Social scientists have argued that a treatment imperative shapes experiences of biomedicine. This is evident within oncology, where discourses of hope are tempered by persistent fears surrounding cancer. It is within this context that genomic decision‐making tools are entering routine care. These may indicate that a treatment is not appropriate for a particular disease profile. We draw on qualitative interviews and observations centred on gene expression profiling to consider the implications of this technique for the treatment imperative in early breast cancer. Influenced by sociological perspectives on medical technologies, we discuss how fallibilities of established tools have forged a space for the introduction of genomic testing into chemotherapy decision‐making. We demonstrate how high expectations shaped patients’ interpretations of this tool as facilitating the ‘right’ treatment choice. We then unpick these accounts, highlighting the complex relationship between gene expression profiling and treatment decision‐making. We argue that anticipations for genomic testing to provide certainty in treatment choice must account for the sociocultural and organisational contexts in which it is used, including the powerful entwinement of chemotherapy and cancer. Our research has implications for sociological perspectives on treatment decision‐making and clinical expectations for genomic medicine to resolve the ‘problem’ of overtreatment.