The views expressed in this article are those of the authors. Publication in Gates Open Research does not imply endorsement by the Gates Foundation.

Adolescent girls represent 16% of all women of reproductive age (15–49) ¹ . Adolescence is a critical moment for intervention, as it is a period that establishes future behaviors and patterns ² and an important window before discriminatory gender norms are fully entrenched ³ . Adolescents require age-appropriate information, opportunities to develop key life skills, and essential health services in a safe and supportive environment to grow, develop, and pursue healthy futures. Increasing access to tailored, adolescent-focused services and expanding opportunities for adolescents to engage in the design and delivery of interventions related to their own health are critical components of a rights-based approach to adolescent sexual and reproductive health (ASRH) programming ⁴ . Yet, the specific needs of adolescent girls are often neglected or underemphasized in global, regional, and national-level priorities.

As girls transition into young women, they face many barriers including school dropout, child marriage, early pregnancy, physical and mental health problems, and gender-based violence ⁵ . Lack of sexual and reproductive health (SRH) knowledge and access to comprehensive SRH services, and cultural norms that promote early childbearing contribute to high rates of unplanned pregnancy among adolescents ⁶ . Adolescents aged 15–19 have a low contraceptive prevalence rate (CPR), estimated to be around 10.2% and have double the unmet need of all women of reproductive age (43% vs. 24%) ¹ . Simultaneously, discriminatory gender norms and patriarchal structures limit adolescent girls’ ability to achieve economic well-being. These barriers are built-in from an early age and constrain the extent to which girls can gain the skills and experiences necessary to pursue improved economic outcomes and economic autonomy. As adults, women are less likely to participate in the paid labor force compared to men and are more likely to be unemployed and over-represented in informal and vulnerable employment sectors ⁷ .

There is a growing evidence base demonstrating adolescent girls’ need for a holistic package of services that can offer reproductive health information and services, build their agency, and establish a positive enabling environment for them to pursue their life goals ⁸ . There is increasing interest from the broad community of practice (donors, implementers, and local stakeholders alike) in addressing adolescent girls’ needs through programming that combines SRH with age-appropriate economic empowerment. For adult women, evidence suggests that reproductive health improvements can lead to improvements in economic empowerment ^{9– 11} . Yet, for adolescents, the evidence on the interplay between economic empowerment and SRH is more complex. There is some information, albeit limited, on successful programmatic approaches that integrate components of SRH into economic programming for adolescents ^{12, 13} . Some evidence also shows that interventions which have sought to prevent early and forced marriage, as well as early childbearing, have had significant impact on the economic empowerment of adolescent girls and young women (AGYW) ¹⁴ . Yet other research demonstrates diluted effects when combining SRH and economic empowerment components in programming for adolescent girls ¹⁵ . Overall, there is a growing body of evidence on the types of interventions which demonstrate significant and meaningful impact on adolescent girls’ economic empowerment ¹² .

Across all geographies, the objective of the design process was to adapt the existing SRH interventions to offer participating girls the opportunity to further identify and progress towards their economic goals. As part of our intentional pursuit of gender transformative programming, we aimed to design age-appropriate program adaptations that not only supported girls to gain additional income, savings, and/or assets but did so in a way that built their self-efficacy and economic autonomy. Given the growing evidence around multi-component approaches that pursue both SRH and economic outcomes, we also wanted to understand how, if at all, these more intensive and intentional economic empowerment components would influence contraceptive uptake and continuation among participating adolescent girls.

The HCD timeline and process was similar across geographies, with some notable differences. We followed the same three overarching design phases as we pursued in our formative design phase for A360’s SRH interventions in 2016 – insight gathering, rough prototyping, and live prototyping. Within each of these phases, potential program participants and other key stakeholders who might interact with the final intervention – such as mothers or husbands – were invited to participate in the testing and give feedback on the evolving concepts. To ensure we included youth perspectives and to help “bridge” the gap between adult researchers and adolescent girl participants, 18–24-year-olds were hired and trained as Youth Innovation Officers (Nigeria / Tanzania) or Young Designers (Ethiopia). These young people participated as full members of the design teams, leading research conversations and prototype testing in the field.

