Digital tools play a dual role in an experiment: to assist farmers to perform farm ope- rations efficiently, and to create and combine data to add value to the farm produce. We present here an original case of digitally-enabled on-farm experimentation, which was not conducted in fields but rather at the farm and value-chain levels, through the creation of an integrated ecosystem of farm-level digital services for individual farmers to better market grain. This is akin to the iterative development of a prototype but de- serves the term experiment because of the multiple functions and variables that were investigated within real systems setups. The digital system presented here to inform grain quality on farm was built as a fede- rated system infrastructure where a data intermediary service delivered data between the participating digital systems for reaching the farmer’s target. We found that the experiment validated the approach that could be replicated to other arable production and contribute to solving the challenges of siloed data faced in agri- culture. However, this type of ecosystem needs extra time to reach its full functional capacity. A single partner delivering digital service or data may face difficulties to see its role in far- mer’s value chain and to understand how the ecosystem benefits its own business. The ecosystem requires so-called meta-agents to orchestrate the ecosystem, to promote partners’ commitment, and to maintain adequate development speed.