Short Abstract

A feasibility study of Restart, a multi-agency domestic abuse programme, showed how evaluators can add value even without full recruitment by supporting system learning, strengthening programme design and improving conditions for future evaluation.

Description

Feasibility studies rarely unfold as intended, particularly for complex domestic abuse programmes operating within dynamic multi-agency systems. This presentation uses the evaluation of Restart, delivered by the Drive Partnership and Cranstoun, to show how meaningful insights and value can still be generated even when formal study recruitment is not achieved.

Restart takes a multi-agency, whole family approach to hold perpetrators accountable for change, to prevent escalation of risk and helping (ex-) partner and child victim-survivors remain safe and together at home. It brings together professionals from Children’s Social Care, housing, and domestic abuse sector services to identify, change, and disrupt patterns of harmful behaviour at an early stage.

Our role evolved during the study from testing feasibility for a traditional impact design to acting as a learning partner, focused on understanding what was helping or hindering delivery, strengthening programme design, and identifying the conditions needed for robust future evaluation. Working closely with delivery partners, we explored where Restart sat within local systems, what was enabling or constraining implementation, and which model adaptations were needed before progressing to an outcomes-focused study. Key insights from our study included:

Mission drift around Restart’s aims and adaptations to Restart’s model, highlighting the need to refine the Theory of Change, clarify the programme’s vision, codifying its core delivery elements, and consider extending the intervention’s timeframe.

Engagement and retention barriers, including practitioner hesitancy around thresholds, competing priorities and uncertainty about Restart’s place within local pathways. There was also inconsistent understanding of eligibility criteria for Restart which influenced communication of the target cohort to referrers.

Challenges to embedding evaluation processes within delivery and barriers to recruiting participants to the research study

System-level enablers and barriers, especially the influence of strategic priorities, leadership visibility and alignment across CSC, Early Help, Housing and VAWG teams.

Rather than treating these issues as limitations, we used them as catalysts for programme and system learning. We brought together insights from multiple stakeholders to help partners understand the wider conditions shaping delivery. Our final report set out what is, and is not, currently feasible to evaluate, alongside clear recommendations for strengthening programme clarity, referral pathways, data systems and practitioner confidence in using validated tools.

While formal funding for Restart has now ended, the feasibility study has laid strong foundations for future collaboration between evaluators and practitioners. Partners identified several areas that would benefit from further development, including:

refining programme aims and mechanisms through further Theory of Change work;

strengthening eligibility and informed consent processes;

improving data monitoring systems to support future evaluability;

piloting outcomes measurement tools; and

further embedding meaningful Expert by Experience involvement in programme design and practitioner support.

This case study illustrates how evaluators can add significant value, even when recruitment to research is limited, by taking an adaptive approach that prioritises system learning, programme development and evaluability. It highlights the importance of evaluation as a developmental tool for shaping policy and practice within complex social systems.