With COVID-19 cases again on the rise, schools are once again closing their doors this year. Meanwhile, the toll on many educators, students, and families feels crushing. It may be the biggest understatement of 2020 to say that the pandemic has thrown K–12 environments into turmoil.

By forcing conventional classrooms online, the pandemic exacerbates the challenges of keeping students engaged and on-track through single-paced, whole-group instruction. Frustrated by this reality, a sizable proportion of students, parents, and teachers have left their schools to seek alternatives such as homeschooling, virtual schools, micro-schools, and learning pods. These shifts beg the question: are we witnessing the beginning of K–12 disruption—the type Clayton Christensen described?

What is Disruptive Innovation?

Before I dive in on answering that question, let’s clarify what Disruptive Innovation means. Disruption, as defined by the Christensen Institute, is more than just upheaval. In our research on how organizations and sectors change, Disruptive Innovation describes a phenomenon whereby the incumbent technologies and business models in a sector get displaced by new technologies from new entrants in that sector.

The phenomenon starts when new entrants come up with seemingly inferior solutions to serve people who are unserved (nonconsumers) or overserved by established incumbents (low-end customers). Then, over time, these new entrants improve their offerings into compelling alternatives to incumbent solutions. The disruption of the sector unfolds as the people previously served by incumbents migrate to good-enough solutions from new entrants that also offer greater affordability, customizability, and convenience. Consider, for example, how Amazon disrupted conventional retail by initially setting up a website to sell books; or how Netflix disrupted Blockbuster by initially mailing people DVDs.

When Clayton Christensen first described this phenomenon in the mid-1990s, it captivated the attention of the business world because it explained how leading incumbents in an industry—the organizations that seemed to be doing everything right—were routinely upended by new entrants with seemingly inferior products. The implications for business strategy were huge, and the term quickly became a buzzword in both board meetings and startup communities.

Nonetheless, Disruptive Innovation is a phenomenon, not an agenda. Like gravity, it happens as a course of nature when certain conditions are true. It is neither inherently good nor bad, but it can have both positive and negative consequences. Disruptive Innovation is often painful for the people who lose jobs and investments in the decline of incumbent organizations. Furthermore, any technological progress, whether disruptive or otherwise, can have unanticipated consequences—consider social media’s impact on political polarization and mental health. Nonetheless, Disruptive Innovation routinely produces marked social benefits. The technologies that flow from disruption are more affordable, customizable, convenient, user-friendly, and accessible than their predecessors. This means that advances in communication, transportation, health care, and the like that come about through Disruptive Innovation end up helping more people in more places.

Are K–12 schools getting disrupted?

The big disruption-related question right now in K–12 education is whether learning pods, micro-schools, and other alternatives will disrupt conventional schools. At first glance, these models seem to fit the pattern: in their infancy, they’re unable to offer everything conventional schools provide; but they initially serve students and families who are locked out of childcare, structured learning environments, and other services they need from conventional schooling. However, checking the box on a few disruption look-fors doesn’t guarantee disruption any more than having wings can guarantee that ostriches and penguins will fly.

The nonconsumption problem

The pandemic has certainly created nonconsumption—a key starting point for disruptive innovations. Until last March, many families relied on schools to provide a range of services, including reliable childcare, social communities, and a structured setting for learning. Then COVID-19 created a void, and pods and micro-schools sprung up to fill in.

But this response to nonconsumption differs from the normal pattern of disruption. Normally, serving nonconsumption gives disruptive innovations an opportunity to grow and improve without inciting a competitive response from incumbents. But the nonconsumers of schooling today are not families and students that established schools routinely ignore. Rather, established schools see students who leave for pods or micro-schools as lost enrollments; and as the pandemic draws on, conventional schools will respond to try to rake those enrollments back. Some are already setting up learning hubs and virtual schools of their own.

