inBloom’s collapse offers lessons for innovation in education


Dec 4, 2014

By now the story is well known. inBloom, a non-profit that offered a data warehouse solution designed to help public schools embrace the promise of personalized learning by helping teachers integrate seamlessly the number of applications they use in their day-to-day teaching, collapsed and has ceased to exist, as privacy concerns from interested parties mounted over a period of many months (full disclosure: I served on the inBloom board of directors).

What’s notable about the collapse is the massive investment to the tune of over $100 million from the Gates and Carnegie Foundations that went into inBloom to create a safe and secure solution that would guard student privacy far better than the patchwork quilt of makeshift solutions that most school systems use currently. For those worried about privacy and security, that’s worth pondering.

So why did it collapse?

Some of the external reasons include the fact that inBloom entered the education conversation during an intense time of broader concerns around privacy, thanks in large part to events involving the National Security Agency, which fostered an environment of suspicion around privacy and motive; concerns and suspicions in certain corners around the Gates Foundation’s investments to improve education, as well as the fact that Wireless Generation, now owned by Rupert Murdoch’s News Corporation, helped build the product; that inBloom’s customers and partners—states and school districts—did not have the proper policies in place around data and privacy nor did they have a deep enough understanding of what inBloom was doing to explain it to their communities (and inBloom did not help them out enough in these regards, as it sought to remain “just a vendor”); and a narrative that tied inBloom to concerns around the Common Core, assessments, and the use of data in teacher evaluations.

inBloom did plenty to hurt its cause as well, including a lack of a clear business plan initially with wildly optimistic assumptions around adoption; a poor understanding of the external political climate into which it was entering; an inability to explain in clear language what inBloom was and was not and why the solution it was offering mattered—and to know when to use different narratives; poor execution, as it needed to work far more intensely to win new customers, as well as to build support for its solution with its existing customers to experience speedier implementation and proof of concept; and repeated bad PR moves early on that created confusion among a large number of constituencies in the education world—from school districts to vendors to community groups—and added to the narrative that inBloom was complicated and scary.

But just as importantly—and far more overlooked—some of the seeds of inBloom’s demise were sown much earlier and remained far more fundamental throughout its existence, an understanding of which yields important lessons for innovation in education. The biggest evidence of that flaw was in the $100 million-plus investment poured into the development of the solution.

The initial framing for inBloom, which was originally called the Shared Learning Infrastructure before the service debuted with the inBloom name, was just what its original name said: infrastructure. Those who conceptualized the solution thought of it as infrastructure—akin to the pipes that carry water to our homes—to support public schools and teachers and enable personalized learning at scale. It was envisioned as a ubiquitous solution created from the top.

With this analogy in mind, just as fundamental universal infrastructure typically requires lots of up-front investment, so, too, would inBloom. A large investment followed.

The solution would need to be reliable and bulletproof from the start and set a gold standard in the education industry for privacy and security. inBloom accordingly built up a large cost structure to realize the vision.

None of this was bad per se, but there was one big problem. inBloom would have to be sustainable on its own ultimately—the foundations would not perpetually support it—which meant that it would need to develop a business model with revenue, which meant that it would need customers. This required a bottoms-up adoption, which was fundamentally at odds with the original conception of the one beautiful and universal solution from the top. It could not simply be a free service gifted to districts or states.

With its massive cost infrastructure, that meant inBloom needed a lot of customers and fast. That ultimately pushed it even further into the limelight, which increased the stress on the solution’s performance—and further strained inBloom’s resources, as it could cut its costs only so much.

But inBloom had a big problem in its business plan. Early on the thought was that inBloom would be sold to states, but that proved untenable for a variety of reasons—and left inBloom without any significant revenue commitments in advance to tide it over as it developed its solution. So it had to market to districts, which, as anyone selling to public school districts will tell you, takes a lot of time—even without the difficult external environment and mistakes referenced above. Time that inBloom did not have because of its cost structure and resulting funding needs.

A concurrent problem complicated inBloom’s situation. Because the initial framing for inBloom had been that it would serve as infrastructure—nice and boring infrastructure—the solution had been framed as being for everyone. Early business planning contained wild assumptions about adoption.

But inBloom was a solution intended to enable personalized learning; if a district was not trying to implement personalized learning, there was a good chance that the narrative explaining the rationale for inBloom would not make as much sense to the community. Although the number of districts seeking to move to a world of personalized learning is growing, it is still quite small.

That meant that the ready-and-willing market with a pressing job to be done that required inBloom’s solution was small. inBloom always talked about the need for its service, but the bulk of its potential customers did not have the motivation to adopt it—a big reason why needs-based product development lacks the power of and is fundamentally different from a jobs-to-be-done understanding of the world.

So inBloom was stuck trying to sell a product rapidly to a small group of slow-moving customers. Needless to say it would be difficult to sustain inBloom’s cost structure.

Meanwhile, the team talked constantly about inBloom’s “disruptive” nature—but of course, it was anything but. Yes, it was intended to support the emerging disruption in education, but inBloom itself was a massive, expensive top-down, high-profile solution intended to be better than its alternatives and high quality from the outset that had been framed originally as infrastructure—the opposite of a disruptive innovation along every dimension. This misunderstanding of the word disruptive—not an uncommon misuse but an unfortunate one, as it strips away its meaning and predictive power—left the team without an obvious playbook for the road ahead.

The tragedy here is that school districts are increasingly clamoring for a solution like the one inBloom promised to provide. In our recent report, “Schools and Software,” small- and medium-size school systems talked openly about their desperate need for a data warehouse solution to act as middleware that simplified their management of the explosion in educational software. Framing the solution as a disruptive innovation—to start small with a small investment; gain organic adoption over time from the ground up; begin with a primitive solution and iterate to improve gradually and unpredictably as it grew and learned what customers did or did not in fact want; be patient for growth but impatient for evidence of success and sustainability; and avoid the limelight and pushback from an ecosystem’s incumbent organizations—as opposed to being the one-perfect solution or infrastructure with a massive up-front cost, could result (and still may very well result for another start-up entity) in a vey different picture.

But inBloom had little chance from the outset—and then kept compounding its chances to revolutionize our nation’s schools.

Michael is a co-founder and distinguished fellow at the Clayton Christensen Institute. He currently works as a principal consultant for Entangled Solutions.