Why good management makes great companies fail

1

Good managers still fail

Companies fail not from lazy or arrogant managers but because rational, customer-focused decisions systematically blind them to the innovations that later destroy them.

Christensen opens with disk drives as a laboratory: product generations turned over every two years, so leadership turnover could be watched in fast motion. Firms that dominated one size class lost the next transition almost every time even though they held more capital, engineering talent, and market data than the entrants who beat them.

The incumbents were often first to prototype the new drive. They failed at commercialization, not invention. Standard stories of bureaucratic decline or engineers who cannot innovate do not fit the evidence.

He treats failure as a predictable output of good management aimed at the wrong situation. Listening to best customers, funding highest-margin work, and chasing largest markets—the pillars of sound strategy—can lead straight into collapse when the innovation at hand is the kind current customers reject. Technology here means any process that turns inputs into outputs, including mundane organizational changes as well as engineering breakthroughs.

In practice: Seagate and Quantum repeatedly lost the next drive-size transition despite superior resources; they often built working prototypes yet could not commercialize them.

2

Sustaining versus disruptive

What matters for survival is not radical versus incremental technology but whether an innovation sustains the trajectory current customers value or disrupts it with a different, initially worse proposition.

Sustaining innovations improve performance along metrics mainstream customers already care about, whether by small steps or big leaps. Incumbents almost always win these because existing customers and profit structures reward the investment.

Disruptive innovations introduce a new value proposition—typically simpler, cheaper, smaller, or more convenient—that underperforms on the traditional metric yet wins on a different one such as price or ease of use. Disruption is not about technical simplicity; some disruptive drives required serious engineering. It is about the mismatch between what the offering gives and what the current customer base values when it appears.

This distinction predicts which projects incumbents will champion and which they will starve. Once the disruptive product improves enough to meet the old market's needs, it invades from below and the incumbent has no answer, because its cost structure and organization were built for the prior format.

In practice: Each shrink from 14-inch to 8-inch to 5.25-inch to 3.5-inch drives started slower and lower-capacity than mainframe customers wanted, so incumbents under-invested while the smaller drives found homes in minicomputers and desktop PCs.

3

Value networks decide attractiveness

A firm's profitability formula and its place inside a chain of suppliers and customers determine which innovations look rational, regardless of their ultimate importance.

Christensen calls this context a value network: the suppliers, distribution channels, and customers within which a firm competes and makes money. Each network carries its own definition of performance and profit. A component that is a clear improvement in one network can be worthless or harmful in another.

That is why competent managers rationally reject disruptive technologies. Inside the mainframe network a smaller, lower-capacity drive genuinely was a worse product; heavy investment would have ignored real customer feedback and real financial data.

The same managers who missed a disruption in one generation often executed a nearly identical move well when it aligned with their existing network, proving the problem was structural rather than personal. A company never evaluates technology in the abstract; it evaluates it through the profit logic of the network it already serves.

In practice: Bucyrus-Erie and other cable-and-bucket excavator makers dismissed early hydraulic machines as toys because they could not yet handle the large buckets commercial customers valued; hydraulics grew instead in small trenching, then later took the high end.

4

Funding starves disruption quietly

Disruptive projects lose the internal money fight not by executive veto but because middle managers answering to customers and targets quietly starve them before top management ever sees them.

Strategy is what actually gets funded, not what leaders write in planning documents. Middle managers build careers by championing projects that serve large known customers with clear near-term revenue, because those proposals survive finance scrutiny.

A disruptive opportunity usually promises a small uncertain market, lower margins, and unproven demand—the exact profile that loses against a sustaining project aimed at the firm's best accounts. This happens even when the CEO wants disruption, because day-to-day allocation sits with people whose incentives point elsewhere.

Resource allocation is a bottom-up process shaped by the company's values—its implicit rules about which opportunity sizes and which customers deserve attention. A large established firm calibrates those values to need big revenue numbers, making it constitutionally allergic to small emerging markets even when they will later become huge.

In practice: Engineers at several drive makers built working prototypes of the next disruptive generation on their own initiative, only to watch the projects die for lack of internal sponsorship while entrants succeeded with the same technology.

5

Small markets cannot feed giants

A company's own size and growth targets, not merely its culture, make early entry into a disruptive market nearly impossible because those markets are too small to matter on a large income statement.

Even a perfectly rational executive at a large firm faces arithmetic. A multi-billion-dollar company that must grow at a high percentage rate needs hundreds of millions in new revenue; a fledgling market measured in tens of millions cannot supply it, no matter how bright the long-term path.

Large firms therefore wait until the disruptive market is big enough to be interesting. By then an entrant with lower revenue hurdles has already captured customer relationships, the learning curve, and brand position.

