Disruptive innovation is one of the most influential ideas in business strategy. It is also one of the easiest to misuse.
A startup threatens an incumbent and gets called disruptive. A new technology changes customer behavior and gets called disruptive. AI enters an industry, a company grows quickly, or an established product loses market share, and the same label appears.
Used this broadly, disruption means little more than something important is changing.
That is not what Clayton Christensen’s theory originally tried to explain. Disruptive innovation describes a particular competitive process in which an entrant begins somewhere an incumbent has weak economic reasons to defend, develops under different constraints, and eventually becomes capable of competing for customers the incumbent values.
That mechanism is useful because it explains something counterintuitive about successful companies: they can recognize an emerging competitor, understand what it is doing, make decisions that are perfectly rational given their current customers and economics, and still create room for that competitor to grow.
The theory becomes less reliable when we ask it to do something else.
It cannot tell us which startup will complete that journey, which emerging technology will reshape a market, or when an incumbent should sacrifice a profitable business to defend against something that remains uncertain.
Disruptive innovation theory is therefore more useful as a way of identifying conditions than predicting outcomes. It can tell us what kind of competitive trajectory might be developing, but the trajectory still has to happen.
Disruption Begins Somewhere the Incumbent Has Little Reason to Defend
Christensen’s argument is more specific than a new company beating an old one.
One version begins at the low end of an existing market. Established companies improve their products for demanding customers, often adding performance, features, service, and sophistication that support higher margins.
That is usually sensible. The best customers are asking for more, and serving them well produces attractive returns.
Eventually, however, some customers may become overserved. They are paying for performance or complexity they no longer value enough to justify the price.
That creates an opening for a different proposition.
The entrant may offer something that is objectively worse on dimensions the incumbent traditionally uses to measure quality. It can still be attractive because it is cheaper, simpler, easier to obtain, or more convenient.
A second route begins with nonconsumption rather than overserved customers. A new-market entrant makes a product or service accessible to people who previously could not afford, obtain, or conveniently use the established alternative.
The common feature is not technological inferiority. It is that the entrant begins somewhere the incumbent does not find particularly attractive.
INCUMBENT
Best customers
Higher performance
Better margins
↑
│
Mainstream market
│
↓
Low-end / nonconsumption
↑
ENTRANT
Different trade-offThis distinction immediately excludes many things commonly described as disruptive innovation.
A competitor that enters at the premium end with a better product for an incumbent’s most profitable customers may transform the industry. It is not following the same mechanism.
A technology can therefore be revolutionary without being disruptive in Christensen’s specific sense.
The Foothold Matters Only If the Entrant Can Move Beyond It
Being cheap, simple, or accessible is not enough.
An entrant can remain in a small low-end niche indefinitely. A product serving nonconsumers can discover that its initial market is too small, its economics are poor, or its limitations prevent it from satisfying more demanding customers.
The important part of disruption is therefore not merely the foothold. It is the trajectory that follows.
The entrant needs economics that allow it to survive where it starts. Its product then has to improve enough to become acceptable for customers who previously required the incumbent’s stronger performance.
At some point, what was once obviously inferior can become good enough.
That phrase matters because customers rarely require the maximum technically possible performance. They require enough performance for the job they are trying to accomplish, alongside acceptable price, convenience, reliability, and other trade-offs.
Once the entrant crosses that threshold for more customers, the competitive relationship changes.
Foothold
↓
Viable economics
↓
Performance improves
↓
Good enough for more customers
↓
Mainstream competition?The question mark belongs in the model.
Nothing about occupying a low-end or new-market foothold guarantees that the entrant will progress through the rest of the sequence. The economics can fail, technology can plateau, customer requirements can move, regulation can intervene, or the incumbent can respond effectively.
Disruption is a process whose outcome becomes clearer as evidence accumulates.
That creates the first major limitation of the theory: identifying the beginning of a plausible trajectory is much easier than knowing whether the trajectory will finish.
Why Good Incumbents Can Rationally Ignore the Threat
This uncertainty is what makes Christensen’s incumbent problem so powerful.
Imagine an established company choosing between two investments.
The first improves its flagship product for large customers who are already asking for the improvement. Those customers have substantial budgets, the company understands their requirements, and the projected revenue is attractive.
The second investment targets customers the company currently considers marginal. Revenue is small, margins are lower, demand is uncertain, and the proposed product may compete with part of the existing business.
