Zipline Drone Delivery Offers Lessons for Health Care Innovation in Africa

(Photo courtesy of Zipline)

According to the United States Mission to the African Union, the US alone has invested more than $100 billion in African health over the last 20 years. Despite the billions that international donors have poured into African health systems, expert reviews, such as the Lancet Global Health Commission, have found the performance of these systems to be unsatisfactory. Those authors estimate that poor-quality care is responsible for 60 percent of deaths caused by treatable illnesses on the continent. Reviews of African health systems call for fundamental change in how we operate and a paradigm shift in health innovations because more of the same will not be enough.

Unfortunately, conversations with innovators from the private and NGO sectors indicate that international donors and their funding processes, attitudes, and relations with local governments can inadvertently act as barriers to developing effective health innovations. The innovation journey of Zipline, a technology company that uses drones to deliver medical supplies and where two of us (Sofer and Burton) are senior executives, vividly illustrates many of the challenges that innovators face. We see several strategies that could help overcome these barriers to executing and scaling effective health solutions.

An Unhealthy Context for Health Innovations

Most African countries have struggled to lower their dependencies on foreign investments. This reliance on international donors comes with significant opportunity costs in terms of limited local ownership and leadership of health agendas. Too little time is spent on local program design and genuinely innovative approaches addressing local health issues. The philanthropic sector has also been criticized for using its unprecedented funds and political influence to dominate local health agendas and leaving little room for local innovators. The Gates Foundation, for example, has been accused of focusing too much on vertical health issues such as polio or malaria eradication that fit the foundation’s preferences for technical interventions and concrete measures. Many global health initiatives operate similarly, funding expensive treatments for specific diseases in a highly siloed fashion rather than promoting local innovations that are also integrated into health systems to improve performance and accessibility.

Innovators may also have a hard time accessing resources. Not only are financial and other resources scarce, but they are also not put to their best use. The WHO identified significant inefficiencies in translating financial resources into health outcomes: “For every $100 that goes into state coffers in Africa, on average $16 is allocated to health, only $10 is in effect spent, and less than $4 goes to the right health services.” International donors are also prone to mismanaging resources. The Bureau of Investigative Journalism and Devex published an investigation into a $9.5 billion contract issued by USAID to improve health supply chains in lower-income countries. The report triggered a congressional inquiry over concerns about “evidence of fraud, waste, and abuse, and that USAID and the Washington, D.C. consultant managing this project jointly manipulated performance indicators.”

Needed: Scaling Effective Innovations in Health

Many hurdles stand in the way of formulating an innovation case, sourcing ideas and resources, and supporting their exploration and development to improve health services. Two of us (Seelos and Mair) previously published an article, “When Innovation Goes Wrong,” outlining an innovation pathologies framework that can be helpful in understanding those barriers. Six pathologies systematically hinder innovation efforts: never get started, stop too early, stop too late, too many bad ideas, too little scaling, and too much innovation. To illustrate how these pathologies play out in the real world, we share insights from the concrete experiences, observations, and learnings of the Zipline team developing and scaling health innovations in Africa.

Never Get Started

Starting an innovation effort is necessary when public health ambitions cannot be achieved merely by investing in the expansion and incremental improvement of existing health systems. Failing to kickstart innovations can thus close the door to important strategic opportunities and sustain an underperforming status quo. Saying no to ideas too often and preventing innovations from starting may also frustrate innovators and create a risk-averse context that, over time, loses its ability to start and manage effective innovations even when their strategic necessity becomes clear.

One reason for this pathology may be what we have called a lack of “constructive dissatisfaction” with one’s performance. This attitude drives an ambition for constant improvements but also provides legitimacy for innovations where improvements may not suffice. When donors and ministries do not openly acknowledge the poor design, management, and performance of many existing health systems, they undersupply strategic investments in innovation efforts. In the context of Africa’s health systems, the institutions that influence the political and operational aspects of health and development may have little incentive to change the status quo. Large contractors and international NGOs benefit financially from sustaining their role as a workaround to underperforming systems. Supporting innovation and change could mean working themselves out of a job. For decades, a well-established multi-billion-dollar “development-industrial complex” has limited the opportunities for smaller entrepreneurial actors or actors from other sectors to engage, causing a disincentive to innovate. The World Health Organization has expressed concerns that health ministries likewise neglect and fail to utilize the innovation potential of the private and voluntary sectors. However, the reality is that the existing incentives fail to encourage a change in attitude, and therefore, important innovations grounded in novel ideas and unique competencies “never get started.”

