The $36 Million Company That Fails 99 Percent of Its Clients

The press release arrived on a Tuesday, August 25, 2026, stamped "PR Newswire" and distributed to thousands of financial websites simultaneously. An inventor from San Diego had developed a patent-pending practice set for golfers — a way to swing at biodegradable balls over pools, backyards, and water hazards without ever retrieving one. It was available for licensing. It was the future of golf training. It was also press release number SBW-145 in a never-ending parade of identical announcements, and it was the key to understanding one of the most profitable failures in American business.

InventHelp generates an estimated $36 million a year.

Over 99 percent of its clients lose money.

The two facts should be incompatible. In almost any other industry they would be. A school that fails 99 percent of its students closes. A brokerage that loses 99 percent of its accounts is defrauding them. But InventHelp does not sell invention success. It sells the performance of it — and the performance itself is the product.

The Caveman on the Rock

Robert Susa founded InventHelp in Pittsburgh in 1984, forty-two years ago, as Invention Submission Corporation. The company's early television advertisements featured a cartoon caveman sitting on a rock, banging a wheel with a hammer, with the tagline: "You don't have to be a caveman to bring your new invention to market." It was a brilliant ad because it sold two things simultaneously: the idea that invention was a universal human impulse, and the idea that InventHelp was the civilized gateway to cashing in on it.

The company grew into what it calls an "invention service company," with sales offices across the United States and Canada, Germany, and Australia. It employs roughly 375 people. It has served more than 10,000 clients since its founding. Under Susa's leadership as president and owner, it became one of the most recognized invention promotion firms in North America.

The public mask was straightforward: InventHelp helps everyday inventors protect and present their ideas. The private ledger told a different story — one the company itself was required by federal law to disclose.

The Ledger That Discloses Its Own Failure

In 1994, the Federal Trade Commission accused InventHelp of misrepresenting the nature, quality, and success rate of its services. The settlement required the company to set aside $1.2 million for customer refunds. The company admitted no wrongdoing. But the FTC action triggered something more consequential: the American Inventors Protection Act of 1999, which required invention promoters to publish their success and failure rates before selling services.

So InventHelp publishes its failure rate. On its website. Before the contract. In plain view.

Between 2022 and 2024, out of 3,507 clients, five made more money from their invention than they paid InventHelp. That is a 99.86 percent failure rate. The rate has not improved over time. In 2015–2017, the profit rate was 0.75 percent. In 2007–2009, it was 0.5 percent. If anything, the machine has gotten slightly worse.

This is the irreversible bet of the business model: InventHelp does not need its clients to succeed. It needs them to keep arriving. Every press release about a "patent-pending" golf practice set, skin cream, or filing system serves the same function — it tells a new dreamer that someone, somewhere, is getting help. It does not tell them that the person getting the press release is almost certainly one of the 99.86 percent.

How the Money Actually Moves

The service package costs between $10,000 and $16,000, with some clients paying as much as $30,000. What they receive includes referrals to patent attorneys, marketing materials like 3D videos and prototypes, invention websites, and submission to a "Data Bank" of more than 9,000 companies.

Investigative analysis suggests the market value of these deliverables ranges from $2,650 to $7,000. The markup is not a margin. It is the entire business.

But the most important line item in the package is the one that sounds most promising: submission to the Data Bank. Court documents in class-action lawsuits revealed that many companies listed in this database were defunct, had never heard of InventHelp, or explicitly denied any relationship with the company. In one 2018 case, a listed company called "Abrams Gentile Entertainment" was traced to a Manhattan address containing only vacant office space and empty boxes.

The company the client is paying to reach does not exist. The client does not know this. The client keeps paying.

This is why the failure rate does not kill the business. The client's expectation — that a corporation will license their invention — is not fulfilled, but the client cannot distinguish between InventHelp's failure to deliver and their own invention's lack of marketability. The disappointment is diffuse, deniable, and ultimately personal. The golfer from San Diego did not sell a company his practice set. Maybe he just needed a better design. Maybe he needed better timing. Maybe he needed luck. He does not blame the press release that cost him nothing extra. He blames himself.

The Believer, The Enabler, The Named Plaintiff

Etta Calhoun was 64 years old, a retired patient educator from Spring, Texas, and a devout Christian when she invented "The Word of God Bedding" — bedding products printed with Bible verses to comfort the sick. She saw a television commercial featuring George Foreman, two-time heavyweight champion and ordained minister, speaking about how InventHelp helped bring inventions to market. She trusted him. She trusted the company.

Over six or seven months, she attended meetings in Pearland and The Woodlands, Texas. She paid a $200 down payment for a basic information package, then signed up for the next tier at $9,950. InventHelp connected her with Universal Payment Corporation. The loan carried 18 percent interest but was misrepresented as "interest-free". She paid $281 a month.

