Social Innovation Doesn't Have an Ideas Problem. It Has an R&D Problem.

In November 2020, I decided to start a nonprofit.

Over the previous six months, my team and I had been testing a different approach to food assistance: providing families with consistent, community-based access to fresh fruits and vegetables through a model built around neighborhood schools, food recovery partners, and local farms. We wanted to answer a bigger question: Could food assistance do more than relieve hunger? Could it also improve nutrition and health, support student success, center the human experience, and strengthen environmental and food-system sustainability?

I came to the work after spending six years as a founding executive at LearnPlatform, a high-growth education technology company built around another big question: How can school districts know whether their technology investments are actually working?

At LearnPlatform, experimentation was part of the operating model. We developed products, tested assumptions, gathered evidence, learned from users, refined our approach, and raised investment to support growth. When something showed promise, there was an ecosystem designed to help us take the next step.

When I started +BOX, I assumed much of that same entrepreneurial approach could be translated into the nonprofit sector. While it could—it was not nearly as easy.

I certainly borrowed from my LearnPlatform experience, but I quickly discovered that the operating systems, incentives, infrastructure, and funding models of the nonprofit sector make experimentation unusually difficult. Organizations tackling some of society's most complex problems are often given remarkably little room to test new ways of solving them. Over time, I've come to think about that challenge through what I call the Four Gaps.

The Four Gaps That Stifle Social Innovation

1. The Knowledge Gap

Many social impact funders lack the time, expertise, or incentives to distinguish between fundamentally different approaches to social impact, often rewarding familiar models over innovative solutions. Individual donors, family foundations, community foundations, governments, and institutional grantmakers make decisions across extraordinarily complex issues. They cannot reasonably be experts in every intervention, operating model, or theory for achieving social impact. As a result, fundamentally different approaches can be grouped together based on broad charitable outcomes rather than their underlying design, strategy, or effectiveness. 

+BOX, for example, was often evaluated alongside traditional food banks and pantries despite operating from a fundamentally different nutrition-focused, outcomes-driven, community-based model. When those distinctions are difficult to see or evaluate, familiarity can become a proxy for confidence—and genuinely different approaches can struggle to attract investment.

2. The Value Gap

Because funders are rarely the direct beneficiaries of the work they support, nonprofit organizations can be evaluated more on perception and familiarity than on the value they create for the communities they serve. In the private sector, customers generally experience a product or service directly and decide whether it creates enough value to warrant paying for it. That creates an imperfect but powerful feedback loop.

Social impact works differently. The people funding a service are typically not the people receiving it. That separation weakens traditional market feedback loops and makes it harder to determine which approaches beneficiaries actually value, which produce better outcomes, and which should receive greater investment. Storytelling, reputation, organizational familiarity, and easily communicated outputs can therefore exert disproportionate influence over funding decisions.

3. The Resource Gap

Nonprofits are expected to solve complex social challenges while operating with chronically insufficient and often restricted resources, limiting their ability to invest in the people, infrastructure, innovation, and capacity needed to achieve lasting impact. Funding is frequently restricted, short-term, or tied to specific programs. Meanwhile, investments in staffing, technology, operations, infrastructure, organizational capacity—and especially research and development—can be difficult to fund.

Compare that with the private sector, where investing in product development, technology, talent, experimentation, and growth is considered necessary to build a successful organization.

Nonprofits are often expected to innovate while simultaneously minimizing the very investments that make innovation possible. And when promising ideas do emerge, there are few established mechanisms for financing their continued development outside the organizations and grants that produced them.

4. The ROI Gap

Because social impact is difficult to measure and compare consistently, uncertainty can make investing in innovation feel unnecessarily risky. Social outcomes are complex. They can take years to materialize, depend on numerous external factors, and resist simple comparisons across organizations or interventions. In the absence of clear measures of return, funders can understandably gravitate toward lower-risk, highly visible activities with easily quantifiable outputs. Pounds of food distributed. People served. Classes delivered. Beds provided.

Those measures matter. But they don't always tell us whether people's lives actually improved—or whether a different approach might produce greater value. The result can be an environment in which incremental improvement is easier to finance than experimentation, even when experimentation could uncover fundamentally better approaches. 

Taken together, these gaps create a difficult cycle: we don't always have enough evidence to know what works; our funding systems don't always reward the approaches creating the greatest value; organizations lack resources to experiment; and the resulting uncertainty makes investment in new approaches feel risky.

Until we begin closing those gaps, social innovation will remain harder than it needs to be—and promising opportunities to create both social and economic value will continue to go unexplored.

What Happens When We Actually Fund the Experiment?

There are, of course, exceptions.

Some philanthropists, governments, and other funders are willing to invest in questions rather than predetermined answers. One of those opportunities allowed +BOX to test another idea.