We partnered with IDEO.org as our HCD technical assistance partner. The COVID-19 pandemic limited in-person engagement and as a result IDEO.org provided design strategy and facilitation remotely, guiding the design and implementation teams in Nigeria, Tanzania and Ethiopia who drove the design process in country. In Ethiopia, our design work began in August 2020 compared to January 2021 for the remaining geographies as the result of a complementary investment from Maverick Collective philanthropists. However, in Ethiopia we also had a more intentionally limited engagement from IDEO.org and as a result our design process converged with that of Nigeria and Tanzania in July 2021 during live prototyping. Across all design phases, we engaged 679 adolescent girls (169 in Ethiopia, 96 in southern Nigeria, 156 in northern Nigeria, and 258 in Tanzania) along with 551 other stakeholders (88 in Ethiopia, 149 in southern Nigeria, 196 in northern Nigeria, and 118 in Tanzania) ( Table 1). Quotes from these engagements are provided in this Open Letter to substantiate our learning and experience. The raw data set containing these quotes is currently unpublished, but a synthesis of the data from this design process will be made publicly available at the end of 2023.

In acknowledgment that economic empowerment programming is not one of PSI’s core technical competencies, throughout the design process, A360 collaborated with BRAC as a conduit to the global evidence base around girls’ economic empowerment. Their technical assistance role, beginning in 2021, included reviews of all design materials, rapid literature summaries, and development of skills or competency-based curriculum, drawing from their evidence-based Empowerment and Livelihoods for Adolescents (ELA) Model ²⁴ . Their participation in the HCD process was crucial, sharing best practices and evidence that responded to identified needs of adolescents and compiling literature to help guide inquiries that arose in the design process. As the design process evolved, the need for additions to existing A360 curricula content and follow-on support for economic empowerment outcomes became evident. BRAC then engaged in integrating selected ELA content into the A360 curriculum and training staff on its usage. BRAC continued to provide support to our teams as the pilot launched in 2022.

Kore Global provided technical assistance to pursue a gender-intentional design. With IDEO.org they co-developed a gender in design toolkit to guide the process and ensure the resulting program designs were gender responsive.

The result of the design process is interventions that are informed by the evidence and directly responsive to our users’ lived experiences. While each program aims to achieve the same outcomes, namely increases in girls’ income, savings, and assets, and improved economic autonomy, the specific pathways differ by geography. All programs begin with their respective SRH intervention. Girls are then offered the chance to continue to participate in a more intensive economic empowerment program.

In Ethiopia, girls are brought into the economic empowerment program, Smart Steps ( Figure 1), through a Smart Start counseling session which focuses on how delaying and spacing pregnancies can support girls and couples to pursue financial stability ¹⁹ . If they wish to participate in the expanded economic empowerment program, they are invited to a ‘ future mapping’ session, first jointly with their husbands and then individually. During these sessions girls are supported to identify their medium- and long-term goals and what activities they can take to achieve those goals. Girls’ husbands are supported to understand the value of girls’ economic and market participation in accelerating progress towards achieving their joint goals as a couple ( Figure 2). Girls are then gathered into an adapted version of Village Savings and Loan Associations (VSLA), called a ‘Step Up Association.’ These associations are for married adolescent girls only and self-governed by participants, with support from a program-hired mentor. Each week, the associations meet and provide a platform for girls to collect savings from each member. In our adapted program, girls also go through a participatory curriculum specifically designed for low-literacy audiences to teach essential soft-skills and business skills. After four or five weeks of meetings, girls can choose to access loans from the pooled savings to fund investments in existing businesses or to begin new income-generation activities. Groups are eligible for a matching grant after the first 16 weeks if: a) all members achieve 80% attendance in the initial period; b) the group has saved a minimum amount; and c) they have complied with their constitution and maintained proper bookkeeping. If the group meets these criteria, they are eligible to have their total savings matched at 150%. This money is transferred to the group as a whole and is then available as part of the loan pool.

