The First Visible Failure Is Usually the Last Stage: What Neurodegeneration Teaches Business Strategy

What if the most dangerous mistake in both medicine and business is not misunderstanding the system, but measuring it at the wrong time?

A company can appear healthy while its business model is quietly accumulating contradictions. A person can carry a devastating genetic mutation while showing no obvious symptoms. In both cases, the visible outcome arrives late. By the time the failure is recognizable, the underlying process may have been unfolding for years.

This creates a powerful connection between two apparently unrelated disciplines: the study of neurodegenerative disease and the practice of business development. Both are concerned with hidden states, delayed signals, and the difficult task of distinguishing a cause from the consequences it eventually produces.

The central lesson is this: a roadmap is only as intelligent as its model of what cannot yet be seen.

The visible event is rarely the beginning

In C9orf72 expansion associated amyotrophic lateral sclerosis and frontotemporal dementia, a genetic alteration can exist long before the full clinical picture becomes apparent. At post mortem, patients may show a familiar pathological pattern associated with sporadic ALS and FTD: TDP 43 has moved from its normal nuclear location and formed aggregates in the cytoplasm.

That observation is important, but it is also easy to misread. The visible aggregate is not necessarily the moment the disease began. It is a late, observable trace of a process that may have involved molecular stress, impaired transport, altered RNA regulation, cellular compensation, and progressive loss of resilience long before a pathologist could identify the final pattern.

The same error appears in business. A company notices declining revenue, rising customer acquisition costs, or a failed product launch and treats the event as the problem itself. But these are often downstream manifestations. The deeper problem may be a business model that depends on unusually cheap acquisition, a customer segment with weak retention, or a product whose apparent popularity never translated into durable value.

Revenue is therefore a little like a clinical symptom. It matters enormously, but it is not the whole system.

A firm can grow while becoming less healthy. It may acquire customers who never return, sell products at a loss, or rely on a channel that competitors can easily copy. Growth can conceal structural fragility in the same way that the absence of symptoms can conceal biological vulnerability.

The first visible failure is often the last stage of an invisible process.

This is why intelligent decision making requires more than collecting outcomes. It requires a theory of the hidden machinery producing those outcomes.

Business models are causal models, not slogans

A business model is often reduced to a simple question: how do we make money? That question is useful, but incomplete. A serious business model should also answer a more difficult set of questions:

• What must be true for the economics to work?
• Which behaviors create lasting value rather than temporary activity?
• Where does the system become vulnerable as it scales?
• Which signals will appear early if the model is beginning to fail?
• What evidence would prove that the central assumptions are wrong?

In other words, a business model is not merely a description of revenue. It is a causal map. It describes how inputs become actions, how actions become customer value, and how customer value eventually becomes economic value.

Consider a subscription software company. Its simplified model might be written as:

Acquisition leads to activation, activation leads to repeated use, repeated use leads to retention, and retention supports profitable growth.

Each arrow is an assumption. If users sign up but do not activate, acquisition is not creating value. If they activate but do not return, the product may be interesting but not useful. If they return only because of heavy discounts, retention may be artificial. If retention is strong but serving each customer is too expensive, the model still fails.

A business plan turns this model into a sequence of commitments: hire engineers, launch a product, test pricing, enter a market, expand distribution. But a plan without a causal model is merely a calendar. It can tell a team what to do next without explaining what the next action is supposed to teach them.

The distinction matters because a roadmap can either reduce uncertainty or conceal it. If every milestone is an output, such as launching a feature or signing a partnership, the team may become skilled at producing activity. If milestones are designed as tests of assumptions, the roadmap becomes an instrument for learning.

The most useful roadmap is therefore not a prediction carved into stone. It is a sequence of increasingly informative experiments.

The timing problem: leading signals and lagging signals

The connection to neurodegenerative disease becomes especially revealing when we examine timing.

A lagging signal tells us what has already happened. In business, lagging signals include quarterly revenue, annual churn, gross margin, and cumulative market share. They are indispensable for judging performance, but they are poor tools for detecting early deterioration because they summarize the past after the underlying choices have compounded.

A leading signal is closer to the mechanism. It may include the percentage of new users who complete a critical action, the frequency with which customers use a product without prompting, the time required to deliver promised value, or the share of revenue generated by genuinely repeat customers.

Neither type of signal is sufficient alone. Leading signals can be noisy and misleading. A rise in engagement may reflect novelty rather than durable value. A promising pilot may not survive expansion. Lagging signals are more stable, but by the time they move, the organization may have spent months or years reinforcing the wrong assumptions.

The practical goal is not to eliminate uncertainty. It is to shorten the distance between a hidden problem and a trustworthy signal.

A useful framework is to divide organizational metrics into four layers:

1. Exposure: What has entered the system? This includes new users, leads, partnerships, or capital.
2. Mechanism: What behavior or process is supposed to convert exposure into value? Examples include activation, repeat use, referrals, or successful delivery.
3. Outcome: What economic or clinical result eventually appears? Examples include profit, retention, functional decline, or survival.
4. Resilience: How much stress can the system absorb before the mechanism breaks?