The affordability problem

Second, pods and micro-schools are too expensive and complicated to be disruptive. Disruptive innovations don’t enter as premium offerings that command premium prices. Rather, they appeal to people by offering affordability and convenience. In contrast, parents take on a massive workload to create their own pods or micro-schools. Some may outsource that work by hiring a teacher to man the ship, but then affordability goes out the window. Splitting the cost of a teachers’ salary among a group of 2-3 families is nowhere near as cheap as free public education. So when it comes to affordability, the only potential disruptive path for pods and mirco-schools is at the low end of the private school market.

The improvement problem

Last of all, pods and micro-schools won’t have the runway they need to be disruptive. Disruptive innovators start off crummy and only start pulling people away from the mainstream options after they’ve had time to improve on their early iterations. Remember when YouTube was only about cat videos and Netflix was only mailing DVDs? Both had a long way to go before challenging video rental and cable networks. For pods and micro-schools, the runway offered by pandemic-induced nonconsumption will likely run out some time next year—long before they can evolve into attractive alternatives to conventional schooling. When the pandemic ends, most families are going to say “good riddance” to their COVID-19 stop-gaps and go back to the incumbent system they’ve long relied on.

The outside chance

There is one outside possibility that could change this equation. If the strains of the pandemic push a sizable number of families into pods, micro-schools, homeschooling, and other non-conventional schooling options, there’s a chance the needs of those families could generate the political will to change states’ education policy. If advocates can convince states to give education dollars directly to students rather than funding the schools those students attend, micro-schools, pods and other alternatives will suddenly become able to match conventional public schools on affordability. They’d also then have the runway they need to improve over time while serving the small subset of families that will want to stick with them post-pandemic.

Such a massive change in education funding policy isn’t likely to happen before a vaccine brings school buildings back to full operation. But if by chance we see some states adopt policies resembling universal education savings accounts (ESA’s), the conditions for system-wide disruption will be at play.

Where are the opportunities for innovation during COVID-19?

At the Christensen Institute, we’ve long been on the lookout for disruptive innovations that could make education more customized to students’ learning needs and increase access to valuable learning experiences. In their 2008 book, Disrupting Class, Clayton Christensen, Michael Horn, and Curtis Johnson tagged online learning as the type of technology that could catalyze Disruptive Innovation in K–12 education. It’s important to note, however, that the authors never predicted that online learning would disrupt K–12 public schools. Instead, they predicted that disruption would unfold at the level of instruction, not governance. Hence why Christensen, Horn, and Johnson titled their book Disrupting Class, not Disrupting Schools. In response to the burgeoning question of whether COVID-19 has begun the disruption of K–12 schools, the answer is “no.” However, although the pandemic isn’t likely a catalyst for the disruption of K–12 schools, that doesn’t mean worthwhile innovation during the pandemic isn’t happening or won’t happen, or that Disruptive Innovation won’t be at play. In an upcoming post, I’ll detail where innovations are happening in K–12 education as the pandemic carries on into 2021. Here’s a hint: it’s less about different school systems and more about opportunities for existing schools and districts to develop new instructional models—just as Christensen, Horn, and Johnson predicted.

This story on how blended learning can help schools reopen during the COVID pandemic, originally published on June 15, was eSN’s No. 9 most popular story of 2020. Check back each day for the next story in our countdown.

As educators start considering their options for the fall, the future is full of uncertainty. If schools remain closed, they’ll need to prepare for more remote learning. On the other hand, there’s a chance schools might be able to open back up, in which case they’ll likely need to have students come in shifts in order to maintain social distancing.

Fortunately, if bringing students to school part-time is an option, schools don’t have to invent new approaches from scratch. Two of the blended-learning models we’ve documented are well suited to these circumstances: the Enriched Virtual model and the Flipped Classroom model. There is a caveat: They all hinge on internet connectivity, a challenge that both public and private efforts are moving quickly to try to solve.

Related content: 10 blended learning resources for schools

In an Enriched Virtual arrangement, students complete the majority of coursework online at home or outside of school, but attend school for required face-to-face learning sessions with a teacher a few times a week. In Flipped Classroom setups, students learn at home via online coursework and video-recorded lectures, and teachers use class time for teacher-guided practice or projects.

