Forbes published a “Hall of Shame” in late November 2023 — a list of honorees from its 30 Under 30 franchise that the magazine would like to take back.
The framing was a single sentence: “Below are 10, out of 10,000, that we’d like to walk back.” Sam Bankman-Fried was at the top of it.
And the magazine made the case for itself:
“The 30 Under 30 list is, by definition, future-facing, and not even Warren Buffett can predict with 100 percent accuracy. But we’d guess he’d be good with 99.9%.”
Ten out of ten thousand is one-tenth of one percent. The question worth asking is not whether Forbes counted correctly. It is whether ten is the number of people who did it, or the number who got caught.
How the numbers get on the list
The mechanism matters more than any single case.
First, you have to be nominated — and the nomination process is open. You can nominate anyone, including yourself.
Then comes evaluation, which looks at funding, revenue and investors. Forbes has described financial metrics as top of mind.
Here is the problem, and it is structural rather than malicious. These are private companies. Mostly startups. Their numbers are not posted anywhere. The numbers that get you onto the list come from the founders.
Which creates a two-step loop: the numbers get you on the list — and then the list becomes the credential you use to sell the numbers.
Once a founder can say “Forbes 30 Under 30” in a pitch deck, the magazine’s verification problem has become the investor’s verification problem.
Charlie Javice, and the list that helped sell it
The most thoroughly documented case is Frank.
Charlie Javice appeared on the 2019 list for Frank, an online platform built to simplify how students applied for college financial aid. By 2021 she was pitching it to JPMorgan Chase, claiming Frank had more than four million users. The bank bought the company for $175 million.
The real number of Frank users was under 300,000.
Before the sale, she asked her own engineer to generate fake user accounts. He refused. So she paid an outside math professor $18,000 to generate them instead — and mixed in a list of more than four million real students bought from a data broker to make the whole thing look legitimate.
It came apart the way these things do. JPMorgan sent marketing emails to its new customers. Seventy percent of the addresses did not exist.
In March 2025 a jury convicted her of conspiracy, wire fraud, bank fraud and securities fraud — alongside co-defendant Olivier Amar, after a six-week trial.
On September 29, 2025, Judge Alvin K. Hellerstein sentenced her to seven years in prison.
Gökçe Güven, and two sets of books
The most recent case is closer to the mechanics of the list itself.
Gökçe Güven, a Turkish entrepreneur, founded Kalder in 2022 — a loyalty and rewards company that let brands offer cash back without building the technology themselves. She was featured on the 2025 list.
While raising about $7 million, according to the reporting, she sent investors a pitch deck claiming 79 companies were partnered with Kalder. Twenty-six of those companies were using it. The other fifty-three were at some earlier stage. She claimed roughly $1.2 million in annual recurring revenue.
The real figure was closer to $60,000.
Many of the listed companies were on heavily discounted test pilots. Others had no agreement at all — not even for free. And she showed potential investors a company contract that had been forged.
She was keeping two sets of books — one prepared by her outside accountants with the real numbers, and one for investors with the fake ones.
In May 2026 she pleaded guilty to securities fraud. On October 30 a judge sentenced her to 25 months in federal prison.
And there is a detail in that case that closes the loop on how the list is used: Kalder cited its upcoming appearance on 30 Under 30 as proof that the business was succeeding.
The ones nobody has checked
Both of those founders got caught. The harder question is what the list looks like for everybody else.
Take the 2024 list, and the first name on it: Matthew Vega-Sanz and his twin brother Michael. Forbes featured the twins and produced two videos telling their story.
In 2018 they started what became Lula, a car-sharing app offering daily rentals to college students. The cars needed insurance. Forty-seven companies refused to provide coverage. That killed the app — and produced the next idea: if nobody would insure them, they would build the insurance technology themselves. They called it the Stripe for insurance.
It appears to have worked. By 2021 they had raised about $18 million. In August 2023 Lula raised a $35.5 million Series B. By one public accounting, the company has raised around $60 million.
And then there is the announcement. Matthew posted on LinkedIn that the business had grown “nearly 40x” in the previous sixteen months — and linked to a TechCrunch article instead of giving the figure.
The only number in that article that works out to 40x is customer count: from 99 customers to 4,000. The article also said monthly revenue was up twenty times, without stating the actual figures, because the founders declined to reveal them.
To be precise about what that is: it is not fraud, and nobody has alleged that it is. “40x” and “up 20 times” are the kind of numbers a founder can put in a LinkedIn post and a reporter can print because there is nothing to check them against.
Which is exactly the gap the list is built on.
What the 99.9% is actually measuring
Forbes’ defense is arithmetically fair and logically weak, and the difference matters.
Ten out of 10,000 is a real number, and it is presumably accurate as a count of the honorees the magazine has publicly disavowed.
But disavowing is something that happens after a conviction. It requires a jury, a plea, or a headline. The fraud has to be discovered by somebody with subpoena power before the magazine has anything to walk back.
So the 99.9% is not a measurement of the honorees’ honesty. It is a measurement of how many of them were charged. Those are different datasets, and one of them is being presented as the other.
SPLASH’S READ
SPLASH'S READ: The list is not a fraud. It is a citation. And that is the whole problem.
Forbes is doing something every publication does — aggregating claims from people who have an interest in the claims being believed. The difference is that this particular aggregation gets printed on a magazine cover, handed to a founder, and then quoted back at investors as independent verification.
Nobody at Forbes is checking whether Frank has four million users. They cannot. It is a private company. They are checking whether the founder says it does, and whether somebody with a checkbook believed it.
Which means the list is not measuring achievement. It is measuring momentum — and momentum is the one thing a fraudster can manufacture cheaply.
The Javice case is the clearest proof, because the $18,000 to a math professor was not the hard part. The hard part was the four million real students she bought from a data broker to make the fake accounts look plausible. Even the lie required real data underneath it.
And the sentence Forbes wrote for itself is the most revealing thing in this whole story: “not even Warren Buffett can predict with 100 percent accuracy.”
That is true. But Buffett is not predicting. He is reading audited financials. The difference between an investor and a magazine is that only one of them can demand the books.
Ten out of ten thousand is what the magazine found. It is not what is there.