Twin Engines for Propelling Social Impact

(Photo by iStock/RASimon)

In the 12 years since its founding, OneGoal had grown to engage almost 12,500 high school students a year in six metropolitan areas and had become widely recognized for its success in helping lower-income students—mainly young people of color—enroll and persist in postsecondary education. Yet OneGoal was hitting the limits of its relatively costly model, which required large and continuing streams of philanthropic funding. It was growing one school at a time, adding hundreds of students each year—but the need was many, many times that. “If we continued to grow at the current rate,” says OneGoal CEO Melissa Connelly, “we would never come anywhere close to reaching the 1.4 million 11th graders nationwide who could benefit.”

OneGoal’s problem is a familiar one. On one hand, they want to expand their services to meet needs which only seem to grow. On the other, they feel pulled toward identifying new solutions that hold the potential to scale their impact far beyond their current approaches. It is often hard to pursue both.

For the private sector, Bain & Company has created a concept called Engine 1/Engine 2 to describe this dynamic and help leaders effectively manage it. To sustain profitable growth over the long term, companies need to optimize the core business—their Engine 1—while sowing the seeds of reinvention for their Engine 2, which can fuel growth through a new business model, new customer segments, or completely new products or services. Engine 1 focuses on today’s business model and current capabilities, and it involves continuous improvement and careful risk assessment; Engine 2 focuses on new growth opportunities and business models that are outside of Engine 1, and involves experimentation, agility, and creativity. Bain defines Engine 2s as new businesses built within an existing company that use the existing scale benefits—namely, the assets and capabilities of the strong core business (Engine 1)—to grow to a large size faster than an independent start-up could.

Consider how Netflix became the company it is today. Some years back, Netflix was eclipsing Blockbuster by delivering movies to customers by mail, rather than maintaining expensive brick-and-mortar stores. But Netflix recognized the potential of internet streaming, and began developing its Engine 2, even as for some years the vast majority of its revenue would still come from the existing mail delivery model. Other business giants exemplify this process of reinvention, like Apple moving beyond selling computers to music services, the iPhone, and apps, or Western Union transforming itself from a telecommunications company (remember telegraphs?) to a financial services provider. By contrast, as digital photography replaced film, Kodak was eclipsed by its competitors for not investing sufficiently in their Engine 2.

The Balancing Act for Nonprofits

Many nonprofit leaders would be delighted to invest energy in reinvention and innovation, but it’s difficult to carve out the necessary time and resources. After all, what the Engine 1/Engine 2 authors describe as “the hardest act in business” might be even harder for nonprofits, whose core “business” isn’t a business at all. Because social impact affects real people and communities, it’s hard to slow near-term growth or shrink a service footprint to pursue the promise of a better (but unproven) approach. And innovation almost always entails risk, even failure, which, for nonprofits, can have a painful impact on people's lives. Finally, there’s the money. While businesses can draw on profits, debt, or capital markets to finance their Engine 2, most public and philanthropic funding in the nonprofit world is intended to pay for people served, activities undertaken, and outcomes achieved. Few funders have the degree of patience and tolerance for risk required to invest in helping organizations design, test, and build an unproven Engine 2.

At the most basic level, a nonprofit may want to consider an Engine 2 if its current trajectory will leave it far short of reaching a meaningful portion of its target population, making a lasting difference on the lives it does touch, or achieving its mission—even in 20 years. The bigger the gap, the more an Engine 2 is needed.

For OneGoal, the gap was visible when, after years of growth, its Engine 1 was engaging only 12,500 students out of a potential national “market” of 1.4 million students. That realization was a clarion call, explains Melissa Connelly, “We recognized the extreme pressure we had been putting on practitioners to both execute against our current model and reimagine it to meet the need. We decided that if we were going to reinvent ourselves to be better positioned to achieve our mission, it would require us to put real resources against that.” As a result, OneGoal’s District Partnership Pilot seeks to integrate OneGoal’s proven postsecondary success model into the fabric of a school district, training leaders within each partner district to support the model and using personalized learning technology to deliver its curriculum online.

