Holmes Said She Created A Character. Every Founder Plays One.

A nearly three-hour documentary about a convicted startup founder just pulled about $15 million in its opening weekend.

"You Can See Everything," co-directed by Nathan Fielder and Lance Oppenheim, follows Elizabeth Holmes in the month before she reported to prison. TechCrunch reported it was the second biggest opening weekend ever for a documentary, with roughly 2,000 sold-out screenings across only 855 theaters.

Most founders will watch it as a story about someone else.

That reading is too comfortable. The film's sharpest idea, at least as reviewers describe it, is not that Holmes lied. Everyone already knows that. It is that she may have been playing a character for so long that the line between the character and the person stopped mattering to her.

Every first-time founder builds a character too. You need one to raise money, recruit, and sell. The question this film puts in front of you is simple: who is in charge, you or the character?

A month with Holmes before prison, and a long attempt to figure out who she is.

According to TechCrunch's review by Amanda Silberling, Holmes and her husband contacted A24 about filming her last days of freedom in 2023, hoping it would help clear her name in public. Fielder, known for comedy that blurs fact and fiction, took the project and turned it into something stranger.

The review describes a film full of "colliding personas." Holmes insists she committed no crime. Her husband appears to perform for the camera. Amanda Seyfried, who played Holmes in "The Dropout," is brought in to help reconstruct prison conversations Fielder could not record.

One detail stands out for founders. Holmes once told a reporter that Seyfried was "playing a character I created." Holmes and her husband later said she was misquoted and had said "they created." Fielder asks for proof of the correct quote. The review says he never gets it.

Whatever she actually said, the idea lands because it sounds true. A founder's public self is a construction. The danger starts when the construction gets its own budget.

Investors were told the technology worked. It did not.

The facts are on the public record. In March 2018, the SEC charged Theranos, Holmes, and former president Ramesh "Sunny" Balwani with what it called an "elaborate, years-long fraud." The SEC said they raised more than $700 million from investors while making false statements about the company's technology, business, and financial performance.

In November 2022, Holmes was sentenced to more than 11 years in federal prison for defrauding investors. Balwani was convicted separately.

The specific lies matter less here than the pattern. By most public accounts, Theranos did not begin with a plan to deceive anyone. It started with a big idea, a young founder, a powerful story, and a gap between what the story promised and what the lab could do. Every month, the gap got wider. Every month, telling the truth got more expensive.

That gap is the part first-time founders should study, because every early startup has one.

Because nobody funds, joins, or buys from a list of open questions.

Early on, your company is mostly a story. There's no revenue chart to point at. No customer logos. Maybe a prototype that works if you click in the right order.

So you become the evidence. Your conviction stands in for traction. Your clarity stands in for a product. Investors talk openly about betting on the founder at pre-seed, which means they are partly betting on a performance.

That is normal and mostly healthy. A persona helps you:

• Compress a messy company into a sentence people remember.
• Project calm to a team that needs it when payroll is close.
• Hold a long-term vision while the short term looks bad.
• Sell before the product is finished, which almost every startup has to do.

None of this is fraud. A founder who says "we will be the default way small clinics handle billing" is describing a goal, not a fact. Investors know the difference. The trouble starts when the persona starts speaking in the present tense about things that only exist in the future tense.

When you describe the roadmap as if it is the product.

There's a useful test here. Read any sentence from your deck, your website, or your last investor update. Then ask: if someone checked this today, would it be true today?

• "Our model reduces review time by 60%." Measured on what? With how many customers? If the answer is one pilot and a spreadsheet, say so.
• "We work with leading health systems." Paying customers, or two calls and an NDA?
• "Fully automated." Or automated with a contractor fixing edge cases at night?

Small stretches feel harmless because each one is small. But they compound. Once an investor wires money based on a number, that number becomes a commitment. The persona has made a promise, and now the company has to keep it or someone has to admit it was never true.

This is the gap Theranos let grow for years. Most founders never get anywhere near fraud. But plenty end up in a quieter version: a team that knows the real numbers, a board that hears a cleaner version, and a founder who feels trapped between them.

