Sometimes I worry that the call to personalize learning is actually code for asking leaders and teachers to do more than they’ve ever been asked to do—but without additional resources to do it.

And by resources, I mostly mean time.

This is especially true for traditional systems that may be aiming to adopt new approaches to teaching and learning but less willing to do away with legacy structures. Innovation theory shows us that in industry after industry, existing organizations often default to hybrid innovations that combine new technologies or approaches with old ways of doing business. Put differently, rather than actually making any real tradeoffs, organizations may start doing new things without stopping doing old things.

Noble as these efforts may be, school systems risk layering on more and more responsibilities and processes in the name of personalization, without new resources or more efficient processes to support those additional duties. And although dedicated teachers and staff may be able to carry that load part of the way, the “just do more” approach hardly bodes well for scaling such efforts in the future.

Although schools may manage to add more time on the margins, personalized learning at scale will likely require a massive rethinking of how schools use time, alongside pursuing new efficiencies that can save time. I was reminded of this while reading Silicon Schools’ recent report on its past five years supporting new and redesigned schools in the California Bay Area. In it, the fund’s leaders share actionable insights about the promise and pitfalls of personalized approaches. Among these takeaways, the word “time” appears a total of 39 times in the report’s 27 pages.

Specifically, the report highlights a vital reality on the ground: how schools use time is a balancing act. The report enumerates tradeoffs schools personalizing learning have had to weigh: how much time students should spend working on their own versus in groups; how much time students should be in front of screens versus offline; how much time students should work on content that is at versus above their current instructional level.

Looking beyond just Silicon Schools’ portfolio, thinking through such tradeoffs extends to the structures of entire school systems aiming to personalize instruction. It’s in making—or failing to make—these tradeoffs, however, that efforts to build a coherent personalized system risk getting stuck.

How can systems tackle time tradeoffs? In some cases this means looking for new instructional approaches that slice and dice time differently; in others this means seeking out automation and efficiencies; and in other this means wholly rethinking the structure of the school day. Here’s three levers schools could leverage to fundamentally rethink time:

Blending learning

Blended learning can help schools to reorganize how they use time in meaningful ways. By shifting some content delivery online, some blended-learning models can allow teachers to reconsider how best to spend face-to-face time in the classroom—perhaps allowing teachers to deliver more small-group or one-on-one instruction, perhaps allowing them to lead students in projects, or even allowing them to grade individual performance assessments to provide students with meaningful feedback. It’s worth noting that in some circumstances the opposite is proving true: blended approaches can cost teachers more time if technology isn’t living up to expectations or if classrooms haven’t been redesigned to absorb the efficiencies that technology could offer. In particular, some teachers opt to create blended materials themselves, which can increase costs up front. But if a school system is short on time or staff, it should design blended approaches in ways that leverage software to free up teacher time to teach and support students in more targeted ways.

Demanding tools that automate and integrate

Besides just delivering content, technology can also serve to streamline administrative tasks that often fall to teachers. For example, everything from taking attendance to rostering students can take up precious educator time. Tools like Clever have helped alleviate some of these burdens. Ultimately, however, the ability to meaningfully automate tasks will also depend on tools working in concert with one another. This requires that school systems double down on adopting technology tools that can streamline educator administrative tasks and also demand interoperability among various tools.

Mastering master scheduling

You can hardly talk about time without talking about scheduling. In traditional systems, scheduling can become the tail that wags the dog. Strict course schedules in turn risk hamstringing instructional innovations in the name of logistical coherence. To overcome this trap, schools should seek out new processes for building schedules and tools that can help sort out where students and teachers should spend their time. For example, the startup abl just launched a tool to help schools make thorny scheduling tradeoffs in a far more streamlined manner than the ‘spreadsheets and elbow grease’ models that many schools use to coordinate schedules today.

Without rethinking time, well-intentioned efforts to personalize learning could mean piling so much more onto classrooms that the efforts collapse on their own weight. Schools aiming to personalize learning will have to start negotiating tradeoffs not just regarding how students spend their time, but how the system as a whole spends its time. Put differently, to make personalized approaches feasible in the long run schools will have to pursue bold structural shifts that facilitate and make space for promising instructional shifts. How schools use—or abuse—time may sit at the fulcrum of those shifts.

Full disclosure: Our Founder, Michael Horn, sits of the board of Silicon Schools Fund.