Starting and sustaining innovations requires resources. For a social entrepreneur looking to break into the health space, the assessment of the funding landscape can look bleak. Many donors only provide small seed innovation grants, which might be insufficient to kickstart serious innovation efforts. When Zipline started exploring ideas for operating in Africa in 2016, it found a development sector dominated by a small set of actors who claim the lion’s share of the available funding, with the amounts set aside for small, upstart organizations being extremely limited in comparison. Merely making oneself eligible for these funds often requires partnering with one of the larger organizations. Months (or even years) of work can be sunk into pursuing a single funding opportunity, only to yield a tiny award, at best. Funding is also episodic, unreliable, and very difficult to forecast. This unpredictability creates high uncertainty about whether even a promising innovation can be sustained and fully developed. As a result, many innovators conclude that trying to serve the health sector is just not worth the effort.

Stop Too Early

Sometimes, even promising innovations are abandoned. This deprives innovators and their organizations of any hope for eventual success and the possibility to accumulate important learnings that inform future innovation efforts. As long as health problems persist, innovation only ever fails when we give up or stop too early.

Innovations usually evolve non-linearly as an iteration between hope, good news, and a frustrating sense of going nowhere. Managing this rollercoaster requires decision-makers to understand the strategic rationale and the innovation case that legitimizes a sustained effort. Unfortunately, innovation has become a buzzword that excites donors but mostly creates expectations for quick, small-scale solutions to well-defined problems. Today, innovation is an anticipated component of grant applications and responses to RFPs. Easy funding for small pilot projects gets everyone excited. Rarely do donors invest in innovations from a strategic perspective aligned with long-term impact and effective scale. The experience of Zipline illustrates how short-sighted innovation perspectives can hinder the realization of promising innovation efforts and force innovators to “stop too early.”

Zipline initially received early-stage funding from a few donors to launch operations in Rwanda and Ghana. After a few years of successful pilots, these donors declined to provide funding to support Zipline’s further expansion and integration. Some argued the venture capital Zipline had raised obviated the need for donors to fund the system’s use in development contexts. Other donors believed their role was “market shaping” and this assessment made them abandon a high-impact technology platform in favor of trialing new upstarts to compete with it. However, the funds Zipline raised were required to build its drone technology, not to pay itself for its service delivery in Africa. Investments in competing innovations are only effective if the best innovations are subsequently selected and adequately supported. Fortunately, as detailed later in this article, Zipline was able to bridge this chasm between early adoption and scale. Still, most innovators cannot, causing efforts to shut down before demonstrating impact data and potential for scale.

Stop Too Late

This pathology might be the costliest. Stopping too late means continuing to invest resources in innovation processes with little potential to succeed or generate meaningful learning. Decision makers may sometimes be unaware of this pathology. For example, in hierarchical contexts such as the African health system, innovators may be reluctant to provide honest feedback to influential donors or sponsors of projects. Face-saving attitudes by elites may create a “shoot-the-messenger” approach that further stifles feedback about a lack of progress.

Another important mechanism behind the continuation of ineffective innovations is the sunk-cost fallacy. Sometimes, innovators spend their budget on one large pilot study to “force” success rather than maximizing learning by pursuing several smaller studies. This attitude may be fueled by the recent big-bet and system-change attitudes adopted by the philanthropic and development sectors. Not facing reality and failing to pull the trigger prevents innovators from learning effectively and pursuing an evolutionary approach to innovation that creates and enables efficient identification and selection of superior solutions.