When you are building your plan and pitch, tools like Foundra can help you separate what you have proven from what you are betting on, so the two never blur in the documents investors read.

Build a few habits that force the real company and the story to meet often.

1. Label every claim in your deck. Use three tags privately: proven, in progress, belief. Proven means a customer or a dataset backs it. In progress means you're testing it now. Belief means it is your thesis. You don't have to show the tags to investors, but you should be able to say which is which in a meeting without hesitating.

2. Write investor updates with a fixed template. Include the same metrics every month, even when they go down. A template removes the temptation to pick flattering numbers. Investors trust founders who report bad months plainly.

3. Give one person permission to call you out. A co-founder, an early engineer, or an adviser. Their job is to say "that's not true yet" when you overstate something. If nobody on your team would feel safe saying that, you have a culture problem, not a messaging problem.

4. Demo the real thing. If the live demo is risky, show a recording and say it is a recording. Never fake a result on screen and let people assume it's live.

5. Keep a decision log. One line per big claim you make externally, with the date and the evidence. Six months later, you will be glad you can show where each number came from.

Close it early, while closing it is still cheap.

Suppose you already said something in a pitch that was more hopeful than true. Maybe a revenue number that included a pilot that never converted. Maybe a feature you called "live" that was still behind a flag.

The cost of correcting it grows every week. Today it's an awkward email. In a year, it might be a down round, a lawsuit, or a team that stops trusting you.

A simple correction often works better than founders expect:

"Quick correction to our last update. The $40K ARR figure included a pilot that didn't convert. Current contracted ARR is $28K. Here's what we learned and what we're changing."

Most early investors will respect that. They have seen far worse. What they cannot forgive is finding out later from someone else.

Silberling's review ends with Seyfried, reading a line in character, breaking off to say about Holmes: "She did that to herself." For founders, the useful version of that line is earlier and kinder. You are the only person who can stop your persona from writing checks the company can't cash.

Yes, if you watch it as a mirror, not a villain story.

The film reportedly does not try to prove Holmes innocent, and reviewers say it will not change anyone's mind about the verdict. That's not the point for you.

Watch for the moments when someone is clearly performing. Ask yourself where you do the same thing: board meetings, podcasts, all-hands meetings when the numbers are bad.

You don't need to drop the persona. You need it to stay honest with the person behind it.

What is "You Can See Everything" about? It's a 2026 documentary by Nathan Fielder and Lance Oppenheim that follows Theranos founder Elizabeth Holmes in the month before she reported to prison in 2023, and explores how she and those around her perform for the camera.

How much did the Elizabeth Holmes documentary make? TechCrunch reported about $15 million in its opening weekend, based on Sunday estimates, making it the second biggest opening weekend ever for a documentary.

What was Elizabeth Holmes convicted of? She was convicted of defrauding Theranos investors and sentenced in November 2022 to more than 11 years in prison. The SEC had earlier charged her with raising over $700 million through false statements.

Is it normal for founders to exaggerate in pitches? Selling a vision is normal. Describing future plans as current facts is not. Keep goals in the future tense and metrics in the present tense, backed by evidence.

How can a founder tell if their pitch crosses the line? Ask whether each claim would be true if checked today. If a number depends on a pilot, a single customer, or manual work, say so plainly.

What should I do if I overstated a metric to investors? Correct it quickly in writing, explain what changed, and share what you are doing next. Early correction protects trust far better than a later discovery.

1. Everyone's watching the Elizabeth Holmes documentary 'You Can See Everything', TechCrunch, October 18, 2026
2. In 'You Can See Everything,' Elizabeth Holmes meets her equal in comedian Nathan Fielder, TechCrunch, October 16, 2026
3. Elizabeth Holmes sentenced to 11 years in prison for Theranos fraud, TechCrunch, November 18, 2022
4. Theranos, CEO Holmes, and Former President Balwani Charged With Massive Fraud, U.S. Securities and Exchange Commission, March 14, 2018
5. Box office weekend report, Deadline, October 2026