In Education, How Do We Create a Culture of Experimentation?

This conversation first appeared on Bright, Medium’s publication about innovation in education.

Tom Vander Ark (TV): Ten years ago, there was little innovation in education — limited R&D, almost no venture capital, and foundations were funding traditional ideas. Every other sector was transforming itself with technology, but education largely looked like it did a hundred years ago. That was the problem we were trying to address when we launched Getting Smart and Learn Capital in 2008.

The good news is that we’re seeing progress on many fronts. From almost no venture capital seven years ago, we saw $1.6 billion invested in the first half of this year. Most national education foundations have developed an innovation agenda. While it hasn’t been as fast as hoped, I see progress in learning platforms and the explosion of open, maker, and project-based resources. There are a couple of teams innovating on learning environments and tools simultaneously.

Matt Candler (MC): Tom, you were way ahead of me when it comes to innovation in schooling. I spent much of the last five years trying to unlearn a fairly conservative approach to innovation — one focused on replicating schools steeped in decades-old design. I’ve grown more optimistic about where the future might take us, but to be honest, much of what I’ve learned about experimenting in education (both about process and the ideas that can push us forward) comes from outside the sector.

What I think is missing is a disciplined approach to early-stage investing in ideas — and the leaders who have them.

Most investors, whether they’re profit-seeking entities or foundations interested in social impact, act in one of two ways: (1) They assume someone else is doing the work to create new ideas; or (2) They assume new ideas just come from the ether, and that philanthropy can’t actually impact the number, quality, or diversity of ideas. Their approach is mostly passive.

The idea of philanthropy as smart risk capital doesn’t really fly with most foundations. That’s a problem. It means everyone’s waiting for someone else to take the risks on new ideas in education.

My question is this: How can we make as many smart $10, $100, $1,000 and $10,000 bets on promising ideas and promising entrepreneurs as we can? How can investors help push the idea forward when it is tiny, when the leader isn’t sure yet if it will work? What do people need at this stage?

Here’s an example of the path walked by an entrepreneur named Eric Nelson, who we coached at 4.0 Schools. Fantasy Geopolitics started as a simple experiment: Eric wondered if he could get his students more excited about current events by using the same ideas behind his favorite hobby, Fantasy Football. He built a spreadsheet and came up with a game — get kids to pick world leaders and keep score based on how many times they were showing up in a daily news search. It was like Fantasy Football for Social Studies. His students loved it. So he went to an event called Startup Weekend Education. Investors and teachers loved it, told him to keep going. So he did. We then invited him to a 3-day workshop we call Essentials, so he could create a more extensive test. That test worked too. He learned a bunch and kept tweaking. He was ready for a bigger test, which he developed and ran in our three month Launch program. He stayed in the classroom for a year, testing, growing it, talking to partners like Rand McNally. This year, he upped his own personal investment and left the classroom to build his program out.

This is how a lot of education experiments can happen, and right now, it’s where most investors and foundations don’t want to be. How can we provide support, encouragement, and training at these early stages — when ideas are being formed, and perhaps most importantly, when they can be tested at a small scale without risking hundreds of kids’ academic records?

TV: There are still many barriers to innovation. In addition to a lack of investment, there hasn’t been a space to prototype new ideas — it’s really hard to break into the master schedule or open a “school-within-a-school.”

Procurement laws have made it difficult for school districts to work in an iterative way with EdTech partners (say, to test new cloud software with biology teachers). And most school districts don’t have physical places to try new approaches and tools. Many state charter school laws outlawed innovation by requiring proof-of-concept. Innovation is also difficult for districts because it creates temporary inequities inside classrooms — some students may be part of an experiment, and others not — making it really important for districts to communicate goals, equity commitments, planned phases.

When they launched the personalized math program School of One (nowTeach to One) in New York, Joel Klein and Joel Rose started in summer school, moved to after-school, then a pilot school —from no stakes, to low stakes, to limited stakes. It was a smart entry point, but it took a long time to transfer the experiment to scale. We need more sheltered entry points like this.

The charter sector was initiated to support innovation, but soon began to prize what worked in the past. Authorizing new ideas got tougher, took longer, and became more expensive. Larger networks with a proven track record (like Summit, Aspire, and KIPP) are innovating, but it’s hard to open an innovative new school.

There are some bright spots, though. Houston A+, an education nonprofit, launched a small tuition-free private school to pilot A+ UP, a middle school program that will expand as a three-campus charter network this fall.