A New Way of Working

VisionSpring is a global NGO founded in 2001 with the goal of providing affordable, quality eyeglasses to the approximately one billion people worldwide who need, but don’t have, corrective eyewear. Like OneGoal, VisionSpring had demonstrated success by growing its Engine 1 quickly, selling low-cost glasses through its own direct sales force of “vision entrepreneurs” in low- and middle-income countries. However, because VisionSpring was subsidizing each pair of glasses sold, it would never come close to reaching the size of the need.

Engine 1 and Engine 2 require different mindsets, capabilities, and success metrics. “Whereas the core part of our business is more or less set in stone,” says Jordan Kassalow, VisionSpring’s founder, “with our innovation work, the iteration cycles are much more rapid—quicker testing, lots of go/no go decisions.” Engine 1 is a structure for delivering what you already know how to do, which requires discipline, repeatability, and careful risk assessment to be successful. But because Engine 2 is a structure for developing, testing, and refining new solutions, models, and pathways, it requires agility, creativity, and a healthy appetite for risk. The traditional management process—design, then plan, and finally implement—gives way to testing hypotheses and learning from failures as well as success.

Consider a pilot that VisionSpring currently has underway to sell reading glasses through pharmacies. While common practice in the US, it’s virtually unknown in countries such as Bangladesh, where glasses are considered a medical product obtained from doctors. “This innovation is still in the pipeline, and it’s unclear if it’s going to graduate to become one of VisionSpring’s core approaches, or be a failure,” says Kassalow. There are tens of thousands of pharmacies in Bangladesh, and when VisionSpring began its pilot, none of them sold eyeglasses. VisionSpring had to figure out whether this entirely new approach to selling glasses would work, and if so, how. There were many things it would need to figure out along the way, including whether customers would purchase a pair of glasses off the shelf or need hand-holding from a sales agent, and if marketing and social media could work to change current glass-buying habits.

It started small—with just 20 pharmacies. Some of the stores were able to successfully sell glasses; others weren’t. Since VisionSpring was trying to experiment and learn rather than implement a fully-developed plan, both successes and failures provided useful information as to which approaches might work best as VisionSpring expanded the pilot to 50 and then 200 pharmacies. “Our goal is to get to 2,000 pharmacies,” says Kassalow. “The unit economics don’t work at 200 pharmacies, but we have strong indication that they will at 2,000. That’s what we need to confirm. And there are pharmacies all over the world. If we can get this to work in Bangladesh, it might work in other places.”

During the early stages of developing an Engine 2, the innovation process may feel unfamiliar and risky—but many organizations find their confidence growing with each step. (We’ve both written extensively about the new types of tools needed for innovation here and here.) Initially, when uncertainty and risk are high, small, highly-targeted experiments—sometimes called Minimum Viable Products (MVPs)—can give leaders a better handle on the biggest risks before they make more significant investments. If these experiments don’t perform as anticipated—say, if willingness to pay is lower, or fidelity of replication is poor—then it may be time to pivot and consider other approaches. But if early indicators point toward a successful pathway, it’s time to double down and invest.

Measuring success looks different for Engine 2. Traditional metrics for tracking progress such as the number of people reached may not provide much insight as to how the innovations are working. Instead, nonprofits may want to track, validate, and optimize unit-level metrics such as the adoption rate, retention rate, success rate, and unit economics. Improving these underlying drivers will in turn lead to greater and greater impact.

Structuring Engine 2 for Success

The differences between Engine 1 and Engine 2 make it challenging to run them side by side, but they can coexist within the same organization. Finding the right relationship between these two structures depends on the degree of overlap for key elements of their business model (e.g., partners, customers, core approach) as well as what will enable both engines to operate optimally. If they are too entwined, Engine 2 may struggle to get the necessary attention amid more immediate deliverables. Yet, if they are too disconnected, rivalries may emerge and important learning may be lost. “Where there is resource allocation, there’s tension,” says OneGoal’s Connelly. “Since core operations make up 90 percent of our services, we need to continue to prioritize appropriate resource allocation to ensure they can maximize effectiveness; however, since we are not continuing regional site expansion with that model, there were some roles, such as those responsible for training new regions, which were no longer a necessity.”

