As the call for personalized learning gains steam, we’re hearing more about how best to invest in systems that support schools that break the traditional mold. Although advocates are pleased by having a growing number of proof points out there, a handful of individual schools that depart radically from the traditional school design and perform well is not enough: we should be striving for transformation at scale. To get there, oftentimes reformers, philanthropists, and policy analysts talk about nurturing the “ecosystem” for reform in which key players come together to drive change.

But to innovate effectively at the systems level, our efforts need to be about much more than coordinating across sectors, subsidizing a collection of loosely aligned experiments, or getting everyone around the same table. We need a theory about why particular systems-level reforms should lead to different results and why past efforts to invest in whole-system change have often fallen short. If we look to how disruptive and sustaining innovations have unfolded across numerous sectors, then we can get much more specific about what an ecosystem must look like to create lasting change.

In thinking about scale, reformers ought to get far more specific about what they mean by “ecosystem” by exploring the concept of value networks. A value network is the context within which a firm establishes its cost structure and operating processes and then works with suppliers and channel partners to respond to customers’ common needs. As I wrote about earlier this month, disruptive innovations need new value networks in which to operate. When reformers try to nurture innovation-friendly ecosystems they may think they are creating new education value networks; but more often than not, they are just supporting bold experiments within the existing value network of traditional education. So long as the school system is still operating within the traditional value network—based on levers like traditional funding streams, contractual relationships, and accountability frameworks—disruptive innovations in that ecosystem will not thrive or transform the ecosystem in the ways we might hope. This explains why more often than not, lots of investment and even the political will to change in a given geography may yield modest gains relative to the traditional system, but won’t predictably transform schools.

In large part, focusing on building a value network means getting the incentives that animate players in the system—not the inputs—right. For an ecosystem to nurture “next-gen” or personalized learning models, we need a value network that is organized around accountability to individual student outcomes and growth. Without that keystone in place, investments will just fund “innovative” or personalized inputs without necessarily seeding any of the outcomes we hope such approaches would yield. Focusing on outcomes will depend on transforming accountability and funding. A better funding system would reward successfully driving individual student performance among both schools and EdTech providers. Take, for example, the manner in which the state of New Hampshire funds the Virtual Learning Academy Charter School (VLACS), a statewide source for online learning opportunities. Because New Hampshire is one of few states to have gone fully competency-based, VLACS’s instructional and funding models are contingent on students advancing—and being funded—upon demonstrating mastery. Other examples of new value networks include experiments like ReSchool Colorado and the emerging micro-schooling movement. Both are worth watching, as they are attempting to grow an entirely new system with wholly new incentives centered on students and families.

Thinking through an education ecosystem should not be about sketching out the education system of the future in vague, idealistic terms or simply trying to get businesses, schools, and tech firms to align around supporting students. It should be about clarifying the innovations we are trying to seed, identifying new value networks to support different types of innovation, and nailing down the right outcomes before we pour money into inputs. These efforts will not inevitably ensure our success, but they stand to create conditions far friendlier to innovative school models that don’t fit the old mold.