It Was Never UGC. Jeffrey Escobar on Building House AI’s Creator Partnerships

The difference in practice

He describes the transactional version precisely because he does not do it.

“The bad version is: hey, make me a video, here’s the money, goodbye. And then you’re surprised when the second video is worse than the first and there’s never a third.”

The alternative takes longer and starts earlier.

“You get on a call. An actual call. You explain what you’re trying to do and why it matters. You ask what they’re building. Then you figure out how to work together. It’s slower and it’s the entire reason it works.”

Many of the people he partners with are not primarily promotional accounts.

“They’re across different age groups and a lot of them genuinely don’t care about promoting things. That’s not a problem, that’s a filter. If somebody will promote anything, their audience already knows that. The ones who are selective are the ones worth having.”

Which means onboarding is a conversation, not a brief.

The system underneath

The relationship part is the philosophy. The measurement part is the machine, and Escobar is precise about it.

“Every creator gets their own tracked link. We watch where the volume is coming from and where the concentration of downloads is. Then we double down on the winners and work with the ones who need improvement instead of dropping them.”

That last clause is the departure from standard practice.

“Most companies cut anybody who underperforms once. We coach instead, because a creator whose second video does 5x their first is somebody you now have a real relationship with. You can’t buy that. You can only build it.”

The output is what he is actually after.

“The goal is constant video every single day. Not a campaign. A permanent flow. And that beats ads, because an ad stops the second you stop paying and a creator relationship keeps producing.”

He adds the compounding argument.

“Every video is also a page on the internet with the product’s name on it. Ads disappear. That doesn’t.”

What holds up from the original version

Three lessons from the first year that he still stands behind.

Volume beats perfection. “A rough video that ships today teaches you more than a perfect one that ships next month.”

Creators know their audience better than founders do. “Give them the product and the goal, not a script. The first two seconds decide most of it and they know their first two seconds better than I do.”

The comment section is a free research lab. “People tell you exactly what they want the app to do next. We’ve shipped features that came straight out of comments.”

The best organic videos become paid ads later, already proven.

“That’s the sequence. Organic first, and the winners graduate. You never guess at a creative you haven’t already watched perform.”

The honest limits

Asked what this approach costs, he gives three answers rather than selling it.

“It’s slow to start. The transactional version gives you videos next week. Partnerships take months before they produce anything, and if you need numbers this quarter you will hate this.”

“It doesn’t scale linearly. Every relationship is a real relationship, which means somebody has to actually maintain it. There’s a ceiling on how many you can run well, and pretending otherwise is how you end up back at transactions.”

“And you can pick wrong. When you tie your product to a person, you inherit whatever that person does next. That’s a real risk and anybody running this should think about it before the deal, not after.”

Why he thinks it works anyway

“Because it’s a new form of a very old thing. People buy what people they trust are actually using. That was true before the internet and it’ll be true after.”

And because the product is one you can only understand by seeing it.

“Polished ads have a ceiling for consumer AI. People want to see the thing do something in front of them.”