In 2014, she received a low-quality DVD presentation of her invention that she described as unattractive and hastily assembled. InventHelp told her the poor quality was intentional — to prevent idea theft.

A friend whose husband was a lawyer advised her to stop paying. She contacted attorney Julie Pechersky Plitt after reading a newspaper story about her own experience. She became the named plaintiff in the class action against Invention Submission Corp, the class-action lawsuit that consolidated cases across multiple federal districts. The plaintiffs sought $36 million in relief plus $72 million in compensatory damages.

The settlement was $3 million, finalized in March 2023. Calhoun received between $20 and $250. She said publicly that even without compensation, she wanted InventHelp's operations stopped.

George Foreman was never named as a defendant. He expressed surprise at the lawsuits. InventHelp continued operating. The press releases continued.

The golfer from San Diego filed through the same company, the same year, the same system that Etta Calhoun sued and lost to. His press release arrived eight months after Foreman died and InventHelp posted a tribute on Facebook. The machine does not need its spokespeople. It needs its clients.

The Press Release Machine

This is where the golf practice set stops being a product and becomes evidence of the operating model.

InventHelp sends dozens of press releases per month through PR Newswire, each one following the same template: Pittsburgh dateline, anonymous inventor quote, patent-pending description, invitation for licensing. The cost of a single PR Newswire distribution starts around $1,070 for a basic release and can exceed $6,500 with amplification packages. InventHelp almost certainly negotiates volume rates, but even at wholesale pricing, the press release program costs the company hundreds of thousands of dollars annually.

The return on that spending is extraordinary — if you measure it correctly. The press release is not an attempt to sell the invention. It is an advertisement for InventHelp itself. Every time a would-be inventor reads about someone "getting help," the cycle renews. The press release proves that InventHelp is active, credible, and connected to the business press. It looks like promotion of a client's invention. It functions as recruitment for the next client.

The San Diego golfer is both product and billboard. He paid $10,000–$16,000 for services that include this press release. The press release costs InventHelp perhaps $1,500–$3,000. The rest of the margin flows to Susa and his 375 employees. The golfer's invention has a 0.14 percent chance of earning him more than he paid. The press release has a near-certain chance of earning InventHelp another client.

What the Investor Sees

InventHelp is not publicly traded. There is no ticker, no quarterly report, no earnings call. It is a private Pennsylvania corporation owned by Robert Susa, and it has generated estimated annual revenues of $36 million for years while operating with a documented failure rate that would shutter any business built on outcomes.

But that is precisely what makes the case instructive. The investment lesson is not about buying InventHelp stock — it is about recognizing the structure that makes the stock unnecessary.

The company has insulated itself from the consequences of its own failure rate through three mechanisms:

First, the revenue is upfront. The client pays before any outcome is possible. The company never needs to share in invention profits because it has already captured all the value it can extract. The service contract converts hope into cash at the moment of maximum enthusiasm.

Second, the failure is deniable. When a client's invention does not generate revenue, the client cannot prove that InventHelp caused the failure. The invention might have been unmarketable. The timing might have been wrong. The client might have executed poorly. The 99.86 percent failure rate is disclosed — but it is disclosed the way a casino discloses the house edge: technically transparent, functionally invisible to someone in the moment.

Third, the liability is capped. The 1994 FTC settlement required $1.2 million in refunds. The 2023 class-action settlement cost $3 million. Over 42 years and more than 10,000 clients, those are modest penalties relative to the company's cumulative revenue. The cost of failure — for the company, not the client — is a rounding error.

This structure appears in public markets too, just wearing different clothing. It is the education company that charges students upfront and measures success by enrollment, not employment. It is the coaching firm that sells transformation and measures revenue, not outcomes. It is any business that monetizes the belief in a future result rather than the result itself.

The investor who recognizes this pattern sees the same question every time: does the company earn its revenue from the success it promises, or from the hope that precedes it? When revenue is decoupled from outcomes, the business is not selling a service. It is selling itself, repeatedly, to people who cannot distinguish between the sales process and the product.

The golfer from San Diego still has his patent-pending idea, his biodegradable balls, and his email address listed in a press release read by financial websites that will archive it and forget it within a week. Somewhere in the database of 9,000 companies, there may be a manufacturer who notices it. Or there may not. The odds, according to InventHelp's own disclosed statistics, are 1 in 700.

Robert Susa does not need those odds to change. He needs the next inventor to see the next press release, feel the same recognition Etta Calhoun felt, and decide that this time, the idea is different. The machine was designed for that exact sequence. It has been running for 42 years. The golf practice set was never the invention. The press release was.