Let’s be honest about something that doesn’t get said often enough in development circles: reaching marginalised communities is hard. Not hard in the way that filling out a report is hard, or hard in the way that coordinating a Zoom call across time zones is hard. Hard in the way that requires you to completely rethink your assumptions about how change happens — and how long it actually takes.
At Tech4G, we have spent two years operating across 23 counties in Kenya. We have delivered 45 programmes. We have sat in community halls in Mandera, navigated unpaved roads to reach pastoralist households in Wajir, and tried to run a digital skills session in a room where the internet cuts out every twenty minutes. And through all of it, we have learned things that no programme design document ever warned us about.
This is our honest account of what it actually takes.
First, You Have to Earn the Right to Be There
There is a temptation, in this sector, to arrive in a community with a curriculum, a timetable, and a set of predetermined outcomes — and to expect people to show up because the programme is good for them. It rarely works that way.
In Kenya’s marginalised counties, communities have seen organisations come and go. Promises have been made and not kept. Projects have launched with fanfare and quietly dissolved. The result is a completely rational wariness — a community that does not immediately trust a new face with a new offer, no matter how well-intentioned.
Trust, in our experience, is not a soft metric. It is the actual infrastructure of effective programme delivery. Without it, attendance is poor, dropout is high, and the people who need the training most — women, persons with disabilities, older community members — are often the least likely to come forward.
Research on last-mile programme delivery in northern Kenya confirms this. A study on digital financial inclusion in the region found that community members’ willingness to participate was deeply tied to whether they recognised and trusted the people delivering the programme — and that when that trust was absent, even technically well-designed interventions failed to generate meaningful, lasting change (Karlan et al., 2016, as cited in Tanle & Alhassan, 2019).
Building trust means spending time before the training starts. It means meeting with local elders, community leaders, women’s groups, and local government officials — not to tick a consultation box, but because their endorsement changes everything. When a respected local voice says “this is legitimate, these people are here to help,” it opens doors that no marketing campaign ever could.
The Language Problem Is Real — and Deeper Than You Think
Kenya has 68 officially recognised languages. When you work in Mandera, you are working in a county where Somali is the everyday language of most residents, where Kiswahili is understood but not always comfortable, and where English — the language of most digital content — can feel like a third or fourth tongue.
This is not a minor operational detail. It is a fundamental design challenge.
We have learned that delivering training in English or even Kiswahili alone excludes the people furthest from the digital economy — the very people the programme is designed to reach. True inclusivity requires content that speaks people’s language, literally. It means working with Community Digital Champions who are not just digitally literate, but who are rooted in the community and can navigate between languages and cultural contexts with fluency and sensitivity.
It also means accepting that translation is not just a linguistic exercise. Concepts like “cloud storage” or “digital entrepreneurship” do not have ready equivalents in many Kenyan languages. Communicating them effectively means finding analogies that resonate with people’s actual lives — farming, livestock trading, market days — and building understanding from there.
When we get this right, something shifts in the room. Confusion becomes curiosity. Disengagement becomes participation. And the person at the back who came because a neighbour dragged them along starts asking questions.
Connectivity Is Not a Given. Plan Around Its Absence.
Every programme that assumes reliable internet access in underserved Kenya is planning to fail part of the time.
According to the Communications Authority of Kenya and the Kenya National Bureau of Statistics, mobile phone ownership in counties like West Pokot sits at just 29%, Turkana at 29.4%, and Marsabit at 34.8% (CA & KNBS, 2025). These are not just device ownership gaps — they reflect broader patterns of infrastructure scarcity that make consistent online training a logistical challenge.
We provide participants with data bundles specifically to reduce this barrier. But even with data in hand, a session can still be derailed by a weak signal, a power outage, or a device that overheats. The honest reality is that hybrid and offline-first approaches are not contingency plans in rural Kenya — they are the baseline.
This shapes how we design everything: how we sequence content, how we build in repetition, how we make sure that a missed class does not become a missed opportunity. Participants who miss a session can rejoin the same module in a subsequent cohort intake without having to repeat the entire programme. It sounds like a small thing. For a mother who had a sick child the day class was scheduled, it is the difference between dropping out and completing.
You Cannot Separate Digital Skills From Economic Reality
One of the most important lessons we have learned is this: a person who is worried about where their next meal is coming from is not going to fully engage with a lesson on online business registration.
In Busia and Mandera — counties selected for our intensive work under the Strengthening Digital Communities project precisely because of their high food poverty rates (49% and 65.5% respectively) — the daily pressures of economic survival are not background context. They are front and centre in every participant’s mind.
This means that effective programme delivery requires acknowledging the whole person. We build in financial literacy and entrepreneurship components not as add-ons, but as integrated parts of the learning journey. We connect digital skills training to practical, immediate applications: how to use mobile money, how to access government services online, how to reach customers beyond the local market. The digital is never abstract — it is always in service of something tangible and liveable.
