Exciting news hit the presses early last week—the XPrize is funding its first edtech competition. The goal: to handsomely reward the team that develops the best open source, scalable adaptive software to help children in developing countries teach themselves basic literacy and math.

As I’ve written about before, prizes are effective pull mechanisms to expedite R&D across a field and to in turn fill a gap that the market is currently failing to supply at scale. Unlike so-called push mechanisms that reduce the cost of R&D by directly funding research upfront, pull mechanisms incentivize private sector engagement and competition by creating viable market demand for specific products to solve specific problems. The XPrize is a good example of a pull mechanism, as are government Challenge Grants and social impact bonds.

A number of the past week’s headlines have characterized this new XPrize as an attempt to “disrupt” global education. The competition’s parameters indeed have the trappings of a disruptive play. Many disruptive innovations get their start by extracting growth from nonconsumption: that is, they target customers who are trying to get a job done but lack the money or skill and for whom a simple inexpensive solution has been beyond reach. Disruptive technologies also tend to be easier to use—even if the technology underlying a product is sophisticated, it is its “foolproofedness” that creates new growth by enabling people with less money, expertise, and training to begin consuming. Eventually, these more accessible and affordable technologies can improve over time to serve more and more demanding customers and change entire industries.

The Global Learning XPrize explicitly targets massive global nonconsumption of education. As XPrize Founder and Chairman Peter Diamandis told Wired, “We’re aiming at kids who live in villages where there’s nothing. This has to take them from complete illiteracy to basic reading, writing and numeracy.” Moreover, by outlining the need for students in these communities to learn autonomously, the XPrize committee appears to be pushing for the very foolproofedness that can allow new technologies to scale across nonconsumers who lack resources and expertise.

But the prize’s emphasis on software innovation may oversimplify the fact that to grow successfully, disruptive innovations often require a whole new value network. New customers typically purchase the disruptive product through new channels and use the disruptive product in new venues. A “channel” refers to not just outlets like wholesale distributors and retail stores, but any entity that adds value to or creates value around the company’s product as it wends its way toward the hands of the end user. For example, a physician’s practice is a channel through which many health care products provide needed care to patients.

Disruptive value networks and channels are important concepts to consider in any education market and one that any new software product must take into account. In education, a value network is the context in which students consume education—this is often defined by policies or informal norms that supply public funding or dictate behavior, as well as educational resources at students’ disposal. The Global Learning XPrize is aiming to spur the design of software that can serve students in as direct, unmediated manner as possible. But even laser-like focus on developing a software program that successfully allows children to teach themselves should not presume that education itself is occurring in a vacuum—indeed, everything from hardware to physical settings, peer engagement to adult supports, culture to connectivity will likely play into the success of failure of a given software product in the developing world.

Diamandis is already thinking through some of these complementary elements, such as partnerships with hardware providers to ensure that the open source software can be deployed alongside free or affordable devices. Still, the XPrize’s current framing draws a sharp dichotomy between autonomous learning and the concept of education occurring in some broader context or circumstance. As our blended-learning research demonstrates, educational software can provide the fuel for personalized-learning models. But these models are defined not only in terms of their use of software and hardware to drive personalized learning. Models are also characterized by the particular brick-and-mortar settings where students go to learn; new roles for adults—even if these adults are not trained as traditional educators—who care for, support, or motivate students; and sometimes even new distribution channels, like libraries or community centers where students can access online-learning programs. Particularly for younger children, the physical setting and face-to-face supports in many blended-learning models may be important factors in scaling sustainable learning platforms, even if the bulk of content delivery is occurring online.

To ensure that the $15 million dedicated to the Global Learning XPrize spurs new models for learning, the prize committee should thoroughly consider the entire value network in which these new software products stand to operate. As they evaluate what is sure to be an impressive suite of submissions, the prize committee should conduct research on what products appear to be working for what students in what circumstances. They should take pains to describe these circumstances in great detail, and to test assumptions about what contextual factors make a model successful. This would likely help the XPrize Foundation to reveal a range of different models that work for different children—not merely a one-size-fits all approach. Evaluating submissions based on these contextual factors may also encourage software developers to think in terms of models rather than products, shepherding both technical and instructional talent to meet the Foundation’s ambitious goals.