On a winter afternoon in 1975, an engineer at Kodak named Steven Sasson sat in a laboratory in Rochester, New York, and assembled a device that would eventually destroy the company that paid his salary. It was the size of a toaster, weighed about eight pounds, captured images in black and white onto a cassette tape, and took twenty-three seconds to record a single photograph. When Sasson’s bosses at the world’s most famous photography company examined it, they were not dismissive of the technology itself. They understood immediately what it could become. Their reaction, recounted by Sasson decades later, was the quiet fatalism of people who can see the future arriving and feel powerless to prepare for it. A digital photograph, they told him, would one day make film obsolete. And then they put the prototype in a drawer.

This is one of the most frequently told stories in business history, and it is usually framed as a lesson about innovation or about the perils of protecting legacy revenue. But there is another way to read it, one that gets far less attention and may matter more. The Kodak that shelved that prototype did not fail because its engineers lacked vision. It failed because the organization could not act on what it already knew. The knowledge that film was mortal was present inside the company, encoded in the caution of executives, the skepticism of the sales force, and the experience of veterans who had watched other formats rise and fall. Yet that knowledge never translated into the kind of decision that could have preserved the business. Kodak is studied today as a textbook case of technological disruption. It deserves to be studied just as much as a case of corporate amnesia, a company that forgot its own history while it