Ryan Macdonald, a JPMorgan Chase executive who worked on the deal, sat in a federal courtroom earlier this year and described how the bank put 350 people on a single acquisition. The deal was JPMorgan's $175 million purchase of Frank, a student financial aid startup. Not one of those 350 people checked the user count that set the price. Three hundred fifty professionals looked at the numbers. The one number that mattered, nobody verified.
The founder's name was on the right list. The logo was on the website. So 350 trained professionals trusted what they were told and skipped the call they would have made if the name were unknown.
The Mechanism Has a Name
The research on this is worth reading. In 1961, Stanley Milgram ran an experiment at Yale. He showed that people stop thinking for themselves when a figure they see as an authority tells them what to do. Psychologists call it authority bias. It does not just work with lab coats and clipboards. It works with media badges. When a trusted institution stamps a person with approval, the people downstream stop verifying on their own. The badge becomes borrowed diligence. It does not add to the checks. It replaces them.
Charlie Javice, Frank's founder, had earned a Forbes 30 Under 30 spot in 2019. She had the pitch deck, the press, and the logo. What she did not have was the customers. She claimed 4.25 million users. The real count was under 300,000. The gap between what was sold and what was real was 93 percent.
Evidence at trial showed she paid a data scientist roughly $18,000 to build fake user records. Fake names. Fake profiles. Enough to fill a spreadsheet and fool the most resourced bank on the planet.
After the deal closed, JPMorgan sent marketing emails to 400,000 of Frank's claimed users. About 70 percent bounced. The addresses did not exist. Eighteen thousand dollars in fabricated data had carried a $175 million sale past 350 trained professionals.
Javice was sentenced to 85 months in federal prison.
The Pattern, Not the Outlier
If this were one case, you could call it a fluke. It is not one case. A paper from the Boston University Review of Banking and Financial Law gave the pattern a name: "30 Under 30 Pipeline to Prison." Seven fintech alumni from the Forbes list now face or have received federal sentences. The charges run from fabricated revenue figures to forged bank records to synthetic customer data.
Forbes published its own "hall of shame" in 2023. Ten picks they admitted were, in their words, "duds, or far worse." Their defense of the process: "Not even Warren Buffett can predict with 100 percent accuracy." That is a statement about stock picks, not a vetting standard.
The pattern is not old. It is still running. In the past few weeks, Gokce Guven, the CEO of a loyalty startup called Kalder, pled guilty to securities fraud. She had claimed $1.2 million in yearly revenue. The real number was $60,000. She kept two sets of books: one accurate set from her outside accounting firm, and a second set with inflated figures for investors. Her pitch deck listed 26 brands as clients. Some had only entered pilot programs at steep discounts. Others had no agreement with Kalder at all.
Guven was a 2025 Forbes 30 Under 30 pick. In an interview Kalder published to mark the honor, she praised the process. "Forbes was thorough in their evaluation," she said. The evaluation missed two sets of books and a 95 percent revenue gap.
The Flaw at the Level of System
Most people treat this as a character problem. It is a system flaw. The badge did not sit on top of the due diligence. It sat in place of it. When a name appears on a list you trust, you skip the work you would have done if that name were unknown. That is not a motivation problem. It is a measurement problem.
Three Checks That Cost Nothing
The fix is not more research. It is three concrete steps you can run in thirty minutes before you write a check to any vendor, partner, or hire who leads with a credential. Once that is clear, three moves follow from it.
Move 1: Call the Customers
Pick three names from the client list. Call them. Ask what they bought, when they bought it, and whether they are still paying. Do not email. Call. A real customer answers the phone and knows the product.
This is the step most people skip, because it feels like it crosses a line. It does not. It is the price of knowing what you are buying.
Move 2: Ask for Bank Records, Not Pitch Decks
Revenue on a slide is a claim. Revenue on a bank statement or payment processor report is a fact. If a vendor will not show you where the money lands, that tells you what you need to know. Guven's pitch deck said $1.2 million. Her bank said $60,000. The deck got the money. The bank had the truth.
Move 3: Ask to See Signed Contracts, Not Client Logos
A logo on a slide means nothing. A signed agreement with a date, a scope, and a dollar amount means something. Guven listed 26 brands on her deck. Several had no agreement at all. One look at the contracts would have ended the conversation.
What the Checks Show
Running these three steps on every vendor and partner you currently pay does something the badge never did.
It shows you which relationships rest on proof and which ones rest on a story. You see where the numbers match and where they drift. You learn who gets calm when you ask and who gets stiff. The gap between what was sold to you and what you received stops being a guess.
The Feedback Loop
At the end of thirty days, ask three things.
→ Which vendor's numbers held up under a direct check?
→ Which relationship looked strong but left no trail when you asked for records?
→ Which friction point came up more than once across different partners?
That is the difference between a credential that sounds right and a system that proves itself.
Where You Stand
Three hundred fifty people trusted a badge and skipped the call. Three checks, thirty minutes, zero cost would have caught it. A credential tells you what someone wants you to believe. The bank statement tells you what is true.
