Kyle Poplin

Why the Aid Model is Perverse (Part 2): Tech startup mWater dreaming of the day health and wealth structures are under local control

Kyle Poplin: How many apps does mWater have? Which is most popular? Are others on the horizon?

Annie Feighery: We actually have five apps. Our flagship app, Surveyor, is the most used. We call it the Wild West app because anyone can use it however they want without a relationship to us. It would be just as useful to a health NGO mapping and tracking clinics as it is for water. Explorer is a far more streamlined option to Surveyor, designed specifically for NGOs and governments that are trying to align with the new Sustainable Development Goal 6 (“Ensure access to water and sanitation for all”) by mapping water and sanitation sites and monitoring them for safety and function. Explorer is also free, but in its development, we have the ability to build very cheap ($5,000) custom apps for organizations that want their own surveys presented as a white label app with their logo and private data visualization site. This is our first fee-based product in the platform.

One of our apps is the first water app on Facebook’s platform. This is a platform currently rolling out country by country that offers certain Internet apps for free, as long as the apps meet the standards of very low bandwidth pulls. We are trying to close the circle from collecting water safety data to communicating it back to the water end users. The app has been launched in three countries and will be in 13 more soon. We are still in the early iterative stages of learning how to adapt it to users so that it is truly useful. Eventually, we hope it works like Yelp for water. In places where more people have a mobile phone than access to a safe water source, they can use their mobiles to find safe water sources.

We built a surface water app and map for Earth Echo, an NGO that runs the World Water Monitoring Challenge. They use it for crowdsourced citizen water monitoring.

My favorite and, sadly, completely unused app is Broadstreet. We built it for frontline health workers to map diarrheal outbreaks. In five to 60 seconds a user can create a case and add as many details as they are able. It automatically maps the case to correlate with known, mapped water sources. It works on- and offline and on any model of phone with a browser.

(mWater’s monitoring tool in use, left.)

Last year, we adapted it for Ebola frontline workers to map cases with a single, gloved finger press on a screen. The Ministry of Health in Liberia contacted us and said they wanted to implement the app for water source and clinic mapping as well as household-level symptom mapping. This was very early in the global response (to Ebola), in September. To implement it, they needed mobile phones, data plans and small stipends for their workers. We reached out to U.S. groups like USAID, Gates and the Centers for Disease Control and Prevention to try to get Liberia’s Ministry of Health the resources they needed to use the app, but it was never implemented by any of the responding organizations. We would still love to get this app working for Ebola. The outbreak isn’t over and it could come back at any time now after the virus mutates further. I think two things hurt it: The agencies and NGOs in charge were focused on central professionals like doctors and epidemiologists conducting the data collection, rather than frontline health workers; and the “procure private software and deliver a custom product” cycle of emergency aid response did not have the capacity to integrate a built, off-the-shelf product. It’s built and it’s free if anyone wants to use it.

The mWater apps are all managed by the portal, which is a central data management platform. There, we can provide real-time data cleaning and very advanced data visualizations of survey results and sites mapped. In the portal, users can export their data as often as they need.


KP: Is there any dissonance in mWater being a nonprofit organization, yet being actively engaged in finding market-based solutions?

Annie Feighery (right): Working in a revenue model ensures our product is always needed. The aid model is actually a perverse economic model because the end product being useful enough to drive user metrics has no connection to whether the donor likes it and it is continually funded. There are broken feedback loops. For us, if people stop using the apps and we stop receiving revenue for our services customizing, expanding and training the software, we go out of business. NGOs almost never go out of business.

Part of our dream is the world post-aid. Poverty can’t really be eradicated unless the structures for health and wealth currently being built by the aid industry can be transitioned to local control by local governments and markets. We think building solutions that can exist without aid depend on two possible support structures for sustainability: Either they can be supported by local taxes and fees (like water currently is in the U.S.) or they fit into an existing or emerging market. The Surveyor and Explorer platforms meet the earlier case: Fledgling utilities can map their systems and monitor them without adding any load to their budgets beyond a $70 smartphone (and that price is dropping) for the field managers. The infrastructure mapping actually cuts their costs and maximizes their efficiencies because they become aware of non-revenue water from leaks and theft. They are able to provide more consistent services in taps and kiosks and studies show more consistent service results in more fee payment.


KP: Is it surprising that no government entities filled the void that mWater currently fills? Why didn’t they?

AF: I think, given time, UN and government aid could have filled the gap. We are starting to see mobile monitoring tools being built and advertised by some UN branches, for example. As a startup, we move at lightning speed compared to bureaucracies that, often by design, turn like the Titanic. Bureaucratic structures are good at protecting from problems that could come with rapid change, like revolution, but they don’t make good software. The knowledge economy may very well be a new evolution in management that adds to more traditional entities.


KP: Why is it important to have an open-source business model?

AF: The most important part of our model is that the software is free and open access. The alternative strategy that some of our competitors use is a fee structure that can be written into grants when water projects are conducted. That means, after the project ends and the NGO hands it over to the local government, if they want to keep monitoring it, they have a long-term debt to fill. Keeping it free for the end user follows the software model of the most successful platforms, like Google Docs and Facebook. The people who benefit most from this existing – in our case, very large NGOs with big mobile monitoring and evaluation needs – pay for it. And the value of their product only goes up with the more end users we can attract.


KP: How will you ultimately determine if mWater is successful?

AF: (Number of) users is my most important metric on a daily basis. One of our biggest successes is the open collaboration of two of the world’s largest water NGOs, WaterAid and, in investing in mWater. NGOs used to compete so constantly, that collaboration was not possible. We are continuing this goal to create collaborations between NGOs and local governments.

Over the course of the next five years, we need to have generated advances in the field of water and sanitation that result in a significant advance in diarrheal disease. Or we need to go out of business.


KP: At what specific point in your startup did you realize that mWater was filling a need and that your concept for the organization was valid?

AF: As a startup, the bootstrapped stage was the hardest. We had so many points where we didn’t know if we’d make it. Then, all at once we hit a tipping point and signed contracts that brought us into the black. That was early spring 2014. I remember thinking it felt like learning to ride a bike – that moment where you don’t realize your mom isn’t holding onto you anymore and you’re balancing on your own. My chief technology officer officially quit his day job that May. We had a paid staff of four. By summer 2014 we had people calling us for our services and we were internally yelling, “I’m doing it, I’m really doing it!” We are still evolving and iterating. We can’t ever really lean back in our seats or we won’t be a startup anymore.


KP: What are the major challenges facing mWater?

AF: We face challenges with expanding budgets and attracting partners/investors and challenges scaling without losing our pace or vision. But our biggest challenge is paper. Our biggest competitor is not another mobile app, but rather paper-based monitoring and evaluation. We like to say paper is where data goes to die. It closes the door on the data collected ever becoming meaningful and ever getting shared. We have to keep doing a better job at getting the word out that digital monitoring can help every single aspect of the aid industry do its job better. There are so very many malthusians out there who say things like, “Technology is not a solution.” Imagine saying that to a librarian organizing card catalogues in 1995. Technology is the solution.


Kyle Poplin is the editor of NextBillion Health Care.

Agriculture, Health Care, Technology