Five-Day Marketing and Sales Programme for Refugee Women: Outcomes, Evidence and Recommendations for Future Cohorts (ARWO Cohort 1 Evaluation Report)

Abstract

This report evaluates the first cohort of a five-day marketing and sales training programme delivered to fifteen refugee women in Kuala Lumpur, Malaysia, in September 2026 by the Afghanistan Refugee Women's Organisation (ARWO). The programme was designed to convert existing production skills — soap-making, candle-making, baking, tailoring and hairdressing — into income by teaching pricing, customer identification, sales communication, mobile-first digital marketing, generative AI tools, packaging and labelling, culminating in a supervised market-testing day with real customers. The evaluation applies Kirkpatrick's four-level training evaluation model (Reaction, Learning, Behaviour, Results) to a structured bilingual Dari–English post-course survey completed by all fifteen participants (100% response rate). Two composite indices are reported: a Programme Quality Index (95.2/100) and a Skill Confidence Index (83.1/100, Cronbach's α = 0.92), alongside pre-set key performance indicators. Headline results: 80% of participants completed a sale or confirmed booking on market day against a target of 70%; seven of the eight participants who had never sold anything previously made their first-ever sale; 87% reported increased confidence in selling; and 100% used generative AI for at least one marketing task. Two digital-adoption targets were not met, for a documented structural reason — most participants hold a single mobile number shared with family communication, which cannot simultaneously run a personal and a business WhatsApp account. The report documents two further findings of wider relevance. First, refugee-related mobility and safety constraints required the selection of safer but lower-footfall market-testing venues, and the evidence indicates that sales volume reflected constrained market access at least as much as participants' selling ability. Second, 20% of respondents read the right-to-left rating grid in reverse, a measurement artefact with implications for survey design in right-to-left languages. The report includes the full curriculum, a competency rubric, the evaluation instrument, item-level statistics and nine evidence-linked recommendations. It is intended as a transferable model for livelihood and economic-empowerment programming with displaced populations.