The most frequent is poorly scoping the initial question, meaning the functional unit: "serve 100 hot drinks over one year" can be modelled; "compare 2 cups" cannot. Next come using global averages instead of actual data, which distorts the result and places the product in the most penalising class; launching a full LCA without prior screening; omitting third-party critical review when results will be communicated; and filing the report in a drawer with no ecodesign decision afterwards.
Read the source article: First product LCA: steps, cost and mistakes to avoid