My first exposure to blended learning was a transformative moment in my life.

At the time, I was in my second year of teaching middle-school math at a K-8 school in Kansas City. There were many aspects of teaching that I absolutely loved, but there were a few characteristics of my job that caused me daily frustration. One such frustration was that I felt like I never had good enough data on my students’ content mastery. How could I address my students’ Swiss cheese understanding of math if I didn’t know where the holes were in the cheese? The second was my simultaneous love and hatred for the idea of differentiation. In theory, differentiating instruction to meet my students varied needs seemed like exactly the right thing to do. But implementing more than token efforts of differentiation always felt like an impossibility. These two roadblocks made me feel like I was never giving my students the education they needed and deserved. Instead, I would plod through the course pacing guide and helplessly observe many of the bored or confused expressions I saw daily on my students’ faces.

One day while driving home from a long day of teaching, I cued up an episode of Freakonomics Radio on my iPod. The story I heard that day was about an experimental new learning model in New York City called the School of One (now called Teach to One) that was being used to teach middle school math. Stephen Dubner, the Freakonomics co-author and podcast host, described this new model as the Pandora Internet Radio of education. At School of One, students received customized “playlists” of activities each day that addressed their individual learning needs and preferences. At the end of each day, the students would be assessed to determine whether they had mastered that day’s learning objectives. The assessment data would then be fed into a computer system that would use algorithms to generate the students’ customized instructional playlists for the following day. The idea sounded so amazing that I burst out cheering and screaming in the solitude of my car. If you had been driving next to me that afternoon you might have thought my college football team had just completed a tie-breaking touchdown at the end of the 4th quarter. That podcast was the spark that inspired me to get into the work I am doing today.

In the days and weeks that followed, however, a nagging concern started to settle in. Teaching a core tested subject like middle school math in the challenging environment of urban public schools is a high-stakes game. I worried that School of One’s new model was so sophisticated and so radically different that it might not be able to “get it right” soon enough to maintain its traction. If its model wasn’t quickly able to outperform traditional classroom instruction, I feared that it would in short order be shoveled off to the ashtray of history as a failed idea.

Historically, new innovations have the best chance for success if they deliberately decide not to start off in the big league with the most demanding applications and customers. Instead, they get traction by initially serving the least demanding customers or by providing access to nonconsumers who are outside of the traditional market altogether. For these “low-end” and “nonconsuming” customers, the feeble performance of the new technology is still “good enough” to meet their needs. Consider, for example, the story of the transistor.

Soon after the transistor emerged from Bell Laboratories in 1957, Radio Corporation of America (RCA) licensed the technology and started doing research and development. They could see that it was going to be the next big thing, so they spent millions trying to make it good enough for the high-end electronics they were selling, such as floor standing televisions and tabletop radios. Unfortunately for RCA, its R&D investment never paid a return. The company just couldn’t figure out how to get the transistor to handle the power levels required by its devices. Meanwhile, the transistor got its start in small, low-power devices such as hearing aids and the staticy-sounding pocket radios sold by Sony starting in the 1950s. The customers for these low-power devices were more than satisfied with their low-quality performance because their alternative was to have nothing at all. You just couldn’t put a vacuum-tube-powered amplifier in your ear, and you couldn’t carry a hi-fidelity table-top radio in your pocket. What followed in the years ahead is a story that is typical of practically every new innovation. Once Sony found a foothold for the transistor in small, cheap, low-power devices, it was able to improve the technology in the decades and years ahead until it was eventually making the TVs and radios that RCA had wanted to sell.

Last week, my colleague Michael Horn wrote about a new study out of Columbia Teachers College that analyzes the student performance results of Teach to One during the most recent school year. The results are mixed, with Teach to One students outperforming their traditional-school peers on average, but with some student subgroups and some school implementations showing less-than-stellar results. When we consider how new the Teach to One model is, however, I think the results are quite commendable. Blended learning, like all new innovations, needs time for iterative improvement. The fact that Teach to One is monitoring its performance and working to improve shows that it is on the right track.

Many of us, like RCA, can see the potential that new technologies have to offer. We need to be careful, however, that we do not make the same mistake as RCA by expecting those new technologies to immediately be ready for high-quality performance in our most demanding applications. If we have the patience to let innovation take its course, it will be exciting to see the results that blended-learning models like Teach to One are able to achieve in the years and decades ahead.


  • Thomas Arnett
    Thomas Arnett

    Thomas Arnett is a senior research fellow for the Clayton Christensen Institute. His work focuses on using the Theory of Disruptive Innovation to study innovative instructional models and their potential to scale student-centered learning in K–12 education. He also studies demand for innovative resources and practices across the K–12 education system using the Jobs to Be Done Theory.