Commitment products are widely regarded as a remedy for self-control problems. However, imperfect knowledge about one's preferences implies that individuals may fail to anticipate their behaviour under commitment, and consequently choose ill-suited commitment contracts. I conduct a randomised experiment in the Philippines, where low-income individuals were randomly offered a regular-instalment commitment savings product. Individuals chose the stakes of the contract (a default penalty) themselves. A majority appears to choose a harmful contract: While the intent-to-treat effect on individuals' bank savings is large, 55 percent of clients default on their savings contract. A possible explanation that is supported by the data is that the chosen stakes were too low (the commitment was too weak) to overcome clients' self-control problems. Both take-up and default are negatively predicted by measures of sophisticated hyperbolic discounting - suggesting that partial sophisticates adopt weak commitments and then default, while full sophisticates are more cautious about committing, but better able to choose incentive-compatible contracts.