From “Opportunistic” to Intentional: How Finemind Built a Scale Strategy (and What We’d Love Feedback On)

Hi everyone,

I’m new to the EA Forum and grateful for the chance to share some recent work and invite critique.

I’m the co-founder and Executive Director of Finemind, a nonprofit delivering community-based mental health care in Uganda through task-sharing and integration into primary care. Since 2019, we’ve supported tens of thousands of counseling sessions, primarily for women and girls in underserved and post-conflict settings.

This post is about how our scale strategy came to be, why we felt compelled to build it, and where I’d genuinely value feedback from this community.

The moment that forced reflection

A while back, we spoke with a funder whose senior investment director offered a candid observation:

“Your growth strategy feels opportunistic and unintentional.”

It wasn’t said unkindly. And they weren’t wrong.

Like many early-stage orgs, Finemind grew through relationships, invitations, and funding constraints rather than through a clearly articulated expansion logic. That comment triggered a lot of internal reflection. If we believe mental health care is a moral priority, and if we hope to scale responsibly, then where and why we grow cannot be accidental.

That question became the seed for the strategy I’m sharing here.

What we set out to build

Our goal was not to create a perfect model, but a transparent, defensible way to decide where to scale that balanced:

• Ethical urgency (where unmet mental health need is greatest)
• System constraints (where systems are weakest or fragmented)
• Feasibility (where we can realistically deliver and sustain care)
• Strategic fit (where our presence is additive, not duplicative)

We wanted something that could withstand scrutiny, evolve with better data, and be legible to funders, government partners, and ourselves.

We’ve written up the full scale strategy, including assumptions, weighting choices, and a technical appendix. The document is available here for anyone who wants to dig into the details: https://docs.google.com/document/d/1SlOh9q6my2kvmbauHDypcTGpiq2wvASrxUIGriyao8M/edit?usp=sharing

Inspiration and methodological grounding

A major inspiration was Fortify Health, who generously allowed us to look at their State Selection Matrix. That framework gave us permission to be structured without pretending to be precise, and to treat prioritization as a moral and operational problem, not just a technical one.

From there, we built a two-stage framework:

• Stage 1: Need, combining mental health burden, system gaps, and access constraints
• Stage 2: Feasibility, combining stakeholder readiness and strategic fit

Throughout, we tried to be explicit about assumptions, use proxies where direct data doesn’t exist, and keep the model auditable rather than opaque.

A bit of BOTEC and cost-effectiveness

At scale (around 20,000 clients annually), our cost per person is projected at ~$80, benchmarked conservatively against comparable organizations and adjusted for Finemind’s operational model.

Using WELLBYs as a rough outcome measure, this implies up to ~32 WELLBYs per $1,000, which places Finemind among the more cost-effective mental health interventions we’re aware of.

One more speculative takeaway, offered with appropriate humility, is that a back-of-the-envelope makes our ~$100,000 transition funding look surprisingly well-leveraged over a ten-year horizon, on the order of 100+ WELLBYs per $1,000. We wouldn’t bet a grant on the exact number, but the direction of the result stuck with us.

A few interesting takeaways from the strategy

• There is no perfect overlap between need and feasibility. Regions with the highest burden are often the hardest to work in. The framework doesn’t resolve that tension; it makes it explicit.
• Access constraints compound need. In Stage 1, we treat limited access not as a delivery barrier but as a need amplifier.
• Urban bias is easy to introduce accidentally. We had to be intentional about not letting density and infrastructure crowd out underserved regions.
• “Where we can work” and “where we should work” are different questions. Keeping those analytically separate was clarifying.

What I’d love feedback on

Given the EA community’s depth of thinking on prioritization and cause allocation, I’d genuinely appreciate thoughts on:

• Whether the two-stage structure makes sense conceptually
• Whether the weighting logic feels defensible or arbitrary
• Whether there are better proxies we should consider for mental health burden or system readiness in LMIC contexts
• Whether the cost-effectiveness reasoning seems reasonable or optimistic
• Any blind spots you notice in how we’re thinking about scale

I’m not presenting this as “the right way,” but as a serious attempt to move from intuition to intention. If others can learn from it, improve it, or poke holes in it, that would be a success in itself.

Thanks for reading, and I appreciate the opportunity to learn from this community.

— Pavel