Jeff Harmon

The 8% everyone dismissed was 30 million customers

A dentist pitched Jeff Harmon's university class on selling his OraBrush tongue cleaner. The class said only 8% of people would buy it online, so he should license it to a giant; Harmon pointed out 8% of the US is about 30 million people and argued for focusing on them.

“And the group got up and just said, you can't really sell these, and nobody wants to buy these online, like 8% of people wanna buy them on the internet, and you need to go do a partnership with Procter& Gamble or Johnson& Johnson or somebody. And I interrupted them and said, 8% of people is like 30 million people just in the United States. Why wouldn't you focus on them?”

Steal thisWhen a market is dismissed as a small percentage, convert the percentage into absolute people and ask whether that's enough to build on.

Find your glitch: be first on an underpriced new ad channel

OraBrush's $500 video worked because YouTube had just launched ads at under a penny per view, letting the company spend $1 to make $2-3. Harmon calls this 'finding your glitch': cheap early traffic buys the data to optimize, so you stay efficient as costs rise, giving you a beachhead (a term he credits to Paul Ahlstrom).

“So it's like the early days of AdSense. And so we were able to spend $1 and bring in $2 or $3 now. a dollar in total costs, both the ads, the shipping, and the brush itself. And so that glitch of being the very first on YouTube is what allowed us to find enough data to take action, to be able to optimize and then become more efficient. And then as the cost of advertising on YouTube started to climb, we were able to just get more and more efficient because we had the traffic coming in now to be able to make those changes.”

Steal thisWhen a platform launches a new ad product, get on it immediately while attention is underpriced, and use the cheap traffic to collect optimization data before prices rise.

The user method: the best businesses solve the founder's own problem

Harmon's first rule for finding a winning product is to solve a problem you personally have. OraBrush began as a dentist's superglued prototype to fix his missionaries' bad breath, and Airbnb began as founders renting air mattresses to cover their rent.

“So he was solving a problem for himself, and it happened to be something that many, many other people wanted to solve. There's a book called The User Method by Jeff Schwarting, and Jeff Schwarting goes through and explains that out of the top businesses, the vast majority of them were made by founders who were just solving a problem for themselves. The Airbnb founders were just trying to pay rent in San Francisco.”

Steal thisStart from a problem you personally live with; if you'd hack a prototype for yourself, others likely want it too.

Build for agentic shopping: let AI bots buy through your API

Harmon says he no longer uses Google for search and that many people now have AI assistants like Grok shop for them. Businesses that expose an agent-friendly checkout, so the bot talks to an API instead of scraping pages, can reach a fast-growing group of shoppers. He calls it a current glitch.

“And if you're a business that has an agentic shopping system where the AI bot doesn't have to go and scrape everything, it just talks to your API, then you can plug into this huge growing, rapidly growing group of people that are wanting to shop through their bots. They don't want to shop themselves. They don't want to click on Google anymore. That's a huge opportunity and a potential glitch.”

Steal thisExpose your catalog and checkout through an API that AI shopping agents can use directly, before competitors optimize for it.

VidAngel's origin: build the legal version of what pirates already want

When Game of Thrones was at its peak, scrubbed versions without sex and violence hit torrent sites within 24 hours and were downloaded by millions. Harmon read that as proof of demand and built VidAngel as a legal way for families to filter content.

“And so, we're looking at that and saying, not only do we want something like this for our kids, but what about the millions of people who are just saying, my teenager wants to watch this. I want them to be able to participate in the cultural conversation, but I don't want them to see rape scenes. And I'm going to use this torrent version. But that's completely illegal. And so what we're saying is, let's build a legal version of what millions of people are already asking for.”

Steal thisLook for things people are already doing illegally or through hacks, then build the legitimate version.

The ad agency that pushed YouTube to make skippable ads any length

After the OraBrush campaign, YouTube spent several days with Harmon's team in Provo, Utah, and they came up with the skippable ad. Harmon argued against capping ads at 15 or 30 seconds, because if an ad holds attention it should be allowed to run long. As he puts it, give me 5 seconds and I get 5 minutes.

