While blockchain-oriented software engineering typically aims to ensure the correctness of smart contract code, the correctness or, more generally, the quality of input data is often overlooked. This paper advances the idea of controlling the transaction payloads in the transaction validation phase, to prevent the poor data quality issues that can affect the output of even correct smart contracts. We study how to extend state-of-the-art transaction validation logics, highlight emerging requirements, and outline future research challenges.