One of the strange things about getting older is that you begin to recognize patterns.

Not just patterns in history, but in business, technology, government, and society. Connections become visible between events that once seemed unrelated.

The details change. The underlying behavior often does not.

A successful system is created to solve a real problem. Over time, people build careers, institutions, measurements, and identities around it. Eventually, the system begins serving two purposes: solving the original problem and preserving everything that has grown around the solution.

That is where adaptation becomes difficult.

Knowing the future is not the same as accepting it

The Kodak story is usually told as a simple warning about missing technological change. A Kodak engineer, Steve Sasson, built the first digital camera in 1975. Management feared that digital photography would threaten film, so the company put the invention on a shelf and eventually disappeared.

The truth is more complicated, which makes it more useful.

Kodak did not ignore digital photography. It developed sensors, cameras, imaging systems, online services, and consumer products. It understood that photography was becoming digital. The deeper problem was that Kodak had built an extraordinarily profitable economic system around film, chemicals, processing, and printing.

Digital photography was not merely a new product category. It dismantled the chain of recurring transactions that made Kodak successful.

Kodak could participate in digital photography while still hoping to preserve the economics of film. What it struggled to do was reorganize itself around a future in which those economics no longer existed.

Kodak did not lack information about the future. It had difficulty accepting what that information meant for the present.

Xerox offers a variation of the same pattern.

At Xerox PARC, researchers helped assemble many elements of modern personal computing. The Alto brought together a graphical display, windows, menus, a mouse, networking, and electronic documents in 1973. Xerox later attempted to commercialize many of those ideas through the Star workstation, but the system was expensive and never achieved broad commercial success.

So it is not quite accurate to say that Xerox simply invented the personal computer and shelved it. The company recognized that it had created something important. What it failed to do was turn that recognition into a business capable of defining the new market.

Xerox knew how to sell copiers to organizations. It understood leasing, service contracts, printing, and document reproduction. Personal computing required different products, economics, customers, channels, and assumptions about how people would work.

The technology was inside the company. The operating model needed to make it successful was not.

When an asset becomes an obligation

Sears presents a third and perhaps more tangible version of the pattern.

The company began as a catalog business, but by the time electronic commerce emerged, it had become inseparable from an enormous and expensive retail footprint. Thousands of stores, many occupying large anchor spaces in shopping malls, carried leases, maintenance costs, inventory, employees, and management structures designed around physical traffic.

Those stores were not merely a distribution channel. They shaped how Sears allocated capital, measured performance, organized leadership, and understood itself.

It is tempting to say that Sears could have become Amazon. The company had a national catalog, established customer relationships, purchasing power, proprietary brands, credit operations, warehouses, and experience delivering products to households far from a store.

But those advantages were attached to a business increasingly organized around protecting physical retail.

The stores could have become fulfillment centers, pickup locations, service hubs, or an early bridge between digital ordering and local delivery. Instead, too many became obligations that absorbed capital without creating a compelling reason for customers to return. As sales declined, Sears reduced investment in the stores. The stores then became less attractive, which accelerated the decline in traffic and made further investment even harder to justify.

The footprint that once represented national scale became a system the company could neither afford to renew nor abandon quickly enough.

Sears did invest in online commerce, and its decline began before Amazon became dominant. The failure was not that nobody inside the company noticed the internet. It was that Sears never reorganized its assets, incentives, and operating model around a future in which the store was no longer the center of the customer relationship.

The company possessed many of the ingredients of a modern commerce platform. What it could not do was stop treating its inherited retail structure as the business that had to be preserved.

Each company faced a different industry and made a different series of decisions. But beneath those differences was a shared problem.

The future threatened more than an existing product. It threatened the logic by which the organization understood itself.

Kodak was not simply selling photography. It was organized around the economics of film.

Xerox was not simply helping people work with information. It was organized around the economics of copying.

Sears was not simply connecting households with products. It had become organized around the economics and internal structure of a sprawling retailer.

That distinction matters because organizations rarely describe resistance to change as resistance.

They call it discipline.

They defend margins. They protect jobs. They avoid confusing customers. They honor existing commitments. They wait for the new market to mature. They demand that an emerging business meet the financial standards of the established one.

Each decision can sound reasonable in isolation.

Together, those reasonable decisions can make transformation impossible.

A counterexample in progress

Not every established company responds by defending its existing form.

Germany may be producing an interesting counterexample through the Schwarz Group, the parent company of Lidl and Kaufland.

At first glance, a discount retailer building cloud infrastructure seems like diversification far outside its competence. Look more closely, and the logic becomes clearer.

Operating thousands of stores, warehouses, supply chains, digital services, and transactions across Europe creates enormous internal technology requirements. The Schwarz Group began developing STACKIT in 2018 to operate its own data and infrastructure with greater independence. It used the complexity of its retail businesses as a proving ground, then began offering the resulting cloud capability to outside organizations.