Most organizations overmeasure exposure and outcome. They count how many people entered the funnel and how much money came out. They undermeasure the mechanism and almost ignore resilience.

That is a serious error. A system may be producing acceptable outcomes only because conditions are unusually favorable. Cheap advertising, generous funding, a forgiving market, or an unusually motivated founding team can compensate for a weak mechanism. Stress reveals whether the result is robust or accidental.

In biology, a cell can compensate for damage until its reserve is exhausted. In business, a company can compensate for poor economics with funding, discounts, overtime, or executive intervention. In both cases, apparent stability may reflect compensation rather than health.

The hidden state: a better way to read evidence

One way to improve strategy is to imagine that every organization has a hidden state of health. This state cannot be observed directly. We infer it from imperfect indicators.

Let us call that state business resilience. It includes customer dependence, operational slack, quality of internal processes, strength of cash generation, and the credibility of the assumptions behind growth. Revenue is one observation of resilience, but it is not resilience itself.

This mental model changes how leaders interpret contradictory evidence. Suppose sales are rising while customer support tickets are increasing, implementation times are lengthening, and renewals are weakening. A simplistic reading says the company is growing and has some operational issues. A hidden state model asks whether growth is consuming the very capacity required to sustain it.

Or suppose a product has low initial usage but exceptional retention among a small group of customers. The visible result looks disappointing. The mechanism may nevertheless be promising. The relevant question becomes whether the company can identify and serve that high value segment efficiently.

This approach also helps avoid the opposite mistake: overreacting to every fluctuation. Not every bad metric signals structural failure. A trustworthy interpretation depends on several questions:

• Is the signal close to the underlying mechanism?
• Does it recur across independent customer groups?
• Does it worsen under stress?
• Does it predict a later outcome?
• Could an unrelated factor explain it?

These questions are versions of a general discipline: do not confuse correlation with location in the causal chain.

A TDP 43 aggregate is a pathological feature associated with a disease process, but observing it does not by itself explain every step that produced it. Likewise, a churn spike is evidence of a business problem, but not a diagnosis. It may result from poor product value, weak onboarding, a pricing change, a market shock, or customers who were never a good fit.

Diagnosis begins when an observation is connected to a mechanism.

From roadmap to diagnostic instrument

This is where business development can become more scientific. Instead of treating the business plan as a commitment to a fixed future, leaders can design it as a diagnostic instrument.

Every major initiative should contain three elements:

A hypothesis. What must be true for this initiative to create value?

An early indicator. What should change first if the hypothesis is correct?

A disconfirmation rule. What result would make us stop, revise, or redirect the initiative?

Imagine a company considering expansion into a new industry. The conventional plan might specify a launch date, marketing budget, sales targets, and hiring schedule. A diagnostic roadmap would begin differently: “We believe this customer segment experiences a frequent problem, will adopt our solution without extensive customization, and can generate a gross margin above a defined threshold.”

The first milestone would not be a full launch. It would be a test of urgency and willingness to pay. The second would test repeat use. The third would test delivery economics. Expansion would occur only after the mechanism survives contact with reality.

This method creates a crucial separation between commitment and learning. A team can commit to learning whether an assumption is true without committing to defending the assumption forever.

That distinction is psychologically difficult. Organizations often reward consistency, so leaders continue funding an initiative because they previously announced it. The roadmap becomes a record of promises rather than a tool for discovering truth. A more adaptive culture treats revision as evidence of learning, not as evidence of incompetence.

The best business developers are therefore not simply deal makers or planners. They are curators of uncertainty. They determine which unknowns matter most, which tests will reveal them fastest, and which resources should remain flexible until the evidence improves.

Key Takeaways

• Map the mechanism behind every important outcome. Do not stop at revenue, growth, or engagement. Identify the sequence of behaviors and conditions that must produce those results.
• Pair every lagging metric with a leading indicator. If annual retention matters, monitor early activation, time to value, repeated use, and customer dependence.
• Treat milestones as experiments. Each roadmap item should state the hypothesis it tests and the evidence that would justify continuing.
• Look for compensation before concluding that a system is healthy. Discounts, overtime, funding, or executive intervention may be masking weak underlying economics.
• Define disconfirmation rules in advance. Decide what evidence would cause you to stop, revise, or redirect before emotional and financial commitments become overwhelming.

The deepest lesson is not that business should imitate medicine, or that pathology offers a simple template for strategy. The lesson is more general and more demanding: complex systems reveal themselves indirectly.

A mutation can precede symptoms. A cellular change can precede visible pathology. A weak business assumption can precede declining revenue. In each case, the observer faces the same challenge: distinguish the surface event from the process that generated it, then act while the system is still capable of changing course.

A business plan should therefore do more than answer how money will be made. It should specify what the organization believes, how those beliefs will be tested, and which early signs would expose a hidden loss of health.

The mature question is not, “Are we growing?” It is: “What invisible condition must remain true for this growth to continue, and how will we know when it stops being true?”

That question transforms a roadmap from a schedule into a form of perception. And in any complex system, seeing earlier is often the closest thing to acting sooner.