👋 Hi and welcome to today’s edition of Venture Radar - where we uncover real internet businesses making real money, and break down how they do it.
Here’s today at a glance -
Steal this ideas: Duolingo for skilled trades, Trustpilot for AI agents & clinical trial recruitment.
Real businesses: $6M website, $400K/month AI app & $1.5M per year video library business.
Side hustles & passive income: $400/hour reviewing AI, $150K/year coaching & $7K/month food tours.
Unlock 50+ founder playbooks & templates, a growing database of profitable business ideas and 2-3 premium deep dives every week.
FROM OUR PARTNER - MECO
Reading newsletters in the inbox is frustrating - it is noisy and easy to lose control of subscriptions.
Now with Meco you can enjoy your newsletters in an app built for reading, while giving your inbox space to breathe.
Meco connects to Gmail & Outlook, allowing you to instantly move your newsletters to the app (and clear your inbox)
The experience is packed with features to supercharge your reading including the ability to group newsletters, set smart filters, bookmark your favorites and read in a scrollable feed
You can try the experience for free and available on iOS, Android and web (desktop)
Over 30k readers are enjoying their newsletters (and decluttering their inbox) with Meco.
Partner with us : Get your brand in front of 55,500+ readers including founders, indie hackers, side hustlers and aspiring founders.
🥷 STEAL THIS BUSINESS IDEA
Trade schools aren’t accessible to everyone. They can be expensive, far away, or difficult to fit around a full-time job. At the same time, skilled trades like electrical work, HVAC, welding, plumbing, and construction still need more workers.
That creates an interesting opportunity:
Build a consumer-first learning app for people who want to explore or start learning a trade from home.
Think Duolingo, but instead of learning Spanish, you’re learning how an electrical panel works, how to read wiring diagrams, how HVAC systems are diagnosed, or the fundamentals of welding.
Lessons could be short and visual, with quizzes, simulations, troubleshooting scenarios, progress streaks, and eventually hands-on projects users complete at home. AI could act like a tutor when someone doesn’t understand a concept.
The important part is not pretending an app can replace the supervised, hands-on training or licensing required for many trades. It can become the layer before and alongside that training: helping people discover which trade fits them, learn the fundamentals, prepare for apprenticeships or exams, and continue practicing theory outside the job site.
Start with one trade and one clear outcome. Build the best beginner learning experience for aspiring electricians, for example, before expanding into HVAC, plumbing, welding, automotive repair, and other skilled careers.
Most education startups have spent years fighting over coding, languages, and white-collar skills.
Skilled trades still feel surprisingly underserved by great consumer software.
As AI agents start doing real work, we’re going to run into a surprisingly basic problem: how do you know which agents you can actually trust?
When you hire a person, you have a resume, previous employers, references, reviews, and a work history. An AI agent can show up with almost none of that.
Build a reputation layer that follows agents across platforms and gives each one a persistent track record.
Instead of relying only on star ratings, the profile could show what the agent has actually done: tasks completed, success and failure rates, how long it has operated, permissions requested, complaints, security incidents, verified developers, and feedback from previous users.
That becomes particularly useful when agents start transacting or collaborating with other agents. Before giving an unfamiliar agent access to your inbox, codebase, money, or customer data, your system could check its reputation first.
Start as a simple directory and verification service. Eventually, the bigger opportunity is becoming infrastructure: an API that marketplaces, companies, and other agents query before deciding whether another agent should be trusted.
We spent decades building reputation systems for humans and businesses.
AI agents are going to need one too.
Pharma companies can spend thousands of dollars just to find a single eligible patient for a clinical trial. Yet many trials still struggle to enroll enough people, and some research sites end up recruiting nobody.
That’s where a surprisingly boring agency business exists.
Instead of selling marketing services to another SaaS company, specialize in helping clinical research sites find eligible patients.
Say a site needs adults aged 18-65 with moderate-to-severe eczema who aren’t currently taking a certain type of medication. Your agency builds compliant ads around that study, runs targeted campaigns on channels like Facebook and Google, sends interested people through a pre-screening process, and passes potentially eligible candidates to the research site.
The customer isn’t really buying Facebook ads. They’re buying faster enrollment.
And the economics can be attractive because delays are extremely expensive for drug developers. Clinical-trial marketing engagements can run from several thousand dollars to tens of thousands per month depending on the scope.
The obvious barrier is regulation. Patient recruitment advertising generally needs study-specific review and approval, privacy needs to be handled carefully, and compensation structures can raise ethical and compliance issues. But that’s also part of the opportunity: most generalist performance marketers don’t want to learn the clinical-trial workflow.
Start by becoming extremely good at one therapeutic area and selling to individual research sites rather than trying to land a giant pharmaceutical company immediately. Build repeatable campaigns, prescreening funnels, reporting, and compliance workflows around that niche.
It’s essentially a specialized lead-generation agency.
Except one qualified lead can be worth thousands of dollars to the customer. (More Here)
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.