If Engine 1 and Engine 2 don’t have a clear through-line among strategy, staffing, resourcing, and structure, it can create real dissonance among the management team. It can be challenging to kickstart Engine 2, when an organization’s culture and capabilities have been built around its Engine 1. As a result, building the right staffing for Engine 2 may call for a combination of plucking out existing entrepreneurial leaders and bringing in new talent.

When an Engine 2 is focused on a closely aligned enhancement to an existing product, service, or business model, integrating efforts into the same team can help incorporate Engine 1 know-how into Engine 2. The same team members may be called to balance between keeping the existing program running, while exploring a new one. However, this can place a lot of pressure on team members who are required to have a role in both Engine 1 and Engine 2. “In the past, we innovated mainly by having our existing staff tinker with new ideas on the side,” explains VisionSpring’s Kassalow. This approach may also be born of necessity, in the absence of funding, and contribute to the early validation needed to secure resources for a larger investment.

A second option is a separate team or division, which can help to avoid the pressures and constraints that the Engine 1 program teams are under. Some organizations bring someone who already knows the organization well to lead or staff this team. OneGoal, for example, drew talent from its regional offices, from people who were familiar not only with the model but with the new potential customer, school districts. Other organizations may prefer to bring in someone from the outside. As one nonprofit leader explained, “We felt it was key to get a leader who’d bring a new perspective and wasn’t beholden to how we always did things around here.”

A separate team or division doesn’t have to be a fully staffed innovation laboratory. For an ultimately successful pilot that involved partnerships with factories to help workers who needed glasses get them on site, for example, VisionSpring hired one person to be in charge—focused on the pilot and on nothing else. An initial experiment that may lead to an Engine 2 could start with even less.

If a new model goes so far as to require different partners, funding mechanisms, or approaches, an Engine 2 might be best pursued as a separate entity, free of the organization’s existing constraints. This might involve spinning out a for-profit entity, joint venture, or lobbying group. For example, VisionSpring saw the need for a much broader systems-change effort to increase access to eyeglasses worldwide, so it helped to launch EYElliance, a multi-sector coalition designed to identify highly effective solutions and research and develop methods to embed them into broader public and private systems to increase access to glasses at scale. Innovating on such a broad policy and systems agenda, supported by more than 60 members, made much more sense as the work of a separate organization than as a unit of VisionSpring.

Organizations may find themselves over time pursuing multiple models and approaches through their Engine 2, as VisionSpring did. Some may reinvent how they grow, finding a new vector for scale that overshadows or ultimately replaces the Engine 1 core program. Others may end up using Engine 2 to develop an adjacent service or program that lives alongside the core work, but does not replace it.

Funding Engine 2

Organizations often point to a lack of appropriate funding as the biggest barrier to investing in an Engine 2. Three-quarters of US foundation giving, and most NGO and government funding globally, comes in the form of restricted grants rather than the type of flexible funding that enables innovation. And research from Bridgespan and Echoing Green shows that Black-led nonprofit organizations have significantly more difficulty accessing capital, whether for existing programs or for innovation.

One nonprofit leader we spoke with, who has spent a lot of time working to get his organization’s Engine 2 off the ground (not either of the organizations discussed here), told us, “I thought it would be easier to raise money for it. Funders said: ‘this new approach sounds interesting, come back in a few years when you’ve demonstrated it works.’ We’ve had a long track record with a lot of funders. I was surprised how few of them wanted to give money to try something different.” (Ultimately, two major funders did support that organization’s Engine 2 pilots.)

While funders could certainly do more to incentivize innovation with flexible and risk-tolerant grants, nonprofits have options. At the time of writing, both VisionSpring and OneGoal had attracted and carved out funding for their Engine 2 innovation that amounted to about 10 percent of their organizations’ budgets, with the remaining 90 percent going to Engine 1 core services.