“YouTube came out and talked to us and they said, how can we get more people to do things like Aura Brush? And we, and we came, they sat, I think for 3 or 4 days, they sat with our team in Provo, Utah. And we came up with this idea of a skip ad button. where you could, after 5 seconds, you click skip if you're not interested in the ad. And that concept, originally they were gonna make it like a 15-second ad or a 30-second ad. And I said, no, don't put a time limit on it. If I can maintain your interest, let me just keep them.”

A 20-minute YouTube 'ad' crowdfunded $11M from 17,000 investors

Angel ran The Chosen's short-film pilot as a full 20-minute YouTube ad that ended with a pitch to invest. 17,000 people put in $11 million, using about $800K of ad spend, and the show went on to reach about 200 million viewers.

“We could just take this same concept that we built for these 2, 3-minute ads and just go 20 minutes. Let's just do 20 minutes and see what happens, 'cause it's entertainment. And so, they watch 20 minutes, they're totally invested, and then we pitch 'em, come invest in this project, and let's make a TV series like the Game of Thrones for the New Testament. That's basically what it was at that time. And 17,000 people invested $11 million into The Chosen.”

Paying 8% in ads to raise money beats Wall Street's 8-12% fee

The Chosen's director was upset that about $800K of the $11M raise went to advertising. Harmon's view is that this is a normal cost of capital: Wall Street would charge 8-12% to raise the money, and the ad spend also built awareness for the show.

“And I was like, well, how do you think we built Poo-Pourri and Squatty Potty? We spent money. And if you had gone to Wall Street and asked them to raise you money, they would charge you 12%, right? They'd charge you 8 to 12%. So we're actually right in that world, but we're just having to buy ads, which is getting you awareness on your project.”

Steal thisJudge the ad spend on a crowdfund as cost of capital and compare it with what a bank or underwriter would charge.

It's easier to teach a comedian marketing than to make a marketer funny

Harmon says one lesson from building the Harmon Brothers agency was that comedians are core to a creative marketing team. It's far easier to teach comedians the marketing formula than to teach marketers to be funny.

“In the marketing world, I learned very quickly when we were building the Harmon Brothers ad agency, that it's way easier to train a comedian to be a marketer than it is to train a marketer to be funny.”

Steal thisHire comedians for creative work and teach them your sales formula.

The Harmon Brothers retreat: pick the lead script, then strip the rest for parts

Weeks before a retreat, 3-4 comedy writers use the product and write rough scripts. On day one each reads their script, the team picks the strongest skeleton by evening, and the next three days go into moving the best parts of the other scripts into it. Every stakeholder who has to sign off is in the room, so nothing gets watered down over email.

“We'll have 3 to 4 writers, comedians that are trained. We send them all the product, and they get to use it, and they get to play with it, and then they show up to the retreat with their ideas, their rough scripts. Mm-hmm. Then each one of them, on the first day, or the first morning, is basically doing a script read of what their vision is. And then by the evening, or the afternoon or evening of the first day, You are picking the winner, the bones of the ad out of those scripts.”

Steal thisHave several writers arrive with full drafts, pick one as the skeleton on day one, then spend the rest of the retreat folding the best ideas from the others into it.

Face warming: get a face hot in the algorithm before launching a channel

Social algorithms recognize faces in videos and serve related content, Harmon says. To promote a Kevin James movie, James first reposted old clips to get his face trending, then launched a fresh channel as his art-teacher character. The algorithm recognized him and pushed the new videos to his existing viewers.

“So he was getting himself hot in the algorithm. Hmm. And, and then he's trending. And what we do is we took and said, okay, Matt Taylor launches fresh channel. Nothing on it. But the TikTok algorithm and the Instagram algorithm and the YouTube algorithm recognized this is Kevin James, and this face is trending elsewhere on the system. And so the moment we launched those, they started going viral because people that were— people who had been watching Kevin James videos started getting served Matt Taylor videos.”

Steal thisBefore launching a new channel, get a recognizable face trending with existing content so the algorithm carries that audience over to the new account.