As our food assistance program grew, we encountered a persistent challenge: It can be surprisingly difficult for families to find, enroll in, and participate in programs designed to help them. +BOX tried to reduce that friction. We distributed through neighborhood schools, meeting families somewhere they already went. And we used a no-questions-asked enrollment policy rather than requiring families to prove that they were eligible for help. Those choices made the program easier to access—but created another problem.

Because we intentionally didn't ask families to enroll or demonstrate need, we had limited information about the people participating in our program. That could put us at a disadvantage with funders who reasonably wanted to understand whom their investment was reaching and whether resources were reaching intended populations. 

That tension led to a new question: Could we integrate existing systems to make it easier for food-insecure families to find and access the support they need—while still capturing necessary program information?

With funding from the USDA Local Food Promotion Program and in partnership with Bella Mente Montessori Academy, we launched +BOX Connect to find out. We explored whether existing school data systems could be used to make food assistance easier and more accessible for families. We designed and developed a mobile-optimized web application that allowed families to privately opt in to available food assistance programs, reducing barriers to access while automating much of the administrative work required of schools and service providers. Then we tested it.

A working minimum viable product was piloted in a school setting, allowing us to test assumptions, identify improvements, and establish a potential pathway toward broader implementation. It was exactly the kind of experimentation I believe the social impact sector needs more of.

And then we put it on a shelf.

The Messy Middle of Social Innovation

The reason wasn't particularly dramatic.

The +BOX team had managed the Connect feasibility study while simultaneously delivering our core program to more than 600 households every week. When the funded project ended, so did the dedicated resources to keep developing it. We had a nonprofit to run.

So a promising idea had been researched, funded, prototyped, and tested. We had learned enough to identify a potential path forward. But we didn't have the capacity or funding to take the next step. +BOX Connect is still sitting on that proverbial shelf today.

I suspect there are thousands of versions of that shelf across the social impact sector: feasibility studies, prototypes, pilot programs, research findings, half-developed partnerships, and promising ideas that got far enough to demonstrate potential but not far enough to attract the resources required for what came next. And many ideas never even make it that far.

The people closest to our most persistent social challenges often have extraordinary insight into how things might work differently. What they frequently lack is the time, funding, technical expertise, organizational capacity, or institutional home necessary to turn that insight into a testable idea. That's the gap I'm increasingly interested in.

Other sectors have built infrastructure specifically for this messy middle. Businesses invest in R&D. Startups have accelerators, venture capital, product-development methodologies, and established ways of testing concepts before scaling them. Universities have research infrastructure. The social impact sector has pieces of this infrastructure, too—but access to it is inconsistent, fragmented, and often unavailable to the people and organizations closest to the problems we're trying to solve. 

The challenge isn't simply a shortage of funding. It's the absence of reliable pathways for turning promising research and experimentation into practical solutions—and, where the opportunity exists, sustainable products, services, or businesses.

Why ARC

That's why I'm launching ARC Innovation Lab.

ARC exists to give promising ideas for social change a better chance to succeed. We combine applied research with entrepreneurial product development to identify, test, and build better solutions to complex social challenges.

We believe some of our most persistent social challenges contain overlooked opportunities to create both social and economic value. ARC exists to discover and develop those opportunities.

Working with the people and organizations closest to a challenge, we investigate problems, identify opportunities, design and test potential solutions, and generate evidence about what works. Sometimes that work produces valuable research, practical tools, or new operating models. And when an idea demonstrates sufficient social impact and commercial potential, we can develop it into a sustainable product, service, or business.

Importantly, the objective isn't to prove that every idea works—or that every solution should become a business. Some experiments should lead to scale. Others should lead to refinement, a different approach, or a decision not to proceed at all. Finding out that an idea doesn't work before investing millions of dollars in it is also a successful outcome of R&D.

I've seen what can happen when that process works. The original +BOX program began as a small experiment serving 50 families at one elementary school. Over the next four years, we tested, refined, and expanded the model, ultimately distributing more than 1.4 million pounds of fresh produce. In January 2025, +BOX merged with Brighter Bites, a national nutrition and health nonprofit that had the infrastructure to take the work further. Since then, the San Diego program has more than doubled its reach while operating at roughly half the cost.

One experiment found its pathway forward. Another—+BOX Connect—is still waiting for one. 

Both experiences helped lead me to ARC.

The social impact sector doesn't lack entrepreneurs, practitioners, community leaders, or promising ideas. And I don't think the answer is simply asking social impact investors to "take more risks." We need better ways to reduce the uncertainty around those risks. We need more opportunities to ask difficult questions without already knowing the answer. More capacity to turn ideas into experiments. More funding for the messy work between concept and scale. And better evidence for deciding which innovations deserve our next dollar, our next year, and our collective attention.

That's the kind of infrastructure I hope to build through ARC.

And the Four Gaps are only half of the story. If they help explain why social innovation gets stuck, the next question is more important: What would we need to change to help more promising ideas move forward?

That's where I'll go next: The 10 Shifts Needed to Unlock Social Innovation.