Disruptive Innovation Explained
Disruptive innovation in one page
- Source: Clayton Christensen, The Innovator’s Dilemma (1997), Disrupting Class (2008).
- Sustaining innovation: A better, more expensive version of an existing product, sold to existing high-end customers.
- Disruptive innovation: A simpler, cheaper, “good enough” product, sold to customers the existing market ignored. Over time, it improves until it serves the mainstream and the incumbent collapses.
- Pattern:
- New entrant launches a cheaper, simpler version.
- Existing leaders dismiss it as low-quality and serve their best customers.
- The new entrant improves and moves upmarket.
- The existing leaders lose, often quickly, once the new product is “good enough” for the mainstream.
- Why education is ripe: The mainstream brick-and-mortar school is expensive and slow. Online and hybrid models started low-quality, served students who could not access the mainstream, and have been improving fast.
Clayton Christensen’s The Innovator’s Dilemma described a pattern that had played out across dozens of industries in the twentieth century. A new entrant launches a product that is cheaper, simpler, and lower-quality than what the existing market sells. The existing leaders dismiss it as inferior and stay focused on their best customers. The new entrant gets better, slowly at first, then quickly, until the existing leaders find their core market has moved to the new product. By then it is too late to catch up.
He called this pattern disruptive innovation. In Disrupting Class, written with Michael Horn and Curtis Johnson, he argued that the same pattern was about to play out in education.
The pattern, step by step
Disruption is not the same as competition. Two restaurants on the same street competing for the same customers are not disruptors. Disruption follows a specific four-step pattern.
Step 1: Cheaper, simpler entry. A new entrant offers a product that does less than the existing product but at a much lower price or with much less friction. The first customers are people the existing market did not serve well, often because they could not afford the better product or did not have access to it.
Step 2: Dismissed by incumbents. The existing leaders look at the new product, note that it is lower quality on the dimensions their best customers care about, and conclude it is not a threat. Their best customers agree. So far this is correct.
Step 3: Improvement and movement upmarket. The new entrant keeps improving. Each year it gets a little better. At first it still only serves the customers the incumbents ignored. Over time it gets good enough to serve customers a tier higher.
Step 4: Collapse of the incumbent. At some point the new product becomes good enough for the mainstream. Demand for the existing product falls fast. The incumbents, locked into expensive operations built around the old product, cannot adapt fast enough. They lose share, lose money, and often close.
The trap is that step 2 looks correct at the time. The incumbents are not stupid. Their analysis is right, given what they see. They are listening to their best customers, who do not want the cheaper product. By the time the new product threatens those customers too, the incumbents have lost their lead.
Cheaper entry, dismissal by incumbents, improvement and upmarket movement, collapse of the incumbent.
The trap for incumbents is that step 2 looks correct at the time. The new product really is lower-quality. The mistake is assuming it will stay that way.
Three worked examples
The pattern is easier to see in finished cases.
Radio: Sony vs RCA. In the 1950s, RCA and Zenith made large vacuum-tube home radios, sold through showrooms, with most of their profit coming from the regular replacement of tubes. Sony introduced a small transistor radio that was tinnier, lower-fidelity, and clearly inferior for the home market. Sony first approached the incumbents about distribution; the incumbents declined because the new product had no replacement-part revenue and looked cheap. Sony sold instead through department stores and general retailers, to teenagers who could not afford a home radio. Transistor radios improved quickly. By the end of the 1960s, the home-radio market the incumbents owned had largely collapsed, and the new portable form dominated.
Photography: digital vs film. Eastman Kodak invented digital photography in 1975 and chose not to commercialise it aggressively. Their core profit came from film, prints, and chemicals. Early digital cameras produced grainy images at low resolution; serious photographers laughed at them. Smaller manufacturers improved digital cameras year by year. By the early 2000s, digital had become good enough for the mainstream. Kodak filed for bankruptcy in 2012, holding patents to the technology that disrupted it.
Computing: PCs vs mainframes. In the 1970s, IBM dominated mainframe sales to large corporations, while Digital Equipment and Data General were strong in minicomputers used by departments and institutions. The personal computer started as a hobbyist product, vastly less powerful than either, and was dismissed by IBM as a toy. Apple, then IBM’s own PC division almost as an afterthought, then Compaq and Dell, kept improving the PC. By the 1990s, the PC had eaten the workstation market and was eating into the mainframe’s customers. Digital Equipment, once the second-largest computer company in the world, was sold to Compaq in 1998.
In each case, the same shape. The new product looks worse on the metrics the incumbent measures. The new product gets better. The incumbent looks correct, then loses.
Why education fits the pattern
Disrupting Class argued that K-12 schools and universities show the conditions for disruption.
The incumbent product is expensive. A school year costs thousands of dollars per student in fixed costs that are not directly tied to learning. The incumbents serve the customers who can afford this, and they serve them moderately well.
The customers the incumbents do not serve are large. Students who cannot reach a physical school for any reason, students whose schools cannot afford the staff for advanced subjects, students who need remedial work the schedule does not allow, students in regions where there are not enough schools.
The new entrants started small and looked low-quality. Early online courses were thin and unstructured. Massive open online courses (MOOCs) started with high promise and low completion rates. Educational software was clunky.
Improvement has been steady. Better-designed online courses, adaptive learning platforms, classroom response systems, blended models that mix online and in-person work, micro-credentials and bootcamps. None of them is a complete replacement for a school yet. Each one keeps getting better.
Christensen’s argument was that disruption would spread unevenly. Online learning would gain ground first where students and schools were not well served by traditional providers, then move into mainstream school models as the tools improved. K-12 schools are slower to disrupt because the building also provides childcare and supervised time, not only learning.
The argument is not that every school will close. It is that the mix will shift, and the schools that survive will look different.
How disruption usually happens to the incumbents
A few patterns recur when incumbents try to respond.
They add the new product as a side line to their existing offering, but route it through the same expensive structure. A traditional university launches an online programme but staffs it like an in-person one and prices it the same. The economics do not improve.
They wait too long, then try to make a single big change. The cost of the building, the staff, and the long contracts makes fast change hard. By the time the change happens, the new entrants have moved further upmarket.
They dismiss the data. Internal reports show that the new product is improving fast, but the leaders look at this year’s revenue and conclude things are fine. Steve Sasson’s digital camera at Kodak in 1975 was real; Kodak’s choice to slow-walk it was the failure.
A few institutions adapt successfully. The pattern there is usually to set up a separate unit, run it with its own cost structure and its own staff, and let it cannibalise the parent organisation rather than waiting for an outside disruptor to do it.
Their existing cost structure and customer relationships make fast change hard.
Their best customers do not want the cheaper product; their staff and buildings are sized for the old model; their revenue this year is still strong. By the time the new product threatens their core, the gap is too wide to close from inside the old structure.
What this means for thinking about ICT in education
Disruption is not the same as “use more technology.” A school that adds tablets and projectors to the same classroom routine has not changed its model. The disruption question is whether the underlying form of delivery shifts: cheaper, more flexible, reaching more students, with the economics built around the new form rather than retrofitted onto the old.
Specific models like the School of One break the one-pace assumption of the traditional classroom with daily personalised playlists. Frameworks like Stratosphere argue that pedagogy, technology, and organisational change have to combine for any of them to work.
How was this article?