Tarpit Ideas
A category of startup ideas characterised by: (1) attracting many founders repeatedly over many years; (2) generating easy positive validation from potential users; (3) rarely succeeding. The danger is precisely the positive feedback — it draws founders in and keeps them stuck.
Coined (or at least popularised) by Dalton Caldwell and Y Combinator from observing thousands of applications and batch companies.
Anatomy
What makes a tarpit idea distinct from a merely bad idea:
| Property | Bad idea | Tarpit idea |
|---|---|---|
| Initial feedback | Negative | Positive |
| Prior attempts | Few | Many (often decades) |
| Detectability | Obvious | Requires research |
| Failure mode | Obvious at pitch | Emerges after launch |
The tarpit is a trap because it passes early validation. Users say they want it; a few will sign up; the product can reach a prototype stage. The structural problem (network effects won’t form, willingness to pay is too low, the job-to-be-done is too infrequent) only becomes visible after substantial investment.
Examples
- Friend/social-coordination apps (where to go out tonight, where to meet up): attempted since the 1990s; generates high initial enthusiasm; usage evaporates rapidly.
- Foursquare clones: all anyone built for years after Foursquare’s rise; almost none survived; Foursquare itself pivoted to B2B.
- Music discovery: high initial engagement possible; structural economics (licensing costs, low willingness to pay for discovery alone) make sustainable businesses extremely rare.
Diagnostic questions
Before committing to an idea:
- Has this been tried many times before? (Search for prior companies in this space.)
- Do you get positive feedback suspiciously easily? (Easy validation is a warning sign, not a green light.)
- Can you specifically explain why all prior attempts failed in a way that is uniquely true of you or your approach? If not, you may be in a tarpit.
Connection to information diet
Tarpit ideas cluster around popular information flows. Founders reading the same podcasts and Twitter feeds end up with the same ideas — including the same tarpits. Diversifying your information diet (personal expertise, unfashionable domains) is one way to avoid the tarpit cluster. See Dalton Caldwell on Startup Survival, Pivots, and Tarpit Ideas.
Related
- Dalton Caldwell on Startup Survival, Pivots, and Tarpit Ideas — primary source
- Annie Duke on Better Decisions, Kill Criteria, and When to Quit — kill criteria for recognising when to exit a tarpit
- Sean Ellis on Product Market Fit and Growth — PMF diagnostic; low PMF score may indicate a tarpit
- Christopher Lochhead on Category Design, the Better Trap, and Languaging — “The Better Trap”: competing in existing categories is an adjacent failure mode