Kevin James, a tux, and 12 empty Super Bowl seats

The Kevin James art-teacher campaign got over a billion views. It ended with a Super Bowl stunt: Angel bought 12 seats, sat James in the middle in his wedding costume with flowers, and placed influencers nearby to post 'is that Kevin James?' He ended up on the stadium screen without any commercial message.

“And I think that campaign got over a billion views. And then at the end of the campaign, on Super Bowl Sunday, which is the weekend before, before Valentine's Day, he showed up at the Super Bowl, and we bought 12 seats at the Super Bowl, and we put Kevin in the center of 12 empty seats. And he was in a tux with his flowers in the exact costume he is in the movie, his exact wardrobe.”

Angel's 3 million paying members can veto any film before release

Members of the Angel Guild pay a subscription, vote on every film headed to theaters, and get two tickets to each release. If the audience vetoes a film, Angel can't distribute it. Harmon calls this 'the user method at scale.'

“And what they're doing is, as a member of Angel, you get to vote on every movie that comes into Angel that goes to theaters. So, you get the chance, if you want to, you can go and you vote. And if you veto, if you, if the audience vetoes a film, we can't take it. We can't send it to theaters. We can't do anything.”

Steal thisGive paying members a binding vote on what you release, so curation and demand validation happen before you spend on distribution.

Sound of Freedom: a $14M film that grossed $250M worldwide

Harmon describes Angel as a base-hits business that gets the occasional home run. Sound of Freedom cost $14 million to make plus about $40 million in marketing, and grossed $250 million worldwide.

“I think we're a base hits business that gets home runs every once in a while. So we had Sound of Freedom was a home run, did $250 million worldwide. That was a $14 million movie.”

Hollywood has a cost problem, not a revenue problem

Harmon calls Hollywood a knife fight and quotes a retired executive who said his best friend in Hollywood was worse than his worst enemy in Las Vegas. He argues more money than ever is coming into the industry, and that Angel's job is to fix how much it costs to make and sell films.

“And our job as a business is to fix what's broken in Hollywood, because Hollywood hasn't been a good business. It has a— it's not that they don't have enough revenues. They have plenty of— I mean, there's more revenues coming into Hollywood than ever. And it historically has been America's biggest export is media, is movies. That's our number one export. But this industry, the Hollywood industry, has a cost problem, not a revenue problem.”

If AI makes everyone 10x more effective, you want more people, not fewer

Harmon thinks companies blaming layoffs on AI are looking for excuses. In his view, when a tool makes everyone 10x more productive, the rational move is to hire more people, and he doesn't expect AI to cause mass unemployment.

“I don't, I think they're just looking for excuses to fire people because economics work like this. When you get something that's highly, that makes everybody 10 times more effective, you're going to become so much more productive that you need more people. You don't say, I want less people that are 10 times more effective. You say, I want more people that are 10 times more effective. It's so stupid.”

Have kids sooner, and more of them, than you think you should

Asked what he'd tell his younger self, Harmon says nothing in his professional life compares to marriage and kids. He advises having children earlier and having more of them than feels sensible.

“Well, the first thing I'd say is like, no matter what you do in the workplace, nothing will compare with getting married and having kids and having kids sooner than you think you should and having more of them than you think you should. Like, nothing in the professional workplace will even compare. That's where my real success comes from, is marriage and kids.”

The $5M rule: don't fight a dispute worth less than $5 million

When clients refused to pay the agency after successful campaigns, a mentor told Harmon not to fight over anything under $5 million. A fight takes 18 months to 3 years, and in that time an innovator can create more value. Harmon walked away from a few million dollars, and says doing so changed the course of his life.

“he said, unless it's worth over $5 million, it's not even worth fighting the battle over because you're gonna end up up fighting for 18 months to 3 years. You're going to drain the life out of yourself. And you may win, probably win, but then where did you get— He's like, but in that period of time, somebody who's innovative is going to make more value than that's worth. And not just that, it's just not a happy life.”

Steal thisBefore suing over money owed, compare it with what you could build in the 18-36 months the fight would take. If it's under the threshold, let it go.