Normally, the Flipped Classroom model is used by teachers whose students come to class every day, whereas the Enriched Virtual model is generally deployed to increase support for students in virtual schooling who normally wouldn’t attend a brick-and-mortar school at all. In COVID-19 circumstances requiring a reduction of the number of students in school buildings on any given day, however, the distinctions between these models become a bit blurry. Here are some examples of what these types of models look like.

Combining independent online learning with face-to-face instruction

Scott Nolt, a high school history teacher in North Carolina, has been using blended learning for years as a way to teach his students without daily direct instruction. At the school where he first developed his approach to blended learning, roughly half of his students met in class with him on any given day while the other half learned independently, often at home. At his current school, students were in class with him every day until COVID-19 closed their campus, but spent most of their time working independently on course activities while Nolt worked with small groups in a study-hall-type arrangement. This approach allowed him and his students to transition seamlessly to distance learning once schools closed.

During their online learning time students dive into new content by studying texts, watching online videos, and taking assessments that help them develop their basic understanding of course content. They then turn to individual assignments and group discussions that help them develop their analytical, critical thinking skills. While working online, students communicate with Nolt through an online messaging system and through the comment features built into online assignments and discussions that allow him to give feedback on their work.

The blended-learning approach allows students to explore concepts at different depths of understanding, based on their interests. All students are expected to learn the “big idea” for each lesson or unit, but can then choose activities for going deeper on topics that interest them. They then demonstrate their mastery of course concepts through essays, PowerPoint presentations, or other types of assignments.

Using blended learning to strengthen relationships

Stacey Roshan, a high school math teacher in the Washington, DC, area, became a teacher because she wanted to share her love of math with students and help combat the notion of “I’m not a math person.” She wanted her class to be a place where exploring math was fun and where students knew she cared about them as individuals. But for many of her AP Calculus students, it was hard to love math given their nightly struggle through homework assignments with unanswered questions and the enormous pressure to get straight A’s. Meanwhile, back in the classroom, Roshan found there was little time to focus on students’ needs when she was busy trying to ensure that she delivered all the content they needed to tackle their nightly assignments.

To address these challenges, Roshan turned to blended learning. She “flipped” her lessons into online videos for students to watch at home. She then shifted class time toward addressing students’ questions and misunderstandings as they worked through problems, discussing and exploring math concepts conceptually, and engaging with students in a more personal way.

As Roshan explains in her book, Tech with Heart, “It’s important to understand that my flipped classroom is not about videos at home and textbook work in class. It is about easing students’ anxiety by giving them time to work through problems with their peers and with me. It is about personalizing the learning space, building relationships with students and gaining their trust, and being there to support them when they need me the most.”

Making these models work for the fall

Most of the flipped classroom and enriched virtual models that we’ve studied come from the high school level because elementary students are normally at school every day and have relatively little homework. But considering that COVID-19 will likely require all students to learn at least part-time at home, these models are worth consideration for elementary schools as well. Both models offer a blueprint for how teachers can design instruction to meet the constraints—and even opportunities—of students coming to school in shifts to meet distancing requirements.

To learn more about blended-learning models similar to those described above, I recommend exploring the content on the Blended Learning Universe and browsing the blended learning mini-course available at Khan Academy. For those interested in learning the ins and outs of setting up models like these, the Modern Classrooms Project offers some great professional development resources as a starting point.

We’re swimming in a world of technologies that have huge implications for the future of schooling. Computers live in teachers and students’ pockets, the Internet makes information and media ubiquitous commodities, apps offer to aid with all our daily routines, and artificial intelligence unlocks new possibilities for differentiated instruction.

Yet even with these technologies flooding into schools and classrooms, computers won’t be replacing teachers any time soon, and that’s why now, more than ever, teachers should be given the critical support they’re asking for in the classroom. To illustrate this point, consider these three things educational technology can’t do.

1. Technology can’t … provide higher-order feedback

Software is great for generating immediate, automated feedback on students’ mastery of basic knowledge and skills. But higher-order feedback falls outside its purview.