Conventional market research makes the problem worse. Markets for disruptive products often cannot be sized in advance because the eventual customers do not yet exist or do not know they want the product. Tools built for sustaining innovations aimed at known buyers systematically undervalue disruptive chances and reinforce the decision to pass. More rigorous financial analysis, exactly what good training teaches, makes pursuit less likely, not more.

In practice: A $4 billion company needing 20 percent growth must find roughly $800 million in new revenue; a market projected to reach $50 million in five years cannot meet that need, while the same $50 million is an exciting prize for an entrant or spinoff.

6

Markets must be discovered

Because nobody knows in advance who will want a disruptive product or how they will use it, early strategy has to be a series of testable guesses rather than a fixed plan built on market research.

Standard planning tools—forecasts, net-present-value calculations, competitive analysis—assume the relevant facts are knowable up front. For sustaining work aimed at existing customers that assumption mostly holds. For disruptive work it fails; the eventual market and use case are usually different from the one the team first imagined.

Christensen argues for discovery-driven planning: launch cheaply, expect the original plan to be wrong, treat the first product as a probe for learning rather than a bet that must win, and reallocate quickly once real customers show what they value.

This demands a different stance toward failure. A sustaining project that misses is usually bad execution; an early disruptive product that misses its intended market is often just the first iteration finding its real one. The practical rule is patience for growth but impatience for profit, so the effort is not killed before it can iterate.

In practice: Honda entered the early-1960s American motorcycle market aiming at highway riders with larger bikes and struggled; it succeeded almost by accident when small 50cc bikes used by employees for errands drew ordinary buyers who wanted light around-town transport—a market Honda had not planned and at first resisted.

7

Oversupply opens the low end

Companies keep loading products with more performance than mainstream customers can use because sustaining improvement is what the organization rewards, and that overshoot is exactly what lets a disruptive rival enter below them.

Christensen returns to a simple chart: customers' ability to absorb more performance rises along a gentle slope, while engineers chasing the next sustaining win for their most demanding buyers improve the product on a steeper slope. Eventually the offering exceeds what average or low-end customers need, even though top customers still want more.

That gap—performance oversupply—is where disruptive products win. A buyer already getting more than enough has little reason to pay a premium for still more and becomes open to a cheaper, simpler alternative that meets the actual requirement.

The same mechanism repeats across industries. Incumbents keep fleeing upmarket toward their highest-margin customers, ceding each low-end tier until the disruptor has climbed the entire ladder at lower cost.

In practice: Nucor and other minimills entered at the lowest-margin product, reinforcing bar, which integrated steelmakers were glad to abandon; the minimills then moved into structural steel and finally sheet steel while the integrated mills kept retreating upmarket.

8

Separate the disruptive unit

A company that wants to commercialize disruption should not house it inside the main organization, because mainstream processes, values, and cost structure will smother the project regardless of executive intent.

Given that resource allocation and organizational values reliably kill disruptive work, the central remedy is structural separation. Spin the effort into an autonomous organization sized so that a small early market is genuinely meaningful to it, with its own sales force, cost structure, and profit expectations tuned to the emerging market rather than the parent's existing business.

Simply creating a new product team inside the old division usually fails. The project still competes for the same engineers, the same sales channel, and the same finance approval process—all calibrated to larger numbers and existing customers.

Acquisition can work as an alternative, but only if the acquirer leaves the acquired unit's processes and cost structure intact. Immediate integration typically imposes exactly the overhead and expectations that killed internal attempts. You cannot out-argue an organization's values with a memo; you have to change the context so the disruptive project's small scale and different profit profile look normal.

In practice: Control Data Corporation's independent Oklahoma City division, created specifically to pursue 8-inch drives away from mainframe customers and cost assumptions, succeeded where in-house attempts by other firms failed.

9

Match strategy to the battle

Executives must first diagnose whether a threat is sustaining or disruptive, because the correct response—fight hard as incumbent versus spin out and attack as insurgent—is opposite in each case.

When the innovation is sustaining, incumbents should invest aggressively and compete head-on. Their scale, customer relationships, and technical resources give them a real advantage that new entrants rarely overcome; nearly every sustaining transition in the disk-drive evidence was won by the incumbent leader.

When the innovation is disruptive, the same head-on posture fails. The incumbent is optimized for a value network the disruption does not serve, and its resource process will never fund the fight adequately. The right move is a separate organization matched in size and cost to the emerging market, free to develop its own customers instead of forcing the new product onto the existing base.

The frame can be used forward, not only backward, though it cannot dictate exact timing or final form—only that adoption is likely to begin in an overlooked niche rather than by direct displacement of the mainstream product.

In practice: In the mid-1990s Christensen argued electric cars would likely fail as a sustaining substitute for gasoline cars sold to mainstream drivers, yet might succeed as a disruptive product for urban commuters or fleet operators who valued low running cost and quiet operation over range.