In both northern and southern Nigeria, the heart of the economic empowerment programs, MMA+ and 9ja Girls+ respectively, is a scaffolded package of soft, business, and vocational skills training combined with mentorship and coaching ( Figure 3 and Figure 4). Communities are oriented to the aims of the joint SRH and economic empowerment program and girls are mobilized, either through referral from their key influencers (primarily parents for unmarried girls and husbands for married girls) or directly by program mentors. They are invited to attend an initial set of four curriculum sessions, each lasting 90 minutes, that focus on developing health-related knowledge and core soft skills such as decision-making and communications. After completing these initial sessions, girls are invited to move into a secondary curriculum that provides them with a combination of business and technical or vocational skills. After the secondary curriculum, girls can elect to learn one to two vocational skills through either a vocational training center (northern Nigeria) or apprenticeships (southern Nigeria). Following the skills trainings, girls are then given mentorship support to develop and execute a business plan. Their participation culminates in a large, public graduation that doubles as a marketplace for girls to display the products and services from their newly gained skills. This is a chance for the community to show public support for the program and for girls to show to their husbands or parents the value of what they have learned through their participation. The program models in southern and northern Nigeria, although serving different girls, are overall very similar.

In Tanzania, the economic empowerment program , Kishua Academy, provides a similar scaffolded skills-building experience as in Nigeria but incorporates additional elements of peer mentorship and partnership with TVET institutes ( Figure 5, Figure 6). Girls are recruited through the SRH intervention, Kuwa Mjanja, which implements community-based contraceptive service delivery events geared towards adolescents. Girls sign up for the Kishua Academy, are introduced to a cohort of their peers with whom they will be receiving skills training, and meet their ‘Career Dada,’ a peer mentor with whom they can build trust and have open discussions about their business goals, relationships, and SRH needs. They progress through a series of soft and business skills sessions and then are invited to join a specific course at a local TVET institute. Their engagement during and after these vocational skills training sessions includes opportunities to apprentice with ‘Career Mamas’ who are experienced female entrepreneurs within the community who can mentor and support girls to progress. Girls who opted to learn skills which were predominantly dominated by men in the community were sometimes placed with a ‘Career Baba’, an experienced businessman in the community who could mentor them in their chosen skill. After completing their courses, graduating cohorts transition to local government oversight which supports them (through Youth Development Officers / Community Development Officers) to apply for loans to grow their businesses through the local government-run youth development funding program. There is a moment post-graduation for the community to gather to celebrate girls for their accomplishment and participation in the program.

In Nigeria and Ethiopia, work moved forward post-design into pilot, allowing for additional refinement, while work in Tanzania ended after live prototyping and no additional refinement was conducted.

Across the design phases, we used the evidence base to inform and contextualize our design process and findings. In doing so we discovered nuance in how what we know from the evidence base corresponded with the insights we were drawing directly from our program users. We have described three overarching categories which depict how we used the existing evidence and new user insights to strengthen our design process. Though we describe these categories separately, in practice there was often overlap.

1. The evidence base allowed us to expedite finding a solution that worked for our users

We mined existing systematic reviews to identify promising practices for improving outcomes for adolescent girls around economic empowerment, self-efficacy, and health knowledge ^{12, 14, 25– 29} . We prioritized practices that had specific evidence with adolescent populations to build prototypes that could be trialed within our geographies. In the early phases, we intentionally designed inquiry tools and rough prototypes to validate key elements of the evidence base. By speaking directly with adolescent girls and their influencers, we could confirm that they felt these elements were important and obtain additional details on how girls wanted to experience them in practice ( Table 2). This made the process of finding something which worked for our users faster and more efficient. We encountered some limitations in that adolescent girl-focused economic empowerment programming is not traditionally well evaluated, though this is changing with increased investment in this space.

