Big data provides high volume of data to inform product customisation. Understanding which data is relevant remains a challenge. A method is proposed to identify relevant data to inform data-driven customisation. A case study regarding customisation of orthoses was conducted. Verbal protocol analysis was employed to extract time spent on major fabrication phases. Data related to patients, therapists and fabrication time was analysed. Results showed that the number of stabilised joints, experience of therapists and whether the design is for in- or out-patient are key factors for customisation.