Killer-pattern catalog · public

The ten ways ideas die.

A guide to the most common reasons startup ideas die. Real anonymized examples. Counter-examples for the rare ones that survive.

Tarpit idea-class

K10 · §4 / ENGINE · Graveyard of funded attempts

An idea class that looks attractive and gets tried again and again, even though many funded teams already died there: consumer social, "social network for X", events and coordination apps, dating, "Uber for X" marketplaces, generic AI wrappers, anonymous apps, and habit trackers. These classes usually die because of structure, not effort: cold-start network effects, zero consumer willingness to pay, fast copying by platforms or base models, or supply leakage. The engine compares the idea with a curated graveyard of prior funded attempts and asks the only question that matters: many teams already died here, what is concretely different this time? Founder optimism like "better UX", "we are passionate", or "huge market" is not an answer.

Signs your idea has this

  • 01The idea is a consumer social network, feed, or follow-graph "for X".
  • 02It is a coordination, events, or "plan things with friends" app.
  • 03It is a thin wrapper around a base LLM (prompt + textbox).
  • 04It is an "Uber for X" or local-services marketplace where supply can leak away.
  • 05The difference is "better UX" or "underserved niche", not a structural wedge.
  • 06There is no captive distribution channel that defeats cold-start.
  • 07There is no business buyer with real, recurring willingness to pay.

Public idea database

No public examples yet. The idea database is still small, and we are not making one up.

Rare survivors

Discord

Escaped the consumer-social tarpit through a captive beachhead: gamers who needed low-latency voice before the product broadened.

Faire

Escaped the marketplace tarpit by owning a real wholesale workflow and financing wedge, not by being a nicer directory.

Why this pattern exists fundamentally

A tarpit class fails because of its structure, not because every prior team was bad. A new entrant must name the concrete wedge that breaks the graveyard mechanism: captive distribution, a paying business buyer, a regulatory advantage, hard-to-copy data, or an owned workflow. Without that wedge, the history of the class is the best predictor of the outcome.

Related patterns

Got an idea that looks different? Run the same framework before building.

Check whether your idea is a graveyard class