The emerging opportunity might become strategically important in ten years.
The first project can still be the better investment today.
No executive stupidity is required.
Successful companies develop processes for allocating scarce resources toward customers and opportunities that can support the economics of the organization. Sales teams pursue meaningful accounts, business units defend margins, product teams prioritize important customers, and capital moves toward opportunities with credible returns.
Those behaviors normally strengthen the business.
Under the conditions Christensen describes, they can also make an unattractive emerging market unusually difficult for the incumbent to pursue.
An entrant with a different cost structure does not face the same problem. What looks like an unacceptably small opportunity to a billion-dollar incumbent may be an excellent starting market for a young company.
That asymmetry creates room for the entrant to learn without immediately confronting the incumbent where it is strongest.
Knowing About Disruption Does Not Remove the Economics
This is where simplified versions of the theory become misleading.
Once executives understand disruptive innovation, it seems as though they should be able to recognize the pattern and respond. The problem is that recognition does not change the numbers presented at the next budget meeting.
The emerging business can still have lower margins.
Current customers can still prefer the existing product. Building the alternative can still require engineers, capital, distribution, and management attention that could otherwise support proven revenue.
The new offer may also cannibalize the existing business before it creates comparable value.
Organizational incentives can intensify this conflict. Leaders evaluated against annual revenue and margin targets have an additional reason to avoid investments whose benefits may arrive years later.
But incentives are not the entire explanation.
Even a leadership team perfectly aligned around long-term company value still has to decide how much capital to allocate to a market that may never become important. It cannot respond aggressively to every startup, technology, and low-end competitor that might eventually matter.
This is the real dilemma.
If the incumbent waits until the entrant is obviously dangerous, the entrant may already have customers, capabilities, distribution, and improving economics. If it responds substantially to every weak signal, it can waste enormous resources defending against threats that never develop.
The theory identifies why response can be difficult.
It does not eliminate the uncertainty that makes the response difficult.
The Theory Explains Yesterday Better Than It Predicts Tomorrow
After a successful disruption, the trajectory can look remarkably clean.
An entrant began in an unattractive market. The incumbent concentrated on better customers, the entrant improved, more customers switched, and eventually the incumbent came under serious pressure.
Looking backward, every stage is visible.
Looking forward, the picture is different.
There may be ten entrants occupying apparently interesting footholds. Several technologies may be improving simultaneously, while customer preferences, regulation, capital markets, distribution, and incumbent responses are all changing around them.
Nine entrants may disappear.
The tenth becomes the case study.
That creates a survivorship problem. Once we know which company succeeded, it is easy to reconstruct the conditions that made its success understandable and much harder to remember how uncertain those conditions looked at the beginning.
Netflix demonstrates the distinction.
Its early DVD-by-mail model had characteristics that made it easier for a store-based incumbent to underestimate. The proposition differed from visiting a physical rental store, while the economics and customer experience developed along another path.
But the eventual competitive outcome depended on much more than the existence of an initial foothold. Postal logistics mattered, subscription economics mattered, broadband adoption mattered, content licensing mattered, consumer behavior changed, and Netflix itself had to move from physical delivery toward streaming.
Disruption theory helps explain why an incumbent might not initially organize itself around that emerging model.
It does not predict the entire sequence that made the model win.
That is an important distinction between explaining a competitive mechanism and forecasting a market outcome.
Not Every Market Transformation Is Disruptive Innovation
The prediction problem becomes worse when the word disruptive is applied to every major technological transition.
The iPhone is a useful counterexample because it unquestionably transformed markets. It changed expectations around mobile computing, software distribution, photography, navigation, music, and internet access.
Yet it did not enter as a cheap, limited product aimed at customers the established phone industry found economically unattractive. It was expensive, highly visible, and compelling to valuable mainstream customers.
Calling the iPhone disruptive in ordinary conversation is understandable. Calling it a textbook example of Christensen’s disruptive innovation mechanism is much harder.
That difference is not academic nitpicking.
If a superior product attacks an incumbent’s best customers, the strategic problem differs from an entrant quietly developing below the incumbent’s preferred market. If regulation changes an industry’s economics, the mechanism differs again. The same is true when a new distribution channel appears or when an entire product category becomes less relevant.
Different threats require different responses.
Stretching disruption theory until it covers all of them makes the theory sound more universal while making it less useful.
The important question is therefore not whether something caused disruption in the everyday sense.
