Innovations can be seen as chains of non-routine decisions. With each decision, the innovator has to assess how important the various decision attributes are. Because the decisions are non-routine, innovators cannot fall back on judgements of past importance. Most decision support methods elicit importance judgements but do not help innovators or other decision-makers with the mental processes leading to the judgment. The ‘importance assessment process’ can be divided into seven phases (such as (sub-)attribute processing and various forms of weighting). The phase ‘(sub)-attribute processing’ is the most important phase in terms of effort devoted to it, and the most obvious pitfalls that prevent valid importance assessments appear in this phase. This article describes some of these pitfalls. A few simple instruments may provide better-founded importance judgements that can be better communicated to other actors involved in innovation processes.