To support public safety and risk management in snow-covered mountain regions, avalanche forecasts must deliver reliable and credible information on avalanche conditions. To promote greater consistency in avalanche danger level assessment across European avalanche warning services, a revised version of the EAWS Matrix – a structured look-up table that combines snowpack stability, the frequency of snowpack stability, and avalanche size to determine the regional danger level. Developed through expert elicitation, the Matrix includes cells with a single suggested danger level, while about half also display a second option, reflecting cases where a substantial minority of experts proposed a different level. We analyzed its operational use over the first three winters following implementation by 26 European avalanche warning services. Most Matrix cells were predominantly associated with a single danger level in operational use, suggesting potential for structural simplification. However, two cells – <em>poor–some–size 2</em> and <em>very poor–some–size 3</em> – acted as transition zones with substantial overlap between adjacent danger levels and divergent use across services. Assessments with finer granularity – such as sub-classes within the predefined Matrix categories – revealed meaningful tendencies within coarse classes and may help preserve critical nuances in expert judgment. Moreover, incorporating observed tendencies within these classes may enable more targeted guidance on when to assign the higher or lower of the two danger levels shown in the Matrix. Several Matrix cells remained rarely used, supporting the use of white shading to indicate uncertainty or implausibility in danger level assignment. While Matrix application was relatively consistent for dry-snow problems, marked inconsistencies emerged for wet- and gliding-snow problems, particularly in the classification of snowpack stability for the latter ones. These findings underscore the need for community-wide discussion and alignment of stability assessment practices and offer insights for refining both operational avalanche danger assessment and the Matrix itself. However, because neither the danger level nor its input factors can be independently measured, a true validation of Matrix performance remains out of reach. This study forms part of the iterative development process described in detail in the companion paper by Müller et al. (2025).