As blended learning gains greater traction across the country, conversations around blended learning sometimes tend toward celebrating what technology can do rather than what it should do. It is tempting to give in to the magnetic pull of the next best thing in technology or embrace brand new tools, especially if they can offer some cost savings in other areas like textbooks or staff time. But doing so without clear conviction of what an underlying instructional and school model aims to accomplish could derail the potential for online learning to scale in a student-centered manner.

At first blush, our own theories about online learning as a disruptive innovation can oversimplify away from talking about actual school models. Our bold prediction that 50 percent of all high school courses will be online in some form or fashion by 2019 paints a clear picture of the expanding role of technology in schools.

A tidy forecast like this, however, should not tempt people into complacency about the inevitability of online learning’s growth. Such an outlook risks missing the real crux of our argument about why disruptive innovation in education is truly powerful: in fact, this growth should not simply be about digitizing school. Instead, this disruption in the way we deliver content is an opportunity to rethink schooling.

As schools migrate to greater reliance on blended learning, we can potentially refocus our school system on personalization—for example, by letting students move at their own pace with choice among various learning pathways and modalities. And in fact, as evidenced in our blended learning research, new models are emerging. But this move to rethink school is not as inevitable as the very growth that we have predicted: to go from talking about the forecasted growth of an innovation to leveraging that growth to recreate a system requires quite a big leap both in policy and practice. The grey area between absolute growth of online learning and the potential growth of student-centered models underlying technology implementation can generate two layers of confusion as the field contemplates the role of technology in schooling.

The first potential for confusion is to stray from thinking about the model that we want to use technology to take to scale—that is, the underlying learning processes, social interactions, and skills we deem important. Clarity on this is important both in terms of supply and demand in the edtech market. On the demand side, if school systems are not thoughtful about the instructional model that they would like technology to help them scale they will likely not be discerning customers. On the supply side, then, vendors will be incentivized to create the technologies that can grow their market share as quickly as possible, rather than with fidelity to a student-centered instructional model.

One way we might solve for this is to look at models that are attempting to personalize education at a small scale without implementing a lot of technology—for example MC2, a charter school in New Hampshire that has developed a highly personalized model of learning. The school uses some technology, but most of the students’ learning is not based on online content delivery. The school is partly able to do this because it serves a relatively small student population, just under 50 students. If we can codify those elements of personalization we see as promising in such a small environment, we can nail down what processes and strategies technology and online content delivery should recreate at greater scale with fidelity to that smaller-scale, student- centered approach. That way, technology can be deployed with precision to solve particular struggles around scale, but with a clear underlying model in mind.

The second area where confusion can arise relates to the vital question of whether online learning will necessarily disrupt education in a manner that is good for students. We run the risk of failure on this front if we don’t get the policy parameters around online education and the corresponding market forces right. The current debates about gainful employment standards in higher education exemplify the fact that organizations will respond to the incentives codified in policy. For-profit universities have long been funded strictly on the basis of bodies enrolled. Their rapid expansion over the past decade, then, did not occur with a firm focus on quality because there were no policies to guide providers toward that goal. As Michael Horn has pointed out, “[w]hen policymakers sought to change the rules to instead provide funding based on whether students were benefiting, for-profits geared to the old incentives quite naturally pushed back.” In K-12 online learning we likewise need to think about aligning incentives around the quality of education technology rather than merely its expansion. Policies like performance-based funding that incorporate longitudinal measures of student growth are one place to start. Likewise, policies to urge content providers to be transparent about their results can shift our focus to seeing technology as a means of increasing student outcomes rather than the next-best input in classrooms.

Online learning is indeed taking off along a disruptive growth trajectory. As this disruption marches upmarket, however, it’s our duty to students to ensure that that growth is happening in a student-centered manner and that we’re deploying technology to scale a clearly articulated underlying instructional model.