Pareto 20, No. 15: Greg Isenberg, for choosing an idea before you choose a tool

Twice-weekly startup ideas, a sharp line on AI companies living off trial revenue, and a product of his own at the end of most roads. Useful, with the conflict stated up front.

Seat 15 on The Pareto 20 is Greg Isenberg, and it is the second seat in the domain that comes before tools: what is worth building, and for whom.

Who he is

Isenberg describes himself as CEO of Late Checkout, a holding company for community-based internet businesses, with three venture-backed companies sold and advisory roles at TikTok and Reddit, all self-stated. He co-founded LCA, a product and design firm, and publishes a weekly newsletter plus The Startup Ideas Podcast twice a week, with a YouTube channel past 700,000 subscribers. One correction to a record elsewhere: a creator-tracking site sometimes attributed to him is not his; his properties are gregisenberg.com and, in his portfolio, IdeaBrowser.

What to read him for

Curiosity revenue. His most quotable verdict, from April 23, 2026: "There's a bunch of AI companies with 1 to 10 million dollars ARR raising big VC money on what I'll call curiosity revenue, not real sustainable ARR." For a solo operator, the same test applies at a smaller scale: is anyone still paying in month three.

Do not skip steps. His AI and your job piece uses the McDonald's drive-through experiment as the cautionary case: "Don't do what IBM and McDonald's did. The problem was the tech wasn't good enough. Skipping steps is risky." Validate with people before you automate with agents.

Fast reads on new launches, with a money angle. Recent episodes (September 10, September 8, September 2, 2026) turn a model launch or a concept like local AI into "how would a small operator make money with this," which is a different question from the one most tool coverage asks.

Platforming builders. Guest episodes bring in people shipping the thing being discussed, which is where the useful specifics tend to be.

Start here

The podcast feed, most recent first. The free frameworks on his blog are the paywall-free version.

What he sells, and what that means for you

This is the conflict the list was designed to surface. IdeaBrowser, a startup-idea database, is his own product and sits at the end of a lot of his content; its pricing page was not reachable when we checked, so we are not quoting a price. LCA sells product strategy to enterprises. The podcast carries sponsors, including Brex and Claude Code, and Claude Code was also the subject of a solo tutorial episode (August 17, 2026) that carried its sponsor link. Free downloads on his site feed the newsletter. None of this is concealed, but the pattern is content that leads to a property he owns, and you should read the idea episodes with that in view.

Not for you if

You want ideas evaluated against your situation rather than pitched at volume. He generates; the filtering is your job, and Rob Walling is the better filter.

---

Part of The Pareto 20, twenty creators who get a solo operator most of the way on AI decisions. How we chose them. Facts above were checked against the creator's own pages on September 12, 2026.