Household decisions such as adjusting lighting, configuring appliances, and managing waste often rely on abstract or invisible data, making it challenging for non-expert users to make informed choices. While immersive analytics holds promise for embedding data directly into users' environments, most current systems assume a one-size-fits-all model of data presentation. In this paper, we propose a layered immersive analytics framework designed specifically for household decision-making. Our proposed framework supports adaptive visualization across three levels of user engagement: overview (simplified summaries), detail (interactive data exploration and decision simulation), and contextual (narrative-based feedback). These layers are intentionally mapped onto appropriate immersive technologies - augmented reality (AR) for in-situ awareness, virtual reality (VR) for analytic and experiential reasoning, and mixed reality (MR) for consequence visualization and storytelling. We illustrate this design through realistic household scenarios and argue that such a cross-immersive layered system offers a compelling direction for human-factored immersive analytics.