The Adaptive Learning Market Shakes Out

In a recent rush, more than half a dozen adaptive learning companies have been scooped up like M&Ms at a candy counter. One of the most notable is Knewton, whose assets were acquired by Wiley just weeks ago. Last year, Carnegie Learning, Acrobatiq, Knowre and Fishtree were also swallowed in acquisition fever. Meanwhile, ACT, the nonprofit college admission test company, earlier this year invested $7.5 million in Smart Sparrow.

Ed-tech guru Phil Hill, the MindWires partner who is widely followed on his Phil on Ed Tech blog, told me in a telephone interview that most sales were made not from strength, but “from positions of weakness and need.”

In an essay early this year, I sketched what adaptive learning -- which falls under the larger category of personalized learning -- is all about:

As with a pinwheel set in motion, insights from many disciplines -- artificial intelligence, cognitive science, linguistics, educational psychology and data analytics -- have come together to form a relatively new field known as learning science, propelling advances in a new personalized practice -- adaptive learning. Designed to adjust in real time to each student's prior knowledge and skill attainment, adaptive systems respond to variations in ability and diverse student backgrounds, sensitive to unique needs of each learner. Based on each student’s actions, when a student gets stuck, the system automatically suggests strategies on how to get out of it and proceed to mastery.

Recent acquisitions give long-term critics a chance not merely to smirk, but to go in for the kill. After years of predicting that adaptive was a fool’s game, with no chance of delivering on what developers promised, what better time than to kick it when it’s down? One critical observer admits to “a certain schadenfreude.”

With the discovery of fire, believers and skeptics began their interminable heated debates over innovation. For the defense, flames cook your meals, heat your homes and fire your pots. On the attack, defeatists say they could burn everything down. And so it’s been since the web first opened ed-tech gates in the 1970s, when adaptive learning -- also known in its early days as “intelligent tutoring systems” -- caught the fire of pioneers who had a hunch that computers might eventually achieve the human ability to adjust individually to each learner’s struggles. In the recent adaptive industry shakeout, some vendors are still cooking; others have been snuffed out.

While not as economically momentous, say, as consolidation in the automotive industry, recent adaptive mergers and acquisitions reveal a distinctive pattern. Following a Harvard Business Review analysis of corporate consolidation, adaptive systems now mark a stage at which top players own the lion’s share of the market. On the heels of McGraw-Hill’s proposed merger with Cengage just a few weeks ago, McGraw-Hill finds its ALEKS software, acquired just six years ago, leading the pack.

“Ultimately, a company’s long-term success depends on how well it rides the consolidation curve,” cautions Harvard Business Review. “Slower [growing] firms eventually become acquisition targets and will likely disappear.”

The current shakeout also parallels the “disillusionment phase” of the Gartner hype cycle, a popular analysis, charting ups and downs of new technologies, coined by Gartner, the big information technology consulting firm. In this treacherous cycle, “Producers of the technology shake out or fail. Investment continues only if the surviving providers improve their products to the satisfaction of early adopters.”

So what’s happening? Why is the adaptive industry house cleaning now, sweeping up some early, promising start-ups?

“Companies like Knewton and others went straight into black-box algorithms, assuming mastery of what learning data actually means and how students learn,” observed Hill in an email message. “Their customers were really venture capitalists, not academic programs with real teachers and students.”

The biggest lesson is that inventing whiz-bang software is not nearly enough. To succeed, vendors must assemble an adaptive Rubik’s cube, snapping four essentials securely in place. The central one, of course, is brilliantly crafted technology, coupled with a deep reservoir of high-quality content, integrated with shrewd assessment tools, embedded with skilled teacher training at each site -- all at scale to secure market share, sustainability, profits and plugged-in implementation at every campus. Not trivial.

While some were good at this or that, few pulled it all together. The biggest handicap for many is their thin content libraries. In his prescient insight, Bill Gates wrote in his famous 1996 essay, “Content Is King,” “Content is where I expect much of the real money will be made on the internet.”

One thing start-ups didn’t need was arrogance, especially trumpeting extravagant claims. In 2013, five years after Jose Ferreira launched Knewton, he boasted in Time that there will be only one company in the world capable of succeeding at adaptive learning. “I think it’s going to be us because we’re so far ahead now.”

None achieved significant scale, except McGraw-Hill’s ALEKS, a math and chemistry tutoring system, and its sister product, SmartBook. ALEKS reports 4.5 million unique users in K-12 and higher ed courses. Since 2010, it has generated 8.7 billion interactions. In humanities, social sciences, science and business, the company claims 11.8 billion interactions since 2009. McGraw-Hill offers more than 800 titles for adaptive users.

It turns out that adaptive systems are neither the best thing since sliced bread nor half-baked. Like much of ed tech, adaptive research results can be ambiguous, with some saying the software is marginally better than classroom instruction, while others report impressive results.

Alfred Essa, vice president of analytics and data science at Macmillan, says that “in some domains, well-designed adaptive tutors are on a par with human tutors.” A truly remarkable feat. Until recently, Essa was head of research at McGraw-Hill.

The ironic denouement is that an old-line publisher, founded at the end of the 19th century, more than 130 years ago, outwitted high-tech upstarts, countering modern corporate trends in which whiz kids leave the old guard in the digital dust. Curiously, Knewton, an adaptive falling star, is now in the hands of Wiley, another venerable publisher, founded more than 200 years ago in 1807.

Even curiouser -- while there was lots of chatter about the effect on textbooks, very little attention was given to how these deals were shaking out the adaptive marketplace.

Disclosure: The author serves on the McGraw-Hill Education Learning Science Advisory Board and is the editor of Virtual Teamwork (Wiley, 2010).