This sunk-cost fallacy is robustly at play in the development-industrial complex today and employs a tremendous amount of financial, technical, and human resources, which have a high opportunity cost—they could be deployed to support more effective solutions in the health ecosystem. For example, Zipline can deliver across a large radius (38,000 square kilometers), which could include thousands of delivery points. This range dramatically reduces the need for cold-chain systems, making it a very cost-effective intervention for vaccines, as a recent study under peer review shows. Although partners celebrate significant improvements in vaccination rates in the Zipline-served regions of the country, donors have blocked government requests for Zipline to increase its coverage and include facilities already equipped with cold-chain systems. Everyone wants to justify the investment made in the past, even if it means continuing to invest in a sub-par solution today. The unwillingness to ignore this prior investment and commit to the most effective innovation is an example of the sunk-cost fallacy that can prevent innovations from stopping in time and being replaced with superior innovations.

Too Many Bad Ideas

Innovations have a better chance of success when ideas are informed by relevant knowledge about local context. Bad ideas ignore reality and are often driven by naivete or hubris and inflated ambitions that lack the proper competencies to improve and implement ideas. The quality of innovative ideas usually increases with the extent to which donors become embedded in the local context and develop a deep understanding of the characteristics and dynamics of social problems. Unfortunately, most Western donors still operate from an attitude of “we know better” and arrive with a preconceived understanding of local issues from Western mental templates and experiences from other contexts. Local contractors quickly learn that being seen as a reliable partner and accessing funds requires them to comply with donor demands and build capacity to implement donor-designed projects that may not fit local realities. These innovations, grounded in bad ideas, have huge opportunity costs because they capture scarce resources that could be supporting much better-informed innovation ideas.

Innovation ideas are also often over-engineered and overly prescriptive. Frequently, innovators are faced with calls for proposals or investment opportunities where donors prescribe the means through which something should be achieved instead of defining the desired results or outcomes. Some scale-up funds for health innovations specify a few thematic areas, such as “AI” or “drones,” to categorize their investments instead of identifying desired outcomes. But this approach might not be such a good idea. Just because Zipline has demonstrated the significant health impact of a successful drone innovation does not mean any drone innovation has the potential for scale and positive impact.

Instead of taking an approach that leaves many avenues open for experimentation, donors land on mission statements and plans that have been so overly crowdsourced, prespecified, and reviewed for risk and compliance that the original intention to innovate is thoroughly watered down. Too often, innovation ideas are evaluated by whether they are “good” or “bad” based on untransparent and ad-hoc criteria. This approach rubs off on innovators, leading to narrow, inoffensive, and unambitious ideas that fail to move the needle on desired health objectives. Rarely do supporters consider innovation a long-term process where initial ideas play only a small part and will need to be revised or replaced by better ideas. Prescriptive and prespecified innovation practices fail to appreciate that adaptation and change are how innovations can be “built to last.” From this narrow view, seemingly good innovation ideas may quickly become bad ideas that donors dismiss as “obviously” not leading to positive impact. This is how global health institutions talk themselves into maintaining the status quo.

Too Little Scaling

Scaling turns investments in innovation into impact. Scaling health innovations—efforts to expand and improve innovations to serve more people better—can dramatically increase the impact return on the investments in health innovations. However, the positive impact of health innovations often remains elusive because we struggle with the delivery aspect of scaling them, as remarked by the Committee on the Quality of Health Care in America: “The science and technologies involved in health care—the knowledge, skills, care interventions, devices, and drugs—have advanced more rapidly than our ability to deliver them safely, effectively, and efficiently.”

Scaling, unfortunately, is often an afterthought of investments in innovation. Without scalability and ambitions for scale in mind from the beginning, innovators risk ending up with something that does not scale or can only be scaled in specific circumstances. Donors may fail to consider the resource requirements for truly supporting an innovation to reach effective scale. Scaling is often required to reduce operating health innovation costs due to economies of scale and learning curves. This relationship creates a chicken and egg dilemma—we need scaling to arrive at an efficient cost of operations, while donors often shy away from significant investments in scaling.

Another driver of the “too little scaling” pathology is a tendency to add layers of innovations around an underperforming health system. For important innovations to flourish, they often need to be integrated more deeply into a health system rather than scaling in parallel with an existing system. Unfortunately, as donors have rushed to pilot health innovations, they have not put enough focus on the role of integration in making sure an innovation “sticks.” Investing in a fantastic health innovation creates limited impact if donors continue to allocate 95 percent of their budgets to operate health systems in the same old way.

