Rationale: When evaluating complex, tailorable digital behavioural interventions, additional approaches may be required alongside established methodologies such as randomised controlled trials (RCTs). Research evaluating computer-assisted therapy (CAT) programme for substance misuse, Breaking Free Online (BFO), is informed by Medical Research Council (MRC) guidance recommending examining 'mechanisms of action' of individual intervention strategies, which is relevant when evaluating digital interventions with content that may evolve over time. Aims: To report findings from examination of mechanisms of action of tailoring advice within the BFO programme and outcomes from specific intervention strategies. Methodology: Analysis of variance and linear regressions were used to assess intervention completion data, and psychometric and clinical outcomes, for 2311 service users accessing drug and alcohol treatment services across the UK. Results: Tailoring advice provided to users appeared to prompt them to prioritise completion of intervention strategies associated with their areas of highest biopsychosocial impairment. Completion of specific intervention strategies within BFO were associated with specific clinical outcomes, with a dose response also being found. Mechanisms of action analyses revealed the primacy of cognitions, with cognitive restructuring strategies being associated with improvements in mental health, severity of substance dependence, quality of life and global biopsychosocial functioning. Conclusions: The MRC framework provides an evolved research paradigm within the field of digital behavioural change. By assessing baseline profiles of need, BFO can target the most appropriate clinical content for individual users. Mechanisms of action research can be used to inform modifications to BFO to continually update clinical content and the technology platform.