Incumbent disruption
A Socratic walk-through of incumbent disruption — reasoned out one step at a time, not lectured.
The question we started with
THE QUESTION #Why do dominant firms so often lose to a technology their own engineers saw coming first?
Kodak's Steven Sasson built a working digital camera in 1975 — a toaster-sized thing recording a grainy image to cassette tape over about twenty-three seconds. The firm that became the emblem of failing to see digital coming had it in the building first, with a patent. The pattern repeats: the technology that unseats a dominant firm is very often demonstrated inside that firm, years early, by people who understood exactly what it would become.
So blindness is the wrong explanation. What does a large, well-run, well-informed firm do with an accurate forecast that a small one does not?
Reasoning it through
REASONING #Start with what happens to a proposal inside a big firm. It does not meet a judgment about the future; it meets a comparison. Every project competes for money, factory time and — most scarcely — the sales channel, scored against the other things the firm could do instead.
Now ask what a disruptive technology looks like at that moment. Early on it is worse on the dimensions existing customers care about: the 1975 digital image was hopeless beside film. It serves a small market, often at the low end. And it carries lower gross margins, because it is cheaper and simpler.
Run that through the firm's own filters. Customer research says the best accounts do not want it. The margin hurdle rejects it. The size-of-prize test rejects it, and here the arithmetic is unforgiving: a firm with fifty billion in revenue seeking five per cent growth needs an opportunity worth two and a half billion, where a firm with fifty million needs two and a half. The same emerging market is beneath notice for one and existential for the other. Nobody has to be foolish for the proposal to die; every filter works as designed.
Then there is the piece that is genuinely arithmetic rather than culture. If the new business succeeds, part of its revenue comes out of the old business's hide. To the incumbent the project is worth its gains net of what it destroys; to an entrant with nothing to destroy it is worth the gains gross. The same project therefore has two different and both-correct valuations. This is the load-bearing claim of the whole account, and it is why exhortations to be bolder rarely change anything: the incumbent's reluctance is a correct calculation on a different set of books.
Does that hold up if we test it in the other direction? It predicts something specific: incumbents should win when a new technology is sustaining — better on the dimensions customers already value, at equal or higher margins — however radical it is technically. And broadly they do; established firms have absorbed enormous technical upheavals in their own products without losing their position. It is the cheaper, worse, lower-margin entrant that catches them.
But push once more and the account narrows. If the problem were purely allocation, spinning the venture out with its own profit line and customers should solve it. Sometimes it does; often it does not, because the unit still draws capital from the parent and is still measured against the parent's growth expectations. So the invariant is not "big firms are slow" but something tighter: an incumbent loses wherever the new business must be judged against the old business's customers, margins and size — and that boundary follows the accounting, not the org chart.
The analogy
THE ANALOGY #Think of a farmer with a mature, bearing orchard, offered a new crop that yields less per acre today and improves quickly. To plant it he must uproot trees that are earning; the newcomer with bare land plants it against nothing at all. Both are looking at the same crop and reaching opposite, defensible answers.
the farmer makes one visible decision he could later regret, whereas a firm's refusal is spread across hundreds of small, reasonable choices by people each doing their job properly — no one rejects the technology, the budget process does — and the farmer's binding constraint is land, while the firm's is usually the sales channel and senior attention rather than capital.
Clarifying the model
THE MODEL #Two corrections, one of them to the story this piece opened with.
Kodak is a bad illustration of blindness, and the popular version is simply wrong. Kodak invested heavily in digital imaging and was among the leading sellers of digital cameras in the mid-2000s. What destroyed it was not missing the technology but that the profit pool it was moving into was far smaller than the one it was leaving — film had extraordinary margins and a consumable revenue stream, digital cameras had neither, and then the camera dissolved into the phone. Seeing correctly and acting decisively would have produced a much smaller Kodak, not a saved one. Sometimes the honest answer to "why did they not respond?" is that no response was worth the cost.
The second correction is about the theory itself. Christensen's disruption account is contested. Historians and empirical researchers have challenged whether the original case studies — disk drives, mechanical excavators, steel minimills — support the general claim as firmly as presented, and the theory is criticised for identifying disruptors convincingly only after the fact, which makes it hard to falsify in advance. What survives that criticism reasonably well is the narrow mechanism above — asymmetric valuation and internal resource allocation — rather than the broad prophecy that low-end entrants generally win.
A picture of it
THE PICTURE #How to readThis repurposes the journey form, which normally charts one person's satisfaction through a process: here each row is the same proposal at one moment, scored out of five by the internal constituency that must back it, read top to bottom in time. The Engineers' scores stay high throughout — the forecast was never the problem. Follow Sales and Finance through the middle three sections and you can watch the proposal die without anyone deciding to kill it. The last section is the shape of the loss: a good product, correctly built, arriving after the customers it needed have gone.
What became clearer
WHAT CLEARED #Dominant firms lose to technologies they saw first because seeing is not the scarce thing — allocation is. A disruptive product arrives worse on the axes existing customers value, smaller than the incumbent's growth needs, and thinner in margin, so every filter a well-run firm uses scores it down correctly. Worse, the project is genuinely worth less to the incumbent, because part of its revenue is cannibalised from the incumbent's own book while an entrant destroys nothing. The failure is not vision or courage but arithmetic done on a different balance sheet — and where that arithmetic is honest, as at late Kodak, no amount of foresight makes the answer come out differently.
Where to go next
ONWARD #- Why corporate venture units and spin-outs succeed in some cases and are quietly reabsorbed in others.
- How to tell a sustaining technology from a disruptive one before the outcome is known, if that is possible at all.
Key terms
TERMS #| Term | What it means |
|---|---|
| Sustaining innovation | an improvement along dimensions existing customers already value; historically won by incumbents. |
| Disruptive innovation | an offering that starts cheaper and worse on the mainstream dimensions, then improves until it is good enough. |
| Cannibalisation | new revenue that displaces a firm's own existing revenue, reducing the project's net value to that firm alone. |
| Resource allocation process | the routine by which a firm ranks projects, using its current customers, margins and scale as the yardstick. |
| Profit pool | the total profit available in an activity, which can be far smaller after a technological shift than before it. |
Every term the collection defines is gathered in the glossary.