MC: For the most part, the promise of charters as a place where experimentation can happen has fallen short. For the last decade or so, we have valued “scaling” over new model development. I think this is mostly because we have not thought of a creative and responsible way to test new school designs without risking the academic careers of hundreds of kids at a time.

This is where The Tiny School Project might be worth discussing.

Let’s take a look at how food entrepreneurs work today. Back in the old days, it took $1 million to open a new restaurant. These days, my friends in New Orleans who open restaurants follow a very different path. It starts with a pop-up — renting out for maybe $1,000 an existing restaurant to test your menu in one evening. For many, the next step is to buy a food truck for about $10,000, to test your food with more customers in more neighborhoods. In some cities, a food incubator takes the process one step further; aspiring entrepreneurs can set up shop for a month or two in a facility acknowledged to be a place where innovation happens, and where failure is more accepted.

This approach to new restaurant development — I believe — encourages more people to try new ideas while actually lowering the cost and the pain of failure, for both the entrepreneur and eater.

How might we apply this concept to schooling? That’s exactly what my organization seeks to answer.

Our testing of new ideas happens in stages — a $10 test of an idea impacts only a handful of people; a $100 test impacts the same people, but for a few hours and not minutes; a $1,000 test of an idea impacts a willing group of volunteers for a week or two. In New Orleans and in New York, where 4.0 has dedicated innovation labs, we’ve built a network of teachers, students, and family members who have volunteered to test new ideas and education at an early stage. This gives us an opportunity to make sure ideas are ready for prime time. I believe this kind of space in fact changes the nature of experimentation in education.

For too long we have experimented on instead of with students, families, and teachers. We’re still fine-tuning our process, and there’s no doubt that when an idea fails, it can be disappointing. But I think there’s a fundamental difference between realizing an idea didn’t work in a pop-up version, and realizing an idea didn’t work after spending millions of dollars and asking hundreds of people to risk their academic and professional careers on the experiment.

TV: Let’s talk more about the microschool opportunity. It’s getting easier to open a really cool school either as a small private school or an academy within a larger public school. I wonder if this could be part of the answer.

MC: I’m very excited about this. Microschools — typically, schools with fewer than 150 students — are smaller, more responsive, cheaper to run, easier to start. And they have the potential to teach us a great deal about running lean.

Some microschools like AltSchool (the San Francisco school that’s raised $100 million from Mark Zuckerberg, among other investments) are doing great things. But AltSchool isn’t teaching us how to lower costs of delivering education, at least not yet. Despite the hefty investments, tuition costs over $20,000.

I’d argue that schools like Talent Unbound in Houston, NOLA micro schools here in New Orleans, and Streetlight Schools in Johannesburg have even bigger visions than AltSchool. They want to change the way we do school by putting students and their ownership of learning at the center of everything, while lowering costs in order to make these types of schools accessible to more students.

TV: I’d also like to reflect on your earlier point about experimenting with versus on students. Some observers might have a visceral reaction to the phrase “experimenting on kids,” thinking immediately of the risks. It may also conjure images of frightening Victorian-era laboratories. But two-thirds of American kids don’t get what they need and deserve from our schools. We obviously don’t want to create conditions where kids do worse, but there’s a lot of room to improve.

In my experience, there are four safeguards we can put in place to minimize the quantity and scale of failed experiments: clarity on goals of the pilot; opt-in for parents and kids; short feedback cycles that monitor progress; and privacy protection.

LEAP Innovations has created a Chicago network of schools piloting innovations. Next Generation Learning Challenges and NewSchools’ Catapult support innovative new schools. But these are rely on formal educational environments. A national network of “tiny school” trials like you describe would promote the innovation in learning environments and tools we desperately need.

The key to mitigating risk is small experiments and short cycle trials. Digital tools and continuous feedback make it easy to support iterative development. This is the secret sauce behind new models like AltSchool and Summit Public Schools.

MC: I think it’s also important to talk about the issue of race and class when we talk about experimentation. In many neighborhoods, new ideas have been shoved down the throat of generation after generation with very little tangible evidence of breaking the cycle of poverty or leveling the field, especially for children of color. Many of the “new ideas” we’ve thrown at communities have promised equity but too often led to greater segregation and class division.

Your focus on opt-in innovation, together with my call for small-scale experimentation, are attempts to acknowledge that for too long, the education system has been experimenting at the wrong scale, in the wrong ways.

The future of school, and the future of school reform, means experimenting with, and not on, each other.

For more check out:

• Open a Micro-School: Here’s How
• The Micro-School Opportunity