Summary: Supplier involvement in new product development has become increasingly common. When developing an innovative component for a new product, it may be sufficient for the component supplier to merely follow the buyer’s specifications and release the component without a good understanding of how well it will fit the final product. While this approach meets reasonable expectations of supplier involvement, it may increase the probability of the final product’s failure. The alternative for the supplier is to exceed expectations by learning more about the final product, performing extra tests, and gathering more information about component fit. This paper studies supplier incentives to exceed expectations before releasing the component for mass production. We develop a sequential non-cooperative game with endogenous information asymmetry. We then solve the game for reward-only and residual claimant contracts and compare the outcomes against the first-best outcome. We find that offering a higher reward to the supplier toward component success does not always lead to greater effort on the supplier side. Instead, this approach may backfire on the buyer by making the supplier less likely to test the component and, in some cases, more likely to release a low-quality component. We also find that introducing a penalty for component failure does not lead to coordination of the supply chain either. However, an efficient contract exists, but it requires the buyer to introduce a penalty for non-release of the component.