Kevin’s Week in Tech: Theranos, Fraud and the Failure to Fail

Each week, Kevin Roose, technology columnist at The New York Times, discusses developments in the tech industry, offering analysis and maybe a joke or two. Want this newsletter in your inbox? Sign up here.

Hi, it’s me again. Let’s get to it.

The biggest story in tech this week was the Security and Exchange Commission’s announcement of fraud charges against Theranos, the smoke-and-mirrors biotech company; its founder, Elizabeth Holmes; and its former president, Ramesh Balwani. As part of a settlement, Holmes has agreed to pay a $500,000 fine; surrender much of her stake in Theranos, which was once valued at $4.5 billion; and accept a decade-long ban on being an officer or a director of a public company. The settlement doesn’t involve any jail time or admission of guilt, although the company is also facing a separate criminal investigation.

The complaint, which you should really read if you’re interested in the anatomy of a corporate downfall, contains stunning accusations. They include:

■ Theranos exaggerated its revenue by 1,000 times, telling at least one potential investor that it had made $108 million in 2014, when its actual revenue was around $100,000.

■ Theranos employees wrote flattering reports about the company’s own products, put the logos of pharmaceutical companies on them and used them to imply that those companies had endorsed Theranos’s technology.

■ Theranos was paid millions of dollars by Walgreens (in the complaint, it’s listed as “Pharmacy A”) to roll out testing services at its stores. When its own diagnostic tools weren’t ready in time, Theranos passed third-party testers off as its own.

■ Theranos repeatedly lied to reporters from publications like Wired, Fortune and The Wall Street Journal about how many tests its proprietary machines were capable of running, and gave those misleading articles to investors as proof of the company’s accomplishments.

This is juicy, Hollywood-ready stuff, and the complaint makes it clear that Theranos — which was once a legitimate Silicon Valley darling, with more than $700 million in venture capital raised from prominent investors — is one of the biggest scandals to hit the tech industry in years.

The complaint also helped explain the precise nature of Theranos’s deception, and how it differed from other Silicon Valley companies that have struggled to bring their products to market.

Unlike a shady cryptocurrency initial coin offering, Theranos didn’t start off as a sham. Holmes was trying to create something real and useful, a blood-testing machine that would be able to run hundreds of different tests on a machine roughly the size of a desktop computer, using much smaller samples of blood than a traditional machine. Had it worked, it would have been a legitimately important innovation.

The problem is that the miniLab, as Theranos’s testing machine came to be called, didn’t work. Or, at least, it wasn’t capable of handling all the tests Theranos had promised it would, within the time frame it had agreed to. According to the complaint:

But as September 2013 approached — the date for the launch of the first phase of the rollout of Theranos services in Pharmacy A stores — it became clear to Holmes that the miniLab would not be ready. At the time, Theranos had not fully integrated other testing methods into the miniLab and had not completed the scientific verification steps needed to make any of its blood tests available on the miniLab for patient testing.

On its own, this is nothing unusual. Production delays, especially of complex hardware products, are common in Silicon Valley, and overeager marketing and sales departments often get ahead of engineering teams. Apple is notorious for tinkering with new iPhone designs right up until the factory deadline, and practically every major hardware company has shown off products in onstage demos that weren’t yet shelf-ready.

But according to the complaint, instead of asking Walgreens for more time, Theranos decided to lie. Holmes and Balwani told their engineers to modify standard blood-testing machines to make them look like the company’s new miniLabs. They then misled Walgreens executives and potential investors about how their tests were being performed, the regulators said, while furiously trying to get the miniLabs ready.

What’s so odd about the Theranos story is that failure and delays are not just common in the tech industry, they’re practically expected. Apps routinely die on arrival, hardware doesn’t pan out, partnerships fall through. Ideally, these things happen before you raise hundreds of millions of dollars in venture capital and strike distribution deals with major retail chains, but even if they don’t, they’re generally not the end of the world. (For an example of a company that raised millions of dollars for an overhyped product and then imploded without accusations of fraud, look at Clinkle.)

It would have been embarrassing, and possibly career-damaging, for Holmes to admit to her investors that the miniLab was not going to work out, renege on the deals she had struck with chains like Walgreens and give back some of the money she had already raised. But it was an option, and one that might have kept her out of the legal mess that followed.

Instead, according to the complaint, she constructed an elaborate fiction that she would spend the rest of her tenure trying to keep alive, and that would lead her down the path to this week’s charges. In other words, Holmes’s mistake was that she refused to do what so many other entrepreneurs do every day, and come to terms with failure.

Kevin Roose writes a column called The Shift and is a writer-at-large for The New York Times Magazine. You can follow him on Twitter here: @kevinroose.