Moreover, pursuing innovation need not be expensive. A common mistake is assuming that Engine 2 requires a big grant and big team to before getting started. But there is real value in making small bets before making a major investment in an untested new approach. VisionSpring, using what Kassalow describes as a “tinkering” approach, produced some successful innovations before the organization created a more formal budget and process for its Engine 2. It may be possible to run quick, cheap, validation sprints, producing early data points that indicate whether a new approach holds promise (enthusiastic reception by clients, early indicators of outsized results, or engagement from potential renewable sources of funding). This early experimentation can in turn demonstrate enough potential to attract funders to the table with a somewhat larger investment to move to a next stage, which is still not necessarily expensive or elaborate.

Several years ago, VisionSpring created a dedicated fund—drawing dollars from earned income and philanthropy—and a process to develop and test new ideas for scale. However, explains Kassalow, “We don’t build up expensive infrastructure around the innovation until we figure out if it’s a go or no-go.” For those first few experiments that may take a few weeks or months, existing unrestricted dollars could be set aside. In other cases, funding for innovation has come from specialized vehicles such as the Global Innovation Fund, philanthropic efforts such as The Studio@Blue Meridian (which is a major source of support for OneGoal’s innovation work), or high-net-worth individuals, all of which provide unrestricted dollars and are often more open to innovation and risk than some traditional foundations.

Another way that nonprofits operating on restricted philanthropy grants can fund an Engine 2 is to build a small innovation window into their grant proposals. For example, 95 percent of the funding could remain focused on delivering the desired short-term results through Engine 1, while 5 percent is set aside to invest in the early validation of promising new ideas. In cases where such early investments in Engine 2 reveal strong potential, the case for placing a bigger bet with a bigger investment will become much easier.

Destination: Impact

In this article, we have focused on structured innovation—an Engine 2 that runs alongside the organization’s core Engine 1 programs—as a way of conceiving how ambitious nonprofits can think about balancing today’s needs with tomorrow’s potential. While an Engine 2 can help the organization move into an adjacent field or develop altogether new products or services, we’ve focused on how Engine 2 can help to scale an organization’s reach in an existing field. Whichever the approach, the steps you would take to pursue an Engine 2 are similar. 

Scale is usually not an end goal, of course. Scale is a means to achieve greater impact and progress toward equity and justice. Unfortunately, some innovations—while proving themselves successful in terms of feasibility, demand, and economics—can end up steering an organization away from its core mission and population of focus. We know of one nonprofit, for example, that extended its impact with a set of well-designed and market-tested online resources. But most of those using the online resources turned out to be from more affluent communities, while the people at the heart of the organization’s mission—youth, families, and communities of color—were getting little benefit. Recognizing this, the organization’s leaders pivoted to pursue a different, more targeted approach, which was more effective at reaching those communities. The impact you seek matters to the ideas you pursue as part of your Engine 2: If it’s successful, will this new approach benefit those who are most impacted by the problem you seek to solve?

We recognize that starting an Engine 2 may seem daunting. At the same time, the COVID-19 crisis forced many to recognize that they can build this muscle. How many organizations mobilized to pilot new online approaches to serving clients, within days and weeks, which had previously been stuck in debates and planning for months or years? Many quickly discovered that their services could be provided at greater scale, less expensively, and, in many cases, far more effectively than they thought was possible by going online. As restrictions imposed by COVID-19 are removed, a core question many will need to face is how to continue to innovate—not only which of the changes in approach should be preserved, but how many other opportunities for growth and impact could they have been missing? Investments in Engine 2 are an essential means of uncovering those opportunities in the pursuit of equity and impact as a sector.

* * *

The authors would like to thank James Allen of Bain & Company for guidance on the Engine 1 and Engine 2 concept, as well as Katherine Kaufmann, Dorothy Jones, Bradley Seeman, and Larry Yu of The Bridgespan Group for contributing insights and research support.

Read more stories by Ann Mei Chang & Laura Lanzerotti.