Developing representative consumer decision models has eluded marketing modellers, making choice prediction problematic. Single and two-phase decisions are empirically supported with three-phases mooted. Further empirical research, beyond two phases, has been hindered by a lack of clarity of decision phase boundaries. This paper presents a methodology to capture decision phase boundaries and reports initial empirical results revealing that current consumer decision models exclude almost half (49%) of consumer decision processes. We propose a new 'decision waves' model of consumer decision-making that is inclusive of multiple decision phases.