We're hiring across GiveWell's Operations team—with openings in Grants Administration, People Operations, Talent Acquisition, Technology, AI Enablement, and Executive Support. The deadline to apply is September 7.
We’re hosting an Ask Me Anything session with our Chief Operations Officer, Sai Jahann, to answer questions about what it’s like to work at GiveWell, our Operations functions, and this hiring round. We're accepting questions through Friday, August 17.
A year ago, family planning was a new area for GiveWell—since then, we've made around $20 million in grants to family planning programs across commodity procurement, supply chain support, and service delivery.
In our latest podcast episode, GiveWell co-founder and CEO Elie Hassenfeld speaks with Senior Program Officer Dilhan Perera about our grants to the United Nations Population Fund (UNFPA) Supplies Partnership, MSI Reproductive Choices, and Lafiya, how the portfolio fits together, and what we're hoping to learn about bottlenecks in family planning access.
Diarrhea kills more than 400,000 children under five every year. GiveWell recently made a grant to support WASHmobile in Eastern DRC and South Sudan—a water, sanitation, and hygiene program that aims to reach households at heightened risk of diarrheal disease.
In a recent podcast episode, GiveWell co-founder and CEO Elie Hassenfeld speaks with Senior Program Officer Erin Crossett about how the program works, why these locations are so challenging to work in, and what we hope to learn from this grant.
Hi Tony, This is Mark again. Thanks for the nudge on this monitoring point!
We agree that real-time monitoring of critical activities is likely to be a cost-effective investment. For example, New Incentives (a program that provides conditional cash transfers for immunizations working with Northern Nigeria health clinics) tracks and shares with us data on cold chain failures, stockouts, rollout delays, security incidents, fraud, double treatment of children, and retention rates (see more information in this intervention report). We think this type of monitoring is a good investment for New Incentives and want others to do something similar, because it:
We're still trying to understand the trade-offs between grantees doing this internally vs. externally, and we expect to learn as we go. It makes sense to me that it is usually most efficient and effective for grantees to do the core of this monitoring internally since we want them to be able to address these issues in real-time (point 1) rather than waiting for a report from an external firm. In cases where GiveWell is relying on the accuracy of this data for our decision-making (points 2 and 3), I think it makes sense to double-check data accuracy through external data quality audits and/or interviews with local stakeholders.
The same goes with standardizing. We think it's good to have a minimum standard, but we guess there's also variation in what's feasible or makes sense across grantees and expect to figure that out with them. We’ve standardized monitoring and reporting within a few program areas, such as across all our recent water chlorination grants, in collaboration with our grantees (e.g., agreeing on consistent standards on what grantees are expected to track, at what frequency, and how findings should be acted on). However, we haven’t developed monitoring standards that apply across all program areas. I think there could be value in having broader principles, but, so far, we’ve prioritized addressing concrete issues we’ve identified rather than developing these broader principles. We are planning to publish more on the details of these program area-specific monitoring and reporting standards in the near-future.
Hi Tony—This is Mark Walsh, a GiveWell researcher on the team responsible for pressure-testing GiveWell's research processes and conclusions. Thank you for this thoughtful write-up!
At a high level, I agree with the central thesis: we've underinvested in monitoring and evaluation, relative to other components of our analysis. Like you mentioned in your post, we've been working to fix that, and I wanted to share a little bit more on what we've done so far, what we're planning, and some other related gaps in our work.
We think this kind of external engagement with our research is valuable and makes our work better. We'd welcome feedback on the steps we've taken so far and where we should consider doing more.
What we’ve done
Over the past year, we've been working on what we call "M&E red teaming"—a systematic review of the monitoring and evaluation practices of our largest grantees, motivated by the same concerns Tony raises. We're planning to publish our findings soon but wanted to give a brief overview.
From June to December 2025, we dedicated teams of 3-4 research staff to work full-time for roughly 3 weeks each on six program areas: our four top charities plus water chlorination and malnutrition treatment. For each program, we evaluated many of the dimensions Tony highlights in his monitoring checklist: the independence of data collectors from program staff, the neutrality of the sampling frame, the objectivity and precision of measurement approach, data quality checks and backchecks, timeliness of data, triangulation with independent sources, and whether the program is taking timely action to address any issues raised by monitoring.
Since completing the red teaming, we've been doing the following to make improvements:
As Tony suggests, we think the right way to evaluate these investments is based on the value of the information we’ll gain and the impact we think it will have on our future grantmaking decisions. That means that we are more likely to fund expensive M&E in large grantmaking areas for GiveWell (or areas with a lot of room for more funding) and where we have more uncertainty so expect the M&E to affect our grantmaking a lot. For example, we’ve funded expensive independent coverage surveys of insecticide-treated nets and vaccinations programs because we direct a lot of funding to these programs. On the other hand, we are trying to take a rigorous but lighter-touch approach in cases where there is less funding at-stake, or where key uncertainties (e.g., whether an organization can establish partnerships or hire effectively) can be resolved more cheaply before investing in expensive M&E.
What’s next
Over the next several months, we're planning to finalize and roll out our coverage survey standards with grantees; analyze the results from the independent surveys, enhanced M&E, and population/costs work we've commissioned; and dig further on more of these areas (e.g., investigating better ways to monitor the impact of our grantees on disease morbidity and mortality).
I expect we'll learn a lot as we go about what's feasible and what information is actually most valuable. I'm sure our approach will change as we learn more about both the cost and potential impact of this information. That being said, I think we are on the right track.
A broader gap
I also want to flag that we think the issues Tony raises are part of a broader gap. It's not just that we need better quantitative monitoring—we also need to invest more in understanding what's actually happening on the ground with the programs we fund.
We've been trying to gather more "local insights" on our work. This involves site visits, qualitative research, conversations with local experts, and other ways of testing our desk-based assumptions against what’s happening on the ground. One example is funding the Busara Center for Behavioral Economics to observe vitamin A supplementation delivery in Nigeria and interview households, front-line staff, and government officials about the program.
We're still figuring out which approaches are most useful but think shifting more of our research effort "beyond the spreadsheet" in the ways Tony is describing is directionally right and something we’re making progress on. As I said at the start, we welcome feedback on our work so far—and on our future progress as it occurs.
Thank you for flagging this! And a special thank you to Rory for fixing it so quickly. :)