Consider, for example, essay grading. For years, word processors have been able to point out corrections for spelling and grammar errors; and more recently, intelligent software now offers feedback on elements of structure and style—such as whether a student has a topic sentence at the beginning of each paragraph and whether each paragraph contains evidence related to the topic sentence and the essay’s thesis. But software cannot give feedback on many of the qualities that really define great writing—such as whether the students’ rhetoric and logic will resonate with her intended audience. It takes a human to give feedback on the more nuanced aspects of human communication.

Related: There’s an emotional side of edtech—and it’s affecting school innovation

The same holds true in other academic domains. Software can’t tell students if their research questions for a science project are worthwhile and reasonably scoped; nor can it tell them which engineering and design challenges they should tackle to improve a simple machine. And software can’t coach soft skills—such as working effectively in teams, navigating interpersonal conflict, setting personal goals, and persevering through obstacles. In sum, the skills students will need to future-proof their careers against the rise of machines are also the skills they can’t learn from machines.

2. Technology can’t … get to know a student

Software today can potentially access a lot of data about a student: home address, race and ethnicity, diagnosed learning disabilities, family income, attendance records, test scores, browser history, and even keystrokes and mouse clicks. But with all that data, can software really know a student? Can a computer understand how his social status at school leads him to feel when he’s assigned to work with a particular group of peers on a class project? Or can it predict that she’ll enjoy reading a particular novel because it reminds her of her best friend from the town where she used to live? Software can make a lot of useful inferences based on patterns it finds in the data it collects. But it can’t collect data on all the important factors that shape a students’ learning experiences, nor can it model all of the psychological complexity of childhood and adolescence. Real knowing and understanding is a human-to-human experience.

3. Technology can’t … care about a student

Where do students get the motivation to learn? At times motivation may come from pure intellectual curiosity. But more often than not, motivation comes from relationships. For example, a student stays after class for extra tutoring because he cares what his parents think of his grades and he believes his teacher’s confidence that a little extra practice will help him get the grades he wants. Or a student becomes excited about science because a teacher who cares deeply about her also cares deeply about science. Students often work to learn and to achieve for the praise and approbation of people who matter in their lives. Software, for all its wondrous abilities, can’t offer that sense of genuine caring.

Related: Are your edtech tools really working?

But teachers can’t …

As the examples above make clear, teachers shouldn’t worry about software and devices taking their jobs. The kind of sci-fi-level artificial intelligence that might do the things described above is a long way off. This means the odds of teachers being replaced by machines are pretty slim. Teachers are a performance-defining feature in even the most technologically sophisticated learning environments. Any edtech aimed at replacing teachers sets a low bar for quality education.

But here’s the rub: while software can’t do the things listed above, neither can most teachers. For teachers, however, it isn’t a matter of capability, but a matter of capacity. How often do English teachers conference individually with their students about the quality of their rhetoric, especially when correcting grammar and structure already takes up so much of their time? How many teachers have time to meet regularly with each of their students one-on-one just to ask about how they’re doing, let alone attend all their students’ extracurricular activities or visit their students’ homes to get to know their families? Caring about students isn’t constrained by time, but showing that you care is. Unfortunately, when push comes to shove, most teachers’ days quickly fill up with planning lessons, writing quizzes, running copies, covering content, participating in staff meetings, and grading lower-order assignments; with little time left for many of the high-value activities described above.

This is why edtech is so important for teachers, in spite of all the things it can’t do. If we want all students to thrive in an increasingly complex world, we need teachers to spend more time developing students’ higher-order skills, tailoring learning to their personal circumstances, and building strong relationships with them. A key way to unlocking teachers’ capacity comes from using technology to take lower-order work off teachers’ plates.

In short, educators need technology to automate some of their work so they can focus on the important work technology can’t do. For anyone interested in advancing this idea, we’re developing a new framework to help educators think about how technology amplifies teacher capacity. We’d love to hear what you think.

[Editor’s note: This post originally appeared on The Christensen Institute’s blog and a version of this post originally appeared on The 74 Million. ]