digest
Short Summary
Main recommendation: Before building a startup idea, test whether it is a "tarpit" by talking to real users and checking whether the idea has been tried repeatedly without success, since tarpit ideas attract many founders and often fail. A tarpit is defined as an idea space that keeps drawing applicants and keeps failing, not simply an idea the speakers dislike, and the list of tarpits changes as technology changes.
Key Points
- A tarpit is named after the tar pits where dinosaurs got stuck: the idea looks like an appealing opportunity but traps founders who pursue it.
- Tarpit ideas are typically ones many founders independently come up with and that have been attempted and failed repeatedly with no meaningful change in conditions.
- Tarpit ideas often receive heavy praise from friends ("that's a great idea"), which is why praise alone is not a reliable signal.
- Some tarpits stop being tarpits when technology changes, so the presenters are open to new explanations, especially ones that say what the new technology makes possible.
- Two flavors of tarpits were identified: ideas that aim to make the world work better by changing human behavior at scale, and ideas built around quick arbitrage or getting rich fast.
- Generic "X for Y" framings such as "Copilot for lawyers" are common and often pursued, though not all are tarpits.
Details Worth Noting
- Concrete instructions: Talk to potential users and try to get a customer before committing to an idea, because the speakers repeatedly point to user conversations as the way to evaluate whether an idea is a tarpit.
- Concrete instructions: Research the history of your idea space and check whether many people have tried and failed; do not assume the idea is new, and be able to explain what has changed (such as a new technology) that makes it workable now.
- Concrete instructions: Be skeptical when feedback is overwhelmingly positive or when the idea seems obviously easy and already exists, since that is not the pattern of ideas that have worked.
- Concrete instructions: Test any "X for Y" idea by asking why the new tool would solve a real problem for that specific group, rather than assuming a Copilot-style assistant is automatically valuable.
- The speakers cited 'Wall Street Bets'-inspired startups and investment advisory businesses built around "getting rich fast" as examples of arbitrage-driven tarpits.
- A food discovery app for deciding where to eat and an app for coordinating social plans were given as examples of tarpit-style ideas that draw heavy praise from friends but fail to change human behavior.
- The presenters noted that AI-generated podcasts and voice generation are examples of ideas that became newly fundable because of recent LLM and voice technology, showing how tarpit status can shift.
Bottom Line
Tarpit status is not fixed: judge an idea by user evidence and by whether you can explain what has changed, not by how good it sounds or how many people praise it.
