Effective and Scalable Programs to Facilitate Labor Market Transitions for Women in Technology

Women remain underrepresented in technology despite many support programs, partly due to limited evidence on which interventions work and how to allocate scarce program slots. We evaluate two approaches to facilitating technology-sector transitions for women in Poland: Challenges, a scalable and cheap-to-operate ($15/person) online program developing job-relevant portfolios (designed and implemented during this research), and one-on-one mentoring. The programs address distinct labor market barriers: expanding professional networks and teaching their effective use (Mentoring) versus providing credible skill signals (Challenges). Randomizing oversubscribed admissions, we find both programs substantially increase technology employment twelve months post-completion, Mentoring by 15 percentage points (p.p.) and Challenges by 11 p.p. Exploiting treatment effect heterogeneity, we show that algorithmic targeting based on predicted treatment effects achieves outcomes 86% larger than random assignment at current capacity levels. While mentors successfully identify high-benefit candidates (outcomes 48% above random selection), algorithmic allocation using observable characteristics yields an additional 11% improvement. The results demonstrate that (i) relatively low-cost interventions can meaningfully address gender gaps in technology, and (ii) data-driven allocation rules can substantially enhance program effectiveness even when expert judgment performs well.