Why are schools going blended? The answer is certainly mixed depending on whom you talk to. In some senses it doesn’t matter—whether it’s to increase students’ access to a broader range of courses or to maximize small-group instruction, hopefully the end result means that technology is helping schools deliver content in new, effective ways. Still, zeroing in on schools’ precise motivations—and forcing schools to articulate those motivations—might be an important first step toward insuring that we are sorting information about the academic software market to identify successful tools for personalized learning.

Let’s take one example. We often hear that one of the primary virtues of blended learning is the opportunity to maximize small-group and one-on-one instruction. For example, at last year’s LearnLaunch Conference, Diego Arambula of Summit Public Schools told a compelling story about how Summit’s teachers decided to offer a “genius bar”—a tutoring resource, of sorts—in place of whole-class lectures (since then, Summit’s model has continued to shift and no longer uses a genius bar approach). This decision came on the heels of Summit’s teachers testing the online and offline modalities through which students learned best. After allowing students to opt out of whole-class lectures, the teachers discovered that the most effective face-to-face instruction was happening in small groups, wherein intimate question and answer sessions were supplanting top-down lecture-style instruction. In concert with its “genius bar,” Summit was using a Flex blended-learning model, in which students access core academic content online and engage with teachers and peers offline for additional projects, supports, and lessons.

Summit’s story offers a clear example of how one school system explored using a Flex blended-learning model to unlock personalized teaching offline. But what it doesn’t tell us is what was happening online—that is, what the software was doing right or wrong. Although Summit itself is incredibly conscientious about testing what software works for its students, I worry that the field has extrapolated a far too simple vision of Diego’s story: Blending your school or classroom will maximize small-group instruction. End of story.

The shortfall of framing blended learning in terms of its offline upside is that we gloss over a key question on the demand-side of the edtech market: what do we want software to accomplish within these blended models? If schools focus too intently on choreographing their blended-learning model to maximize for small-group instruction, then they risk failing to discover what software works for what purposes.

That may stand to change as schools and teachers begin to articulate the particular use cases in which they elect to implement technology and the particular jobs they are hoping to accomplish using online-learning programs. We highlight one such example in our latest publication, “Schools and software: What’s now and what’s next.” In that study, we discuss findings from our interviews with 30 small- to medium-sized school systems about how they are using technology across the enterprise. At one charter management network, Rocketship Education, educators have started to sort online-learning programs not according to brand name, features, or functions, but rather according to the circumstances in which the programs work best. Rocketship has identified a range of use cases where it hopes technology can support instruction: independent work, whole-group instruction, remediation, collaboration opportunities, homework, corrective instruction, and progress monitoring. Educators then sort information about online-learning programs according to these distinct categories, depending on when programs excel or fall short. For example, some programs are great for remediation but require too much supervision for homework.

These use-case analyses reflect demands that other researchers have also heard coming from the field. For example, a recent study by the Bill & Melinda Gates Foundation, “Teachers Know Best,” investigated what teachers want out of digital-learning tools. The study identified a range of both teacher- and student-centered purposes, such as delivering instruction directly to students, diagnosing student-learning needs, or varying the delivery method of instruction, to name a few.

These findings may point to a new manner of sorting the academic software market. The contours of case-based demand call for a different type of analysis that would refocus the field on personalization, rather than one-size-fits-none products that babysit students while their peers rotate into offline small-group stations or projects with teachers. Identifying the particular demands we want software to meet would also shift the popular question of “what works?” to “what works, for which students, in what circumstances?” These “circumstances” go beyond blended-learning models—such as the Station Rotation or Flipped Classroom approaches—to the particular goals that educators are hoping to accomplish through online modalities within an array of blended-learning models.

Focusing on what we want out of the online modalities in blended classrooms and schools requires a shift among educators, vendors, and researchers alike. School systems should start or continue to articulate and describe these use cases in a clear manner so that developers could respond to these discrete needs with better-targeted products. Developers would also be wise to spend time observing and getting to know these use cases, so as to build software for these more specific applications, rather than a one-size-fits-none approach. Finally, researchers can further this agenda by pushing the conclusions of randomized controlled studies even further to dig in on particular successes and failures of software programs hidden behind average performance statistics.