When the Numbers Say Stay
How Financial Metrics Enable
Disruptive Innovation to Blindside Incumbents
Based on the
work of Clayton M. Christensen
Introduction: Why Disruption Matters Now
In an era defined by generative AI, trade policy upheaval,
and the automation of knowledge work, the concept of disruptive innovation has
never been more relevant. The term gets thrown around loosely in boardrooms and
business media, but its precise meaning, as developed by the late Harvard
Business School professor Clayton Christensen, carries implications that most
leaders still fail to internalize. Understanding what disruption actually is,
and how a company's own financial architecture can accelerate it, is not merely
an academic exercise. It is a survival skill.
Christensen's two landmark works, The Innovator's Dilemma
(1997) and The Innovator's Solution (2003), influenced some of the most
consequential business leaders of the past three decades. Steve Jobs cited
Christensen explicitly when explaining Apple's shift to iCloud, stating that
the people who invent something are usually the last ones to see past it, and
that Apple did not want to be left behind (Isaacson, 2011, p. 532). Jeff Bezos,
for his part, was a fan of the Innovator's Solution. For business professionals,
CPAs, and knowledge workers confronting a world where AI can automate
significant portions of their output, these frameworks offer something more
valuable than any single technology: a way of thinking.
Material wealth, scientific discoveries, and industrial
inventions are all of lower importance than the mental models used to
understand them. The frameworks that follow are not about predicting the
future. They are about recognizing patterns that have played out repeatedly
across industries and asking whether the same dynamics are now playing out in
your own.
The Mechanics of Low-End Disruption: Steel as a Case Study
The steel industry provides what may be Christensen's most
powerful and detailed illustration of how disruption works in practice. The
story is not fundamentally about technology. It is about margins, incentives,
and the rational decisions that lead incumbents to cede their own markets one
segment at a time.
In the 1970s, integrated steel mills like US Steel (USX)
dominated the industry using blast furnaces that required billions of dollars
in capital and needed to run continuously. A new breed of competitor, the
minimill, emerged using electric arc furnaces to melt scrap metal. Nucor was
the most prominent example. The technology was simpler and more flexible:
minimills could ramp production up and down based on orders, their startup
costs were measured in millions rather than billions, and they ran on scrap metal
rather than raw iron ore (Christensen, 1997).
But the steel produced by minimills was not as high quality.
It started at the very bottom of the market, in rebar, which represented only
about 4% of total steel production. The quality was slightly lower, but the
cost was roughly 20% cheaper than what the integrated mills charged – who were
earning a 7% gross margins (Christensen, 1997).
The Repeating Cycle
What happened next is the signature pattern of low-end
disruption, and it repeated itself four times across the steel product
hierarchy. When the minimills entered rebar, integrated steel mills like US
Steel looked at the numbers and made a perfectly rational decision: rebar was
only 4% of the product mix, carried the lowest margins, and was not worth
defending. They ceded the market to the disruptors and focused their resources
on higher-margin products like angle iron, structural steel, and sheet steel.
But when integrated steel exited rebar, something
predictable happened to pricing. With no incumbent setting a price floor, the
minimills were left competing only with each other. Prices collapsed by roughly
20%. The margins that had attracted them into rebar evaporated, and the
minimills had a powerful incentive to move upmarket to angle iron, which
carried approximately 12% gross margins (Christensen, 1997).
The same story then repeated at the angle iron level. The
minimills entered with slightly lower quality but a 20% cost advantage.
Integrated steel again ceded the market. Prices collapsed again. The minimills
moved up to structural steel, which offered 18% gross margins. Integrated steel
ceded once more. And eventually, the minimills moved into sheet steel, the
largest and most profitable segment at 55% of total production, completing the
disruption of big steel (Christensen, 1997).
The outcome is visible in market valuations. By 2025,
Nucor's market capitalization stood at approximately $26.6 billion, while US
Steel's sat at roughly $9.6 billion. The disruptor had tripled the incumbent's
value. At each stage, integrated steel believed it was making a smart strategic
decision by retreating to higher-margin territory. At each stage, it was wrong.
The Innovator's Dilemma: When the Numbers Tell You to Stay Put
The steel story is compelling as a narrative, but its real
power lies in the financial mechanics underneath. US Steel was managed by
capable professionals. The company understood the disruptive threat of the
minimills. Management chose not to respond, as financial metrics clearly
supported maintaining the status quo.
Sunk Costs and Marginal Cost Analysis
Christensen's article "Innovation Killers: How
Financial Tools Destroy Your Capacity to Do New Things" (2008) laid out
the core financial trap. The critical error was that US Steel focused on
marginal costs rather than full costs for new capabilities. The company's
existing blast furnace infrastructure was already built and largely
depreciated. Using that excess capacity to produce steel cost roughly $50 per
ton in variable costs, yielding about $300 per ton in cash flow at a revenue of
$350 per ton. Against a depreciated net book value of around $60 million, this
produced a return on investment of approximately 400% (Christensen et al.,
2008).
Now compare that to the alternative: investing in a minimill
plant like Nucor's. The cost per ton would be approximately $270 in hard cash
outlays, yielding only $80 per ton in cash flow. Total cash generation at
800,000 tons would be about $64 million, against a required investment of $260
million in real capital. The ROI? Roughly 24.6% (Christensen et al., 2008).
This is the innovator's dilemma in its purest form: do you
invest $260 million to earn a 25% return, or do you continue business as usual
and earn 400%? The answer, for any manager evaluated on ROI, EPS, or short-term
performance metrics, is obvious. You stay the course. And by staying the
course, you accelerate your own disruption.