In general, we understood the practices that have the greatest established efficacy, with some limitations acknowledged, to be a) group-based with opportunities to build solidarity and peer trust; b) mentor-led, particularly by mentors who understand the community context; c) combining vocational training, and business and soft skills development; d) introducing concepts around gender and power to shift girls’ critical consciousness; and e) inclusive of follow up after the initial training, such as through mentorship or coaching ³⁰ . These components are often most effective when implemented in combination. Additionally, in some contexts we prioritized savings groups, given that they are linked to positive impacts on poverty reduction and can serve as a practical platform to implement the promising practices described above. Despite differences in the final program design across geographies due to varying contextual factors, the resulting programs retained most of these core evidence-based program components ( Figure 7).

Mining existing evidence-based practice provided value not just in insight gathering and initial prototype building, but also moving into the latter design phases (additional rounds of prototype testing and live prototyping) when we needed to build out detailed systems maps. When the design team identified a barrier that needed to be overcome to promote program effectiveness, we first consulted the evidence base to understand if there were existing practices that could be adopted to fill gaps in the proposed program design or strategy.

For example, in line with the broader project objectives to promote sustainability for our program interventions, we wanted to explore pathways to sustainable scale. We examined the evidence base on scaling economic empowerment programs for AGYW and identified key lessons. We learned that good practice required a locally grounded approach with more than one scale pathway, responding to local contexts and generating local ownership ⁴¹ . We recognized that, unlike our SRH interventions, there was not likely to be one partner or government ministry that would have the capacity to wholesale adopt these designs given the cross-sectoral nature of these interventions. Additionally, there are limited resources within ministries to take up these solutions as they are necessarily holistic and need to be delivered with integrity, making them time-intensive and requiring collaboration across multiple stakeholders. This informed our landscaping of potential pathways to scale during live prototyping and helped us focus on understanding which actors were best placed for what during the scale-up process.

As another example, in live prototyping in Ethiopia we noted that though girls showed interest in participating in savings and loans groups, they felt limited by their small savings pool, which constrained the amount of the loans available. This was affecting program attendance and meant girls were missing out on critical soft-skills training sessions. Working with BRAC, we learned about ways other organizations had managed similar challenges. After exploring the options and weighing them against our program priorities, we decided to institute group-level matching grants which groups were eligible to receive contingent on consistent savings and attendance. This was based on an approach successfully implemented by a BRAC partner in Rwanda ⁴² .

2. At times there was a disconnect between what we knew worked in the evidence base and what our users said they wanted.

Not all attempts to test existing promising practices with our users elicited positive reactions. This is to be expected – adolescent populations are not homogenous, and resonance of these program components is likely to vary across geographic contexts and specific adolescent sub-populations (older vs. younger adolescents, married vs. unmarried, etc.). However, a challenge in HCD can be that users may prioritize what is known or comfortable. If insight gathering and prototyping are not done well, this can lead to users reporting a preference for, at best, the status quo, and worst, harmful traditional practices, as they are the norm. In this instance, having the evidence-base in mind encouraged us to persist in pursuing promising practices and continue learning to understand the barriers that underlay resistance to what the evidence says works. In doing so, we better understood how to effectively apply the evidence base ( Table 3).

3. New insights allowed us to build a greater understanding of our users’ lived experiences where there were existing gaps in the evidence base

The evidence on promising practices for girls’ economic empowerment, though growing, is not exhaustive. In conducting our insight-gathering phase, we were able to uncover additional detail on the lived experiences of our program users that did not feature prominently in the evidence base. While the overall themes of the insights drawn from the design process are not new, the context-specific nuances were important for us in program development and can be valuable contributions to the existing body of evidence.

Increasing resilience through a diversity of skills

In the evidence base we saw that most programs focused on technical skills paired with some business or soft skills. Yet, our research demonstrated an enormous appetite by girls to learn a diverse set of skills that could provide them with a competitive advantage. Girls did not want one-dimensional training; they were eager for continuous upskilling that would allow them to weather shifts in the market and other situational factors, such as the impacts of climate shocks or conflict. They were eager to learn skills that could be layered together for greater profit, such as hairdressing and make-up or poultry rearing and fish farming. Best practice and research, in our exploration of the literature, did not address the need for young female entrepreneurs to have a set of skills that could allow them to adapt to seasons and trends to ensure continuous income throughout the year and in a longer-term time horizon. Designing a skills curriculum around this type of ‘resilience’ could also become even more critical as the impacts of climate change are felt more acutely within these communities.

