A 9-to-5 employee pointed an open-source agent at his own apps and told it to do the marketing: it researched competitors, generated slideshows, posted them, read the analytics and doubled down on winners — 8 million views on TikTok, run off an old PC under his desk.
$700/month from the apps being marketed · runs on an old PC under a desk, open-source agent, no ad spend, no agency · a second operator, Ernesto Lopez, reportedly at $70,000/month with autonomous businesses
"Oliver didn't actually make any TikTok videos or write a single caption. In fact, he didn't even use TikTok at all." The marketing loop — research, produce, post, measure, double down — is now something you delegate to software rather than to an agency.
Fewer bars = easier, cheaper, or faster for an AI-assisted solo builder. Editorial judgments based on the case details.
Oliver Henry has a full-time job and builds small apps on the side — the standard portfolio approach. The difference is what he did about distribution. Instead of hiring an agency or grinding out content himself, he set up an autonomous agent on an old PC under his desk, connected it to the tools it needed, gave it a goal, and let it market his apps. In early 2026 those apps did roughly 8 million views on TikTok, and Oliver made none of the videos and wrote none of the captions. The agent researched competing channels in the niche, generated slideshow-style posts, published them, read what performed, and prioritised the winning formats. Peverelli is careful to attach the honest number: the apps being marketed generate about $700 a month. He then names a second operator, Ernesto Lopez, as reportedly making $70,000 a month with autonomous businesses, and mentions a third case he saw the same day — someone selling swimming pools by using image generation to show homeowners what their own backyard would look like with a pool in it. Our read: $700 a month is not a business, it is a proof of mechanism, and that is exactly why this entry is interesting. The scarce resource for every small builder is distribution labour, and this is the first credible demonstration of that labour being automated end to end rather than merely assisted.
Oliver Henry started with no audience, no ad budget and no spare hours — he holds a full-time 9-to-5 and builds the apps on the side.
The quick-reference card — acquisition channels, replication playbook, and risk map — is members-only.
Join the Unicorn Club — $5/moCancel anytime, in one clickPaste into Claude Code / Codex and get a working version of this product end-to-end.
Two production-grade prompts per idea — a full build spec and an SEO growth plan — plus 10 tailored build specs a week. Members only.
Join the Unicorn Club — $5/moCancel anytime, in one clickEvery claim traces back to a listed source — open any of them and check the number yourself. Timestamps jump to the second each figure is said.
Members read the exact words said here
5:40 · hear it in the videoMembers read the exact words said here
7:01 · hear it in the videoThe exact line behind each figure is in the Unicorn Club. Read the quote, then jump to the second it's said. Join the Unicorn Club
Data credibility: Creator-reported and second-hand throughout. Christian Peverelli describes Oliver Henry's setup and results; Oliver is not interviewed and no dashboard, analytics screenshot or account is shown. The ~8 million TikTok views 'in early 2026' and the $700/month app revenue are relayed figures with no source given. The $70,000/month attributed to Ernesto Lopez is explicitly hedged as 'reportedly' in the video and comes with no mechanism, timeframe or product detail — we have carried it forward only because it was stated, and it should not be planned against. The backyard-pool operator is mentioned as something the presenter 'saw earlier today' with no numbers at all. The technical description (open-source agent, self-hosted on an old PC or VPS, trained on skills, connected tools, autonomous goal execution) is the presenter's account, corroborated only by his statement that his own company Bond uses the same technology internally. The security warnings, including the deleted-repository and deleted-mailbox anecdotes, are as told in the video. Sections 06 and 07 are largely our analysis rather than reported fact.
The money never comes from the views directly. There is no ad revenue in this loop and no client paying for the service — the agent produces TikTok impressions, some fraction of those viewers go and install one of Oliver's small apps, and the apps bring in about $700 a month.
The turning point is step four. Research, generate and post are just automation, and an agent that only does those three is a spam machine with good manners.
This build has the opposite problem from most: the agent produces data every day, so the quiet stretch is not silence but noise that never turns into money.
The rest of this breakdown — 10 more sections — is in the Unicorn Club, with every section of all 1,335 ideas: first customers, first dollar, the flywheel, and the silent stretch.
Join the Unicorn Club — $5/moCancel anytime, in one click