and subsequent collapse actually happened, so honestly I only skimmed a web article about Theranos at the time and thought “no way” in passing. But I had watched closely how Elizabeth Holmes was being hyped up in Silicon Valley, so the shock hit me all the harder for that. (Though I was busy with my own things at the time, so I’d completely forgotten about it lol)

This book meticulously records the confessions of former employees and other people involved, along with the author’s struggle to get the story published after obtaining their testimony. One shocking episode lines up after another, but if you calmly stop and think about it, you land on the thought: “Isn’t something similar to what happened at Theranos, to varying degrees, happening at most startups?”

It’s just that not every company raises as much money as Theranos did, gets as much media attention, involves as many big names, or gets tangled in as much litigation, so the result is less exciting as a story. That’s really all there is to it.

Let me write down a few things that struck me.

“Bluffing”

For early-stage startups, raising money with a bit of bluffing has become such common practice it’s almost considered standard—a bad habit, but a standard one. No, I can’t call it entirely a bad thing. A small but definite number of people actually turn that bluff into reality and make it all add up in the end. Investors, knowing this, are undeniably playing a kind of game of chicken themselves. But that’s a game with its own aesthetic, played strictly between “professional entrepreneurs” and “professional investors.” When amateurs play the same game with each other, it’s nothing but fraud.

“The shareholder whose connection to the company is unclear”

Look at any shareholder list and you’ll find one or two names whose connection to the company isn’t quite clear. That in itself is fine. It’s perfectly natural for someone to feel, “I don’t really understand the business, but I want to support it financially!” But if that one or two people happen to be a big name everyone knows, the story changes considerably. That’s because “once one big name bites, it becomes easier to raise money afterward.” In Theranos’s case, a genuine heavyweight was on the shareholder list from an early stage. Even if all he did was “support a young entrepreneur with a bit of spare investment capital,” anyone who saw that shareholder list would jump to the conclusion: “If Mr. So-and-so is backing this business responsibly, then what she (Elizabeth) is saying must be true.” In practice, I suspect that even professional venture capitalists rarely conduct thorough due diligence entirely on their own. This is even more true for tech ventures. They build their DD on the back of some clue or another—someone’s advice, past data, something. In such cases, the “name” of an existing shareholder, or of the lead investor running that funding round, can become a source of serious misdirection.

“The executive doesn’t listen to a talented subordinate but listens to some famous rich person instead”

A subordinate is a specialist in their field, not a manager. Scientists and engineers with little corporate experience in particular can’t be called anything but naive about politics and money. A famous rich person, by contrast, usually became rich for a reason. They’re versed in both business and politics, and sharp when it comes to profit. An executive can’t help but sometimes think of a subordinate, “(they don’t really see the big picture).” And at that point, it’s regrettably common for the executive to be swayed instead by the opinion of a third party who sees only the bird’s-eye view. But the subordinate doesn’t know that third party, and of course neither respects nor relies on them. If you get the sense that the top of your own organization isn’t looking at you but is instead taking orders from something existing off somewhere else, your feelings will naturally drift away.

“Fragmentation of information”

Theranos certainly took on an unprecedented, groundbreaking challenge, but given how much money and how much talented staff they gathered, many people probably thought, “Shouldn’t they have gotten a little closer to actually pulling it off?” (In reality their testing device was nowhere near workable.) With over $1.4 billion in funding, more than 600 employees, and nearly ten years of time, it almost seems like any great invention should have been possible. But reading the book, I came to think that one reason they couldn’t pull it off was the thorough fragmentation of information. It’s common enough practice to disclose highly sensitive technology or features to only a limited number of people. Apparently even at Apple it was common knowledge that “ask Steve Jobs or Jony Ive about product specs, because the rest of us aren’t told anything.” But if you build a structure where no departments can coordinate horizontally at all, development simply can’t function. Each member probably thought something like, “Well, I’m getting a decent salary, so I’ll just handle what’s in front of me.” As a result, everyone probably ended up feeling “whatever,” “not interested,” or “I don’t really get the big picture anyway” even about the fact that they were building a system that was completely useless.

Americans and division of labor

I’ve noticed that Americans like dividing up labor. I think it’s because clearly defining your own area of responsibility makes things like accountability, performance review, required skills, and task status (i.e., protecting your private time) all explicit. This always carries the risk that the team stops functioning organically as a team, and Andy Grove, for example, pointed this out strongly during the period when Japanese semiconductors were threatening the US (though the practice persists because it’s cultural).

“The top doesn’t understand the current situation, or won’t try to”

Some organizations keep running even with fragmented lateral connections. Those are centralized organizations. They work fine if the person at the top is capable—though it’s tough. But in, say, a tech company where the top executive is a non-engineer, the presence of a close aide (often the CTO) becomes crucial. If that aide is competent and has a solid trust relationship with the top, this problem can be sufficiently compensated for. But once you reach a situation where “the top doesn’t understand the current state,” “the aide doesn’t understand it either,” and “neither the top nor the aide is competent,” there’s nothing more to be done. Game over.

“A promise made somewhere else”

Not a few people have experienced being rushed on a schedule that never appeared in any milestone plan. At companies where the executive also handles sales, it happens that they’ve gone and made some promise somewhere without telling anyone. It’s the result of desperately trying to seize a business opportunity, so you can’t exactly treat it harshly. Properly, that information should be shared with the team. Then the team would grudgingly come to accept it. But if it isn’t shared, trust breaks down. Employees are not slaves.