It took Zipline some time to recognize that it also became trapped in this pathology. Zipline believed that, because it had raised funds successfully and had proven the effectiveness of its technology and operations, donors would naturally want to continue to support its work. Instead, it found itself operating in bifurcated African public health systems in which it served government-led use cases (blood and essential medicines) while donor-led supply chains (HIV, TB, and malaria, for example) operated as if Zipline was not present at all. By the time Zipline realized this, it was already getting stuck in the “valley of death,” the phase between the initial launch of an innovation and achieving significant adoption. The organization’s leadership launched conversations with existing donor partners (foundations and multilaterals), new potential donor partners operating in similar geographies as Zipline, European bilaterals, institutions focused on large infrastructure and development projects, and social impact investors. Not a single one felt it was their mandate to scale innovations. It was startling: Zipline was falling through the cracks. Only through sustained persistence, undeterred leadership among African government partners, and smaller private funders like the UPS Foundation and the Elton John AIDS Foundation—donors that truly sought the greatest ROI on their grant funding—could Zipline buy itself time to educate donors about the need to invest in scaling innovation and begin to turn the tide.

Too Much Innovation

For a long time, the philanthropic and development sectors have promoted an ideology of innovation is good and more innovation is better. Case studies and articles often showcase bold innovators but are low on credible evidence of their innovations’ potential or actual impact over time. At the same time, those who make the effort to manage the integration, delivery, and scaling of innovations into health systems are talked about much less frequently. Their work is what truly generates an impact on health, but it is much less sexy and glamorous than innovation.

A focus on innovation is particularly ineffective when this focus is not coupled with a systematic effort to maximize learning from both successes and failures. One hallmark of Silicon Valley is not being afraid of failure; it’s okay to fail as long as you learn from those mistakes. But in the development sector, there seem to be very few attempts made to learn from failure, and when questions are raised (such as with the US congressional probe mentioned above), people rush to cover their tracks rather than extract any real learnings.

Like many other innovators, Zipline also struggled with donors’ exuberance to promote more and more innovation to the detriment of supporting innovators’ efforts at getting good at something and investing in deep integration over longer timescales. It also suffered from a lack of attention by donors to true performance over time as opposed to shiny novelty and promises of quick wins. In their experience, donors rarely invest in analysis or post-mortem on promising innovations that never succeeded. There is little motivation to learn from past mistakes, but every year a new innovation fund or accelerator is launched.

Successful Innovation Against the Odds

As our account of innovation pathologies indicates, much of what makes African health systems an unnecessarily challenging context for innovations involves the significant hurdles that donor attitudes and processes and their special relations with local implementers and governments unintentionally create. Like many other organizations, Zipline found the context of African health systems almost prohibitive for incentivizing and supporting effective innovation efforts. Most organizations in this position would have given up by this point.

There are two reasons why Zipline did not fold. First, the organization nurtured an attitude of “stubborn fanaticism” and unwillingness to fail. This attitude is deeply ingrained in Zipline’s DNA. The organization genuinely believes there is a better way to create health care access. Second, Zipline’s business is not entirely or solely dependent on succeeding in Africa or the development sector. This does not mean it will operate forever without eventually needing to reach profitability in Africa (to date it has focused simply on financial sustainability). However, it does mean that Zipline has been able to lean on resources from investors who supported its “heart and soul” mission. And while organizations that only receive funding from donors might have been forced to shut down, Zipline could stay afloat and continue to drive forward.

Zipline also realized at an early stage that it must invest in learning and monitoring capabilities. Initially, Zipline tried to push for support of its innovation and scaling efforts by trying harder, selling harder, and generally being more aggressive—not a very successful approach. Then it tried to reason with donors to help them understand how pathologies like the valley of death dilemma hinder innovations. In retrospect, this may have come across as too cynical, disappointed, and negative to be effective. Eventually, Zipline adopted a different approach: It began to collect credible data systematically and to generate quality research about its health impact. This included identifying and working with supportive champions from local government. In that manner, Zipline could concentrate on implementing services below the radar screen of donors at a scale that was effective for generating convincing data. When researchers from the University of Pennsylvania’s Wharton School published a study demonstrating that Zipline’s operations helped decrease maternal mortality due to severe bleeding after childbirth by 51 percent, donors started to listen. Convincingly demonstrating the impact of Zipline’s innovations mobilized fresh funds for scaling.

