—In wireless sensor networks (WSNs), as in every other discipline, people willing to evaluate the performance of an application or a protocol rely on modeling, simulation or experimentation. Simulations and models produce results for large-scale networks in a reasonable time, but trade representation accuracy for speed and hence ignore many physical and system effects, such as interference from the outside world or race conditions inside the nodes. Experimentation provides more representative and precise results, but is limited to small networks. Besides, they require more effort to be deployed and to collect results. These approaches are therefore complementary and should all be involved in the evaluation, which is seldom true, as it requires duplicating the deployment and data collection processes. In this demonstration, we present MakeSense, a framework that simplifies these tasks for both simulation and real experiments environments by creating a whole experimentation chain from a single JSON description file. By using MakeSense, it is possible to organize the compilation, to orchestrate the firmware deployment, to efficiently collect results and to plot statistics. We illustrate the ease of use and efficiency of the complete MakeSense workflow over a simple RPL-UDP deployment scenario evaluated with the Cooja simulator and the FIT IoT-Lab open testbed.