Networks are pervasive for complex systems modeling, from biology to language or social sciences, ecosystems or computer science. Detecting com- munities in networks is among the main methods to reveal meaningful struc- tural patterns for the understanding of those systems. Although dozens of clustering methods have been proposed so far, sometimes including parame- ters such as resolution or scaling, there is no unified framework for selecting the method best suited to a research objective. After more than 20 years of research, scientists still justify their methodological choice based on ad-hoc comparisons with ‘ground-truth’ or synthetic networks, making it challenging to perform comparative study between those methods. This paper proposes a unified framework, based on easy-to-understand measures, that enables the selection of appropriate clustering methods according to the situation. If re- quired, it can also be used to fine-tune their parameters by interpreting them as description scale parameters. We demonstrate that a new family of algo- rithms inspired by our approach outperforms a set of state-of-the-art com- munity detection algorithms, by comparing them on a benchmark dataset. We believe our approach has the potential to provide a fresh start and a solid foundation for the development and evaluation of clustering methods across a wide range of disciplines.