Nowadays, the search for novel active materials for laser devices is proceeding faster and faster thanks to the development of innovative materials able to combine excellent stimulated emission properties with low-cost synthesis and processing techniques. In this context, amplified spontaneous emission (ASE) properties are typically investigated to characterize the potentiality of a novel material for lasers, and a low ASE threshold is used as the key parameter to select the best candidate. However, several different methods are currently used to define the ASE threshold, hindering meaningful comparisons among various materials. In this work, we quantitatively investigate the ASE threshold dependence on the method used to determine it in thin films of dye-polymer blends and lead halide perovskites. We observe a systematic ASE threshold dependence on the method for all the different tested materials, and demonstrate that the best method choice depends on the kind of information one wants to extract. In particular, the methods that provide the lowest ASE threshold values are able to detect the excitation regime of early-stage ASE, whereas methods that are mostly spread in the literature return higher thresholds, detecting the excitation regime in which ASE becomes the dominant process in the sample emission. Finally, we propose a standard procedure to properly characterize the ASE threshold, in order to allow comparisons between different materials.