Navigating business choices and market access for girls in conservative settings

In certain geographies our design research uncovered greater specificity in the restrictions faced by married adolescent girls in conservative contexts that influence the ways in which they can enter and participate in the market. In northern Nigeria especially, where mixing between the sexes is restricted, girls were often severely limited in options for income-generating activities. In Ethiopia, although there is more interaction between the sexes, the result is much the same, as traditional gender roles mean that girls must be at home to take care of children. Yet, girls demonstrated incredible entrepreneurialism in the face of those restrictions. In Nigeria, girls participated in women-only markets on dedicated days or worked out of their home and relied on husbands or brothers to sell their products. In Ethiopia, girls focused on home-based income-generation activities (such as poultry rearing) or negotiated with extended family members to shift childcare responsibilities. These insights point to how adolescent girls, and women in general, can be successful in income generating activities in conservative settings. Nonetheless, the social practice of sex-based restrictions (e.g. male only markets) has a significant impact on entrepreneurship opportunities and the ways these girls must navigate business. This uncovered a need for further research and best practices on how women can work within these existing restrictions even while systems and norm change are in process.

Low trust in institutional support structures

While we expected to find some frustration or disappointment with prior financial inclusion initiatives that had ended or failed to provide the promised impact, the level of mistrust was greater than we anticipated. We were struck by the broad-based level of mistrust expressed by participants stemming from their interaction across non-governmental, governmental, and formal financial institutions. Users pointed to the failures of past NGO projects to promote lasting outcomes. Mistrust of government-run programs often stemmed from corruption, burdensome bureaucracy, or poor management that limited the funds available for community services. Some users, however, reported greater trust in government loan programs, as rates were generally considered favorable. This contrasts sharply with opinions about microfinance institutions. Across geographies, participants shared negative experiences about high interest rates that made loans difficult, if not impossible, to repay.

These negative perceptions raise questions about how NGOs should be interacting with institutional structures to address persistent gaps and promote sustained change. Leveraging existing systems generally presents the greatest avenue for sustainability. However, given existing mistrust and structural deficiencies we question the feasibility of this without significant structural change. Our insights point to a need for the organizations promoting economic empowerment, especially for marginalized groups, to move beyond understanding these barriers to intentionally designing and implementing programming that pursues strengthening of financial institutions in a way which is sustainable and promotes community trust.

Variations in desired and required dosage for achieving economic empowerment

Across documented skill-building programming for adolescent girls, the topics covered, and time dedicated, differed, from a week to over a year long. The research on the appropriate dosage needed to achieve positive outcomes is limited. This leaves organizations to make their best guess about an appropriate timeline and to let budgetary limitations drive decision-making. Dosage is complicated by disparate needs across sub-populations of adolescents. We saw this across our own geographies in how insights led to differences in dosage – in northern Nigeria, our program provided 6–8 hours of SRH and soft skills sessions and 15–20 hours of economic empowerment training including soft and hard skills over the course of 12 weeks. While in Ethiopia, participants met for one and a half hours per week over the course of approximately 30 weeks. We struggled to balance the time girls needed to absorb skills and new information against the other demands on their time. Married girls were busy managing their household and caring for children. Younger unmarried girls were attending school and had a full load of chores at home, and older unmarried girls might already be involved in a basic income generation activity. In cases where girls were unable to leave home due to pregnancy or childbirth, we provided make up sessions to participants. Despite initial resistance mentioned previously to participate in soft skills training, attendance across geographies for these skills sessions was high, demonstrating that localized solutions with attention to the varying needs of different target populations are critical. These insights also point to the value in providing alternative learning pathways for individuals when they encounter challenges in attending regular group meetings for periods of time.