“Executives chosen through personal connections”

Trust is important when selecting executives. That’s true at any company. But that trust should be used in a way that means “not being afraid of clashing opinions.” At Theranos, the COO was Elizabeth’s boyfriend, and other executives included her brother, who had barely any real work experience, along with his friends—selections that weren’t made in a way that could productively build anything up. And the board of directors, which should have functioned as a kind of “older brother” or “father figure,” was likewise packed with members who had personal relationships with Elizabeth, while Elizabeth held over 90% of the voting rights for a long stretch of time. This is an extremely tragic situation.

“People valued for credentials alone, with thin practical experience”

A degree is quite a useful weapon. In fact, pitch decks at Silicon Valley startups typically list a startling number of names from prestigious universities. But everyone knows this isn’t really the essence of the matter. Even knowing that, the trend persists—perhaps out of some kind of deference. Credentials are useful, but they’re nothing more than a bluff. There’s a well-known academic-credential hack in Japan, for example: “If you want the prestige of a University of Tokyo degree, get in through graduate school.” Your undergraduate grades, thesis, and interview are enough to be properly evaluated there. It’s a far better bet than the all-or-nothing entrance exam.

“High pay for a figurehead consultant”

It’s rare, though not unheard of, for people outside the executive ranks to see a company’s accounting records. Buried in the payment history is a certain individual’s name. Google it, and it turns out to be someone quite prominent, a real player. And yet you’ve never heard the name mentioned inside the company, and you have no idea what their role is. Probe gently and you get answers like “we get advice from them about such-and-such,” or “wait, wasn’t that just very recently? (though the payments have apparently been going on for a year…)” In short, it’s a figurehead role—but if money at a level that could pay for several engineers has been flowing there, that’s guaranteed to leave you disheartened.

“The departing employee who suddenly disappears”

At companies with tight security, an employee who announces their resignation—or is told they’re being let go—in a meeting room sometimes leaves the company without ever returning to their desk. This is generally meant to prevent the risk of confidential information being taken out or leaked once they’re back at a regular desk. Beyond that general explanation, I think the real reason is often things like “to keep them from badmouthing the company to people around them” or “in case of some retaliatory act following a forced dismissal”—in other words, a means of preventing the fallout from a departure that wasn’t amicable. When someone leaves suddenly without even saying goodbye to coworkers, despite there being no risk in doing so, it’s usually this pattern. It’s a sign of a bad company.

“Want to buy back the stock of the company you’re leaving?”

This is an extremely serious matter, but as a general rule, when a company takes action to expel an executive who holds shares, the company will fight tooth and nail to get those shares back. Since this too is, in other words, a departure that wasn’t amicable, having an external shareholder who might hold a grudge is enough to keep an executive up at night, even over only a small stake. But from the perspective of the person who was forced out, it’s natural to think, “Why should I, on top of being forcibly ousted, also have to go along with what you people want?” Even so, there are plenty of cases where the company moves aggressively to reclaim shares, brandishing contracts or resorting to legal action. At Theranos, a veteran named Avie—who had served as a close aide to Jobs at both NeXT and Apple—found himself in this predicament, and when he consulted a lawyer, he was told something to the effect of: “Do you think this company has a future? Is this stock worth holding on to, given all the trouble?” Most people would probably say they have no attachment to it at all. And indeed, that’s what Avie did.

“Say the right thing and you get pushed out”

The section about Ian genuinely weighs on you as you read it. It’s the story of a veteran scientist with deep experience who raised objections about the performance of Theranos’s testing device, was branded “uncooperative,” sidelined, fell into depression, and then… Skilled scientists and engineers are good at saying “the right thing.” That’s practically a mystery of life itself, consistent across every country. At many early-stage startups, the gap between the “grand vision” and the “feeble reality” is enormous. Everyone knows this. And yet they keep pushing forward anyway, carried by their own fervor. No looking back, no hesitating. Just keep moving forward. In that kind of atmosphere, criticism built out of “the right thing to say” is naturally unwelcome. A clear-headed executive should be able to make an appropriate judgment call—something like “thank them for the feedback, but push ahead anyway,” or “come up with a workaround and don’t stop moving,” or “this is a line we can’t cross, so we stop here.” But at a company or under an executive already in dire straits, that feedback gets ignored, trampled on, or pushed aside, and everyone moves forward as if nothing happened. Of course, what skilled scientists and engineers say isn’t infallible either. There are startups where their predictions turn out wrong and the company still achieves great success. But in most cases, the outcome is just someone falling, without learning a thing, into a trap someone else has already fallen into before.

“Winning one undeserved award after another”

As I’ve written elsewhere, it often takes several years before the true state of a company in serious trouble actually becomes public. Theranos was no exception, and if anything, most such companies follow that same pattern. So even if a company is carrying some major, fatal flaw, for whatever time it has left, it can go right on being “today’s hot startup.” Naturally, receiving an award is an honor. You should take seriously the expectations of the people who chose you, and your stakeholders are surely delighted as well. But continuing to receive honors you don’t deserve breeds a different kind of harm. Namely, “losing track of where you actually stand.” And that alone is an extremely serious, potentially fatal affliction.

Originally published in Japanese at https://clazytech.com/2021/03/399/. Translated with LLM assistance and reviewed before publication.