Zipline also needed to overcome concerns about being a potentially unreliable American private sector company. Despite the odds, Zipline made progress through several key decisions. First, the company operates in every country with 100 percent local staff. Additionally, Zipline invested in efforts that created value beyond health, such as supporting engagements and visits by local STEM students and researchers to provide access to frontier technology like drones and robotics. The company also trained local staff in AI and physical technologies to support high-quality jobs. Furthermore, Zipline worked on deep integrations of its innovation into local health systems, demonstrating potential health impacts and significant cost savings. Lastly, the company invested in identifying and nurturing relations with influential local leaders who were genuinely interested in transformative long-term change and creating a positive legacy in health. Over time, local governments and communities became the best advocates for scaling Zipline; blocking or routing government-led transformation, or (worse yet) undermining government-led investments in their chosen technology platform, became a hard position for donors to sustain.

A Healthy Ecosystem for Health Innovation

Zipline’s experiences in developing and scaling important health innovations in Africa demonstrate the potential of innovations for improving health systems. However, they also highlight how the aid sector’s business-as-usual approach stifles effective innovations. We propose several key changes from donors and local governments to create a healthy innovation ecosystem that effectively fosters innovation and scaling to enhance access, economics, and quality of health systems:

• Prioritize concrete health outcomes rather than specifying means to achieve them. Exploring effective means is the work of innovators!
• This outcome-focused, hands-off approach should be balanced with monitoring for tendencies to reassert control through bureaucratic formalities and overengineering solutions. Left unchecked, these tendencies could stifle innovation efforts even at advanced stages.
• Put the right incentives in place to ensure a transition from the status quo and be on the lookout for the beneficiaries of the current system who will push to prevent change.
• Help guide innovations through the valley of death, including providing funding, shaping incentives, and facilitating integration into existing health systems.
• Actively promote private sector partners to do business with development actors and governments. This includes shifting to more commercial models versus cost-plus contracts or grants, where every single expense must be accounted for. Additionally, it involves supporting models where governments can more easily contract directly with the private sector.
• Get hands-on experience; if you’ve only been theorizing at your desk, you won’t be able to identify and support effective innovations!

Change is possible. Half a decade ago, Zipline stared down the valley of death, directly opposite one of its initial donor partners, Gavi. Zipline explained the rationale for providing additional funding, including the challenge that innovators undergo—but Gavi’s funding structure simply didn’t align with scaling effective innovation. Instead of giving up and walking away, Zipline adopted a patient relationship-building approach. For more than a year, Zipline met with Gavi counterparts biweekly—always sharing experiences from the ground, compelling data, and insights from its close partnerships with African governments. When it came time to roll out the COVID-19 vaccine in Africa, Zipline was ready to help. During that year, Zipline did not make a single request of Gavi; it just continued providing insights and ideas. And Gavi began to listen. It became genuinely curious about Zipline’s challenges on the ground with integration and scaling. As Gavi sought to do more with its resources in a poly-crisis environment, it considered Zipline’s ideas on paying for performance and how setting up a model like this could reset the lopsided incentives that exist in global health. Meetings between Zipline and Gavi became more frequent, and Gavi started to independently conceive of new ways of working with innovators. They started approaching Zipline and others for feedback, and they admitted where there were structural or process shortcomings and showed a willingness to enact change. In June 2024, at Gavi’s Global Vaccine Forum in Paris, CEO Sania Nishtar announced that Gavi would focus on scaling proven innovations in its upcoming replenishment, including a commitment to deliver 250 million vaccine doses by drone over the next five years. Hopefully, others will begin to follow in Gavi’s footsteps.

Read more stories by Christian Seelos, Miki Sofer, Caitlin Burton & Johanna Mair.