Leveraging joint goal setting and crafting long term engagement plans for girls’ influencers

While the evidence base is clear on the importance of engaging married girls’ husbands to support their participation in economic empowerment activities, there are fewer evidence-based strategies on how to do so effectively. In Ethiopia, where our SRH approach already targets couples, we used an extended goal-setting session to try and build this support. During the session, which included a facilitated conversation between the couple, they identified their long-term goals and how they would each contribute. Using a model couple, we demonstrated how goals could be reached sooner if both the husband and the wife participated in income generation. We took care to note the non-income-generating activities often falling on women (such as childcare or food preparation) that also contribute to the success of the household. At the end of the session, husbands were introduced to the economic empowerment program and asked to support their wife’s attendance. Both husbands and wives responded positively to these sessions, suggesting the value of harnessing shared goal setting and in making explicit the benefits of multiple earners. While this was a positive start, additional contact with husbands was needed. Some husbands required multiple visits from program facilitators to reinforce the importance of their wife’s attendance at weekly association meetings. In addition, it is essential to be clear what participants would and would not receive for participation. Some husbands expected that their wives would receive a cash transfer for participation and grew frustrated when that was not the case. Again, this required additional communication from program facilitators with husbands to resolve. On the other hand, we also saw some husbands who were very supportive of their wife’s participation, as demonstrated through attending sessions to contribute weekly savings if she was unable to attend.

This open letter describes the complexity we encountered while trying to design with responsiveness to the existing evidence and to users’ desires and needs. A challenge we faced was that the evidence for economic empowerment among adolescent girls was not easily translatable to program design. First, the volume of evidence around economic empowerment programs for adolescents often seemed overwhelming. Second, the evidence base has significant gaps in research and practice around girls’ economic empowerment, particularly marginalized communities ¹² . The translation of evidence was exacerbated by the differences between the macro-economic contexts in which we are implementing. For example, a lack of vocational training centers or high levels of unemployment limit the opportunities available and what proven interventions may even be plausible. Additionally, as an organization focused on improving health and not economic outcomes, we were at a disadvantage. We know that girls need – and ask for – multi-sectoral solutions. In the absence of well-established collaboration across sectors, this may require organizations to pursue design and implementation of programs outside their existing wheelhouse.

We found that the HCD process helped us to translate the evidence and make it more easily usable to program designers. For example, as noted above, the insight gathering and prototyping processes allowed us to validate existing insights from the evidence base and hear more from girls about what they wanted to see in practice, for instance the need to help girls get more specific with their goal setting or help them to see the value of soft skills classes. The HCD process also allowed us to fill in gaps in the existing evidence base and supported us to adopt user-generated solutions, such as the need to offer multiple vocational skills to help girls stay competitive.

Our economic strengthening technical partner also provided invaluable feedback and allowed our teams to rapidly expand their technical capacity in this new program area. We adopted two approaches to reduce the burden on teams to absorb too much information at once. The first was that the economic empowerment partner took on the role of synthesizing large volumes of information. Rather than share journal articles or large reviews, the partner developed presentations on key topics that lasted approximately an hour and allowed time for discussion. The second was to have an “on-demand” support mechanism as questions or challenges arose. Given the vast amount of information available, much of which might be irrelevant depending on the program model chosen, it did not feel efficient to try and proactively understand evidence around potential future challenges. It was important to provide design team members, especially those based in implementation geographies, the right information at the right time. In this way having a technical assistance approach which was demand-based allowed us to request detailed information on a specific challenge as it arose and focus our energies on understanding and applying the evidence. Nonetheless, we believe it would be better to have technical expertise fully embedded within the design team to support in the identification of gap areas as they arose, as a less experienced implementor may not always be able to recognize known gaps or program weaknesses.

Though we instituted many ‘layers’ of technical support to promote translation and application of the evidence, we know that despite our best intentions there were pieces of evidence that we missed the opportunity to apply because we didn’t have the bandwidth to digest such a substantive body of knowledge. The volume of information to absorb almost guarantees that organizations will miss out on the opportunity to apply certain evidence-based practices. We encourage researchers to consider how to make evidence more easily digestible to practitioners and for the whole community of practice to work together to identify what questions need to be asked to effectively operationalize evidence in a local context.