It is whether the specific mechanism described by the theory helps explain what is happening.
AI Shows Why the Distinction Still Matters
The current enthusiasm around AI makes this especially visible.
Saying that AI will disrupt an industry tells us almost nothing about the competitive process.
Consider an established enterprise software company that adds AI capabilities to its existing product. The feature makes the product faster and more capable for customers the company already serves.
That may be an important technological innovation. Competitively, however, it can look much more like sustaining innovation: better performance for an existing market.
Now consider a different company using AI to provide a simplified professional service at a fraction of the traditional price. Customers who previously could not afford the service can suddenly accomplish part of the job themselves.
That could resemble a new-market foothold.
The underlying technology is similar. The competitive trajectory is not.
The relevant questions concern the business around the technology. Who is being served, what trade-off are they accepting, do the economics work, what is currently worse about the new product, and is that weakness improving?
Most importantly, are customers actually moving?
Experimentation is not migration. A large number of people trying an AI tool does not by itself establish that an existing market is being displaced.
The theory becomes useful when it forces those questions. It becomes weak when AI is disruptive substitutes for answering them.
Use Disruption Theory to Form a Hypothesis
This suggests a more practical role for the framework.
Instead of trying to classify a company or technology as disruptive in advance, use the theory to form a hypothesis about a possible competitive trajectory.
An incumbent might notice that part of its market increasingly complains about price or complexity. A smaller competitor serves those customers with a weaker but much simpler product, and the incumbent has little interest in matching its economics.
That is worth watching.
If the entrant’s product improves, its economics strengthen, and customers begin migrating from increasingly valuable segments, confidence in the disruptive hypothesis should rise.
If performance plateaus, customers refuse to move, or the entrant cannot make the economics work, confidence should fall.
The reasoning becomes:
Possible foothold
↓
Different economics?
↓
Performance improving?
↓
Customers moving?
↓
Incumbent response unattractive?
↓
Hypothesis strengthens or weakensThis is much more useful than attaching the label disruptive at the beginning.
It also solves part of the incumbent’s resource-allocation problem. The company does not need to choose immediately between ignoring the threat and making a massive commitment.
It can preserve an option.
A credible small investment can create enough capability to learn about the new market, test assumptions with customers, understand the technology, and respond more quickly if the evidence strengthens.
The word credible matters. An innovation team with no production path, customer access, budget, or ability to obtain additional resources does not preserve much of an option.
If the experiment succeeds and the organization still cannot scale it, the incumbent has learned about the future without improving its ability to participate in it.
Eventually, strong evidence has to change resource allocation.
That is when disruption stops being an interesting theory and becomes a strategic decision.
The Theory Is Most Useful Before the Answer Is Obvious
Disruptive innovation theory does not need to predict the winner to be valuable.
Its strongest insight is narrower and more durable: good management can create a systematic bias against emerging opportunities whose customers, margins, and economics do not yet fit the incumbent organization.
That gives decision-makers something concrete to investigate.
Are some customers being overserved? Is nonconsumption becoming economically addressable? Can an entrant survive in a market the incumbent considers unattractive? Is its product improving toward more demanding use cases, and are customers actually moving?
The answers do not produce certainty.
They change the strength of the hypothesis.
That is a better standard for a strategic framework than pretending it can announce tomorrow’s winner from today’s weak signals. Markets contain too many interacting variables for the label alone to provide that confidence.
It also preserves the boundary around the theory.
A market can be transformed by a superior sustaining innovation, regulation, a new distribution model, category convergence, or technological substitution without following Christensen’s low-end or new-market mechanism. Those changes can be enormously consequential without needing to be squeezed into disruptive innovation theory.
The framework becomes stronger when it is allowed to explain less.
Its purpose is not to tell us that change is happening. It is to explain one reason an apparently weak entrant can receive time and space to become stronger while a capable incumbent rationally directs its resources elsewhere.
For an incumbent, the practical challenge is therefore not predicting every disruption. That is impossible.
The challenge is building enough sensitivity to recognize when the conditions are becoming more credible, preserving an option while uncertainty remains high, and being capable of moving real resources if the evidence eventually justifies it.
For everyone else, the discipline is simpler: stop asking whether every new technology is disruptive.
Ask what competitive mechanism is actually unfolding.
Disruptive innovation theory is most useful not when it declares the future, but when it tells us what evidence would make one particular future more plausible than it was yesterday.