Earnings Per Share and the Shareholder Trap
Christensen extended this analysis beyond the operational
level to the capital markets. Management teams are not only beholden to their
internal cost models; they are also under intense pressure from shareholders
and financial analysts to deliver consistent earnings growth. This creates a
second layer of bias against disruptive investment.
Christensen noted that stock buybacks, which were actually
illegal in the 1970s because they were considered a form of share price
manipulation, became a preferred tool for boosting earnings per share. Rather
than investing excess cash in new capabilities, companies returned capital to
shareholders, artificially inflating EPS while hollowing out their capacity to
innovate. Christensen cited research showing that senior executives were
routinely willing to sacrifice long-term shareholder value to meet short-term
earnings expectations or to smooth reported earnings (Christensen et al.,
2008).
The financial press reinforced this dynamic. When Gary
Works, a US Steel facility, announced it would focus almost entirely on
higher-value flat-rolled steel and abandon lower segments, analysts praised the
move as a quiet comeback. However, the framework of disruptive innovation
offers a different perspective: the financial press was endorsing behaviours
that were, in fact, facilitating disruptors in advancing within the market. The
values embedded in the financial architecture, the definitions of what constitutes
a good deal and a bad one, were shared not just within the company but across
the entire ecosystem of analysts, investors, and reporters who shape corporate
decision-making.
The DCF Trap: Two Fatal Assumptions
Christensen's critique extended to the discounted cash flow
model itself, a tool that remains central to corporate finance education and
practice. He did not argue that DCF was wrong in normal operating conditions.
When a company is in a stable competitive environment pursuing sustaining
innovation, DCF works exactly as intended. The problem arises when an
organization is in a state of disruption and does not know it.
Problem 1: The Status Quo Will Not Continue
The first flaw Christensen identified is the baseline
assumption. In a standard DCF analysis, the "do nothing" scenario
assumes that current cash flows will continue indefinitely. But in a disruptive
environment, the actual baseline is not flat. It is declining, often
nonlinearly. A company like BlackBerry in the years before the iPhone could not
assume that its revenue from selling physical-keyboard smartphones would remain
stable. The actual trajectory was a steep drop-off. As Eileen Rudden at Boston
Consulting Group pointed out, the most probable outcome of inaction is not
continuity; it is accelerating decline (Christensen et al., 2008).
This means that the true delta of a disruptive investment is
not the difference between projected new cash flows and a stable baseline. It
is the difference between projected new cash flows and a falling baseline. When
measured correctly, the case for investment looks far more compelling. But most
companies never run the analysis that way.
Problem 2: Conservative Estimates Get Amplified
The second flaw compounds the first. Christensen
demonstrated that even modest conservatism in estimating the cash flows of a
disruptive investment can produce dramatic undervaluation, because terminal
value calculations amplify small differences. In his example, a conservative
estimate of $175 million in Year 5 cash flows, discounted to perpetuity at a 5%
spread (10% discount rate minus 5% growth rate), yields a terminal value of
$3.5 billion and a total NPV of approximately $4.2 billion. But if actual performance
comes in at $571 million in Year 5, the terminal value jumps to $11.4 billion
and total NPV reaches $13.4 billion (Christensen et al., 2008). A difference of
roughly $400 million in Year 5 cash flow translates into an $8 billion
difference in total valuation.
The implication is stark: management teams reject disruptive
investments not because those investments are bad, but because they
systematically underestimate the cash flow potential. Christensen's proposed
solution was not to abandon quantitative analysis altogether, but to recognize
that when an organization is facing disruption, qualitative judgment must
supplement and sometimes override the numbers.
Implications for Today: From Steel to Knowledge Work
The steel example may feel remote to a financial
professional or CPA, but the underlying dynamics are portable. Consider the
parallel with generative AI and professional services. A law firm billing at
$500 per hour for document review faces the same structural question that US
Steel faced: why would we cannibalize our high-margin work by investing in AI
that could do it for a fraction of the cost? The marginal cost of having an
existing associate review a document is low because the infrastructure (training,
office space, institutional knowledge) is already paid for. The full cost of
building an AI-augmented practice requires significant new investment. The ROI
comparison, on the surface, favors business as usual.
But someone else, a startup without legacy infrastructure or
billable-hour economics, will build that capability. They will start at the low
end, with tasks that incumbent firms consider too cheap to bother with. They
will offer "good enough" quality at a dramatically lower price. And
if the pattern holds, they will move upmarket over time, just as Nucor moved
from rebar to sheet steel.
The broader lesson from Christensen's work is that
disruption is not primarily a technology problem. It is a management accounting
problem, a capital allocation problem, and ultimately a values problem. The
financial metrics that an organization uses to make decisions (margins, ROI,
EPS, NPV) shape the values and culture of the organization. Those values
determine which opportunities get funded and which get ignored. And those
decisions, repeated over time, determine whether the organization survives (Christensen
et al., 2008).
The RPV framework, which stands for Resources, Processes,
and Values, captures this insight. The processes and values that served an
organization well during periods of sustaining innovation become liabilities
during periods of disruption. They are designed for one job and are being asked
to do another. Recognizing that disconnect is the first step toward addressing
it.
For financial professionals navigating the current
landscape, the question is not whether generative AI will change their work. It
is whether their organizations' financial architecture will allow them to
respond before someone else does.
References
Christensen, C. M. (1997). The
innovator's dilemma: When new technologies cause great firms to fail.
Harvard Business Review Press.
Christensen, C. M., Kaufman, S. P.,
& Shih, W. C. (2008). Innovation killers: How financial tools destroy your
capacity to do new things. Harvard Business Review, 86(1), 98-105.
Christensen, C. M., & Raynor, M. E.
(2003). The innovator's solution: Creating and sustaining successful growth.
Harvard Business Review Press.
Isaacson, W. (2011). Steve Jobs. Simon & Schuster.
