A leadership programme that works in one country cannot simply be copied into another. The factories, the workforce and the norms are different. That was the challenge when the International Finance Corporation (IFC) and a global apparel partner decided to bring a women's leadership accelerator from Vietnam to Indonesia. The Vietnam pilot offered a strong starting point. It trained 50 high-potential women in junior management through a hybrid of face-to-face and online sessions. Participating firms and individuals reported strong improvements, and almost one in five participants received a promotion within three months of the programme's end. Before replicating it, IFC asked Innovision to find out how the model should be adapted. The scoping study set out a series of design questions. Which position should be targeted to bring about transformational change? Which soft skills are essential? What should the training content be? Should delivery be offline or blended, onsite or offsite? Who should take part? What incentives and nudges might be required? And what follow-up and exit strategy would sustain the impact? To answer them, the team carried out desk research and formative qualitative research in four Indonesian factories, with in-depth interviews, focus group discussions and key informant interviews across operators, line leaders, supervisors and senior managers. The findings were presented in a design workshop to discuss the framework. The design implications were practical. The factories differ in location, workforce make-up, approach to developing workers' skills and gender ratio, so the programme had to be flexible. Some factories have strong internal training systems, while others are still catching up. Career ladders differ from factory to factory, so the accelerator had to be placed carefully on each. The programme that followed was developed with a local university partner and delivered over six months to a pilot cohort of high-potential women in apparel and footwear manufacturing, with a certification for those who completed it. The broader lesson is about scale. Replication works best when it is treated as a design exercise, not a copy-and-paste. Evidence about the local context, collected before the programme begins, saves time and improves results afterwards.