It also means being realistic about timing. Training sessions scheduled during harvest periods, or on market days, will be poorly attended regardless of how good the content is. Working with communities to understand their calendars — not imposing ours on them — is not optional. It is how you show respect, and respect is how you stay.
Local Partners Are Not Just Logistics. They Are the Programme.
Tech4G does not operate alone. We work within a rich ecosystem of local organisations — groups like the Akukuranut Development Trust in Busia, GENCAD in Mandera, and a range of county-level bodies who understand their communities in ways that no headquarters-based organisation ever could.
These partners are not subcontractors. They are co-designers. They flag when a session time clashes with a community event. They know which village elder’s blessing matters. They understand the particular dynamics of gender in their community — which women can attend evening sessions and which cannot, which husbands need to be spoken to first, which spaces are culturally safe for mixed-gender training.
Research consistently shows that community development projects achieve significantly more impact when local stakeholders are genuinely involved in decision-making — not consulted after the fact, but co-designing from the start (Hendricks, 2018, as cited in Local Stakeholders’ Involvement in Community Development, 2022). A framework study on digital cooperation in Kenya found that co-design was “a game-changer in enabling wider participation” — the difference between a programme that a community tolerates and one that a community owns (Digital Future Society, 2020).
We have seen this play out in practice. The programmes that have generated the strongest outcomes are invariably the ones where local partners were most deeply embedded in delivery — where community members were not just recipients, but active participants in shaping how the training worked.
Safeguarding Is Not a Form. It Is a Commitment.
Working with marginalised communities — particularly women, young people, and persons with disabilities — carries a duty of care that goes beyond programme delivery. The communities we serve are, by definition, communities that have historically been let down by institutions. They are often more vulnerable to exploitation, misinformation, and harm.
At Tech4G, we take safeguarding seriously as both a moral obligation and an operational priority. Every participant should feel that this space is safe — that questions will be answered honestly, that personal information will be handled with care, that they will not be pressured or manipulated. Our Code of Conduct is not a document that lives in a filing cabinet. It shapes how every trainer, every partner, and every team member conducts themselves in the field.
This matters for outcomes, too. People do not learn well when they do not feel safe. Psychological safety — the sense that it is okay to not know, to ask a question, to make a mistake — is as foundational to digital literacy training as a stable internet connection.
What All of This Means for Partners and Funders
If you are looking for an implementing partner that will give you a clean story with no complications, Tech4G is probably not the right fit.
But if you are looking for an organisation that has actually done this work — that has navigated the gap between policy ambition and ground-level reality, that has built real relationships with real communities in Kenya’s hardest-to-reach counties, that has failed at things, learned from the failure, and built better programmes as a result — then we would very much like to talk.
The work of reaching marginalised communities is not glamorous. It involves patience, cultural humility, operational flexibility, and a willingness to be wrong about your assumptions and change course. It involves building relationships that take months before they produce results.
But it is also the only kind of work that actually changes things. And after 45 programmes, 23 counties, and thousands of community members reached, we believe — more than ever — that it is worth doing.
Want to Go Further With Us?
Tech4G is actively seeking funders and implementing partners who share our commitment to genuine, community-rooted digital inclusion. If you believe that the last mile deserves the same attention as the first, let’s build something together.
📧 [email protected]
🌐 technology4growth.org
🤝 Partner with us
References
- Communications Authority of Kenya & Kenya National Bureau of Statistics. (2025). Analytical report on ICT based on the 2023/24 Kenya Housing Survey. Communications Authority of Kenya. https://www.ca.go.ke/urban-rural-digital-divide-hinders-ict-uptake-joint-ca-and-knbs-survey-shows
- Digital Future Society. (2020). Bridging digital divides: A framework for digital cooperation. https://digitalfuturesociety.com/app/uploads/2020/04/THINK_TANK_I4_BRIDGING_DIGITAL_DIVIDE_EN.pdf
- Hendricks, L. (2018). Community participation in development projects: Enabling factors and outcomes. Journal of Community Development Research, 11(2), 45–62.
- Karlan, D., Kendall, J., Mann, R., Pande, R., Suri, T., & Zinman, J. (2016). Research and impacts of digital financial services. NBER Working Paper No. 22633. National Bureau of Economic Research.
- Tanle, A., & Alhassan, A. (2019). Barriers to “last mile” financial inclusion: Cases from northern Kenya. Development in Practice, 29(8), 1012–1026. https://doi.org/10.1080/09614524.2019.1654432
- Waldenu Dissertations. (2022). Local stakeholders’ involvement in community development projects in Kenya. Walden University. https://scholarworks.waldenu.edu
