T cells engineered to express chimeric-antigen receptors (CAR-T cells) can effectively control relapsed and refractory haematological malignancies in the clinic. However, the successes of CAR-T cell therapy have not been recapitulated in solid tumours due to a range of barriers such as immunosuppression, poor infiltration, and tumour heterogeneity. Numerous strategies are being developed to overcome these barriers, which include improving culture conditions and manufacturing protocols, implementing novel CAR designs, and novel approaches to engineering the T cell phenotype. In this review, we describe the various emerging strategies to improve CAR T cell therapy for solid tumours. We specifically focus on new strategies to modulate cell function and fate that have precipitated from the growing knowledge of transcriptional circuits driving T cell differentiation, with the ultimate goal of driving more productive anti-tumour T cell immunity. Evidence shows that enrichment of particular phenotypic subsets of T cells in the initial cell product correlates to improved therapeutic responses and clinical outcomes. Furthermore, T cell exhaustion and poor persistence are major factors limiting therapeutic efficacy. The latest preclinical work shows that targeting specific master regulators and transcription factors can overcome these key barriers, resulting in superior T cell therapeutic products. This can be achieved by targeting key transcriptional circuits promoting memory-like phenotypes or sustaining key effector functions within the hostile tumour microenvironment. Additional discussion points include emerging considerations for the field such as (i) targeting permutations of transcription factors, (ii) transient expression systems, (iii) tissue specificity, and (iv) expanding this strategy beyond CAR-T cell therapy and cancer.