The group has since consolidated cloud computing, cybersecurity, artificial intelligence, communications, and other digital capabilities under Schwarz Digits. It is now building a 200-megawatt data center in Lübbenau, with a planned investment of €11 billion and completion targeted for the end of 2027.

It is far too early to call that investment a success. Cloud infrastructure is capital-intensive, the competition is formidable, and strategic ambition does not guarantee commercial adoption.

But the behavior is worth noticing.

Schwarz is not waiting for foreign cloud providers to define the conditions under which its future business will operate. Nor is it treating technology solely as a support function expected to reduce the cost of selling groceries. It is taking an internal capability developed from the needs of the current business, separating it into a platform of its own, and testing whether it can become part of the next one.

That is almost the inverse of the Kodak pattern.

Kodak developed the future inside the company but struggled to reorganize the business around it. Schwarz is attempting to give an emerging capability its own identity, customers, investment, and strategic purpose before the existing retail model forces the decision.

The result is not yet known. That uncertainty is part of the point.

Strategic adaptation does not mean predicting the future correctly. It means building enough capability and freedom of maneuver that the organization still has meaningful choices when the future arrives.

The pattern becomes harder at the scale of a society

The same pattern appears in public institutions and national economies, but with an important difference.

A company can discontinue a product, close a division, or leave a market. A nation cannot discontinue the people, communities, and regions whose lives were built around an older system.

That makes adaptation slower and morally more complicated.

Educational systems, labor policies, public services, and regulatory structures often continue long after the conditions that produced them have changed. That persistence is not always irrational. Institutions carry accumulated knowledge, legal protections, social trust, and commitments to people who cannot simply be treated as stranded assets.

But protecting people is not the same as preserving every institution in its current form.

An educational system does not protect students by preparing them for an economy that no longer exists. An industrial policy does not protect workers by delaying change until employers can no longer compete. A government does not preserve public trust by allowing administrative complexity to make essential services inaccessible.

The responsible question is therefore not whether to choose people or progress.

It is whether we can preserve human capability, dignity, and security while changing the systems through which they are currently supported.

That requires more than technological investment. It requires transition paths, portable skills, institutional learning, and enough freedom of maneuver to adapt before a crisis removes the ability to choose.

The failure of legacy institutions is not always that they care too much about the people inside them. Often, it is that they confuse protecting the existing structure with protecting those people.

That is the same mistake companies make when they confuse protecting the current business with protecting the future.

Cannibalization is not recklessness

The lesson from Kodak, Xerox, and Sears is sometimes reduced to a slogan: disrupt yourself before someone else does.

That advice is incomplete.

An organization cannot pursue every new technology, abandon every profitable business, or reorganize itself around every prediction. Change has costs, and many supposedly inevitable futures never arrive.

The answer is not permanent disruption. It is preserving freedom of maneuver.

A resilient organization creates room to test a future before the present collapses. It gives emerging capabilities different measurements, protects them from the expectations of the mature business, and allows evidence to challenge the assumptions on which current success depends.

It uses the strength of the present to finance, test, and separate the capabilities that may define the future.

It asks uncomfortable questions early:

These questions apply directly to the current AI transition.

Some organizations will protect legacy processes for too long because changing them threatens existing authority, budgets, or professional identities.

Others will make the opposite mistake. They will pursue rapid labor savings, transfer knowledge and workflows into systems they do not control, and remove human capability before understanding what that capability was actually doing.

Both reactions can leave an organization less able to adapt.

The goal is not to preserve every role or reject automation. It is to understand the whole system before declaring something a saving. People often carry judgment, exception knowledge, customer relationships, informal coordination, and recovery capacity that do not appear on a process map.

Removing a cost is easy to measure. Rebuilding lost capability is not.

Research basis: Kodak company milestones; Computer History Museum and Smithsonian records on the Xerox Alto and Star; Sears corporate history and public filings; Schwarz Group and Schwarz Digits records concerning STACKIT and the Lübbenau data center.

The safer decision

The most dangerous moment for an organization may not be when it is failing.

It may be when the old system is still working well enough to defend.

That is when evidence of change can be dismissed as premature. It is when experimentation looks inefficient, established interests remain powerful, and the consequences of waiting have not yet become visible.

By the time everyone agrees that transformation is necessary, much of the freedom to shape it may already be gone.

Kodak did not fail because digital photography appeared.

Xerox did not lose the personal-computing opportunity because its researchers lacked imagination.

Sears did not decline because customers suddenly discovered that shopping from home was convenient.

Their histories are more complicated than any single explanation. But each illustrates the same strategic danger: the capabilities that produced yesterday's success can become constraints when an organization cannot imagine itself without them.

History does not reliably reward those who change first. Moving early can be expensive, mistimed, or wrong.

But history is particularly unforgiving toward institutions that wait until change is no longer a choice.

That is the paradox.

When we are living through one of these moments, protecting the present almost always feels like the safer decision.

Sometimes it is.

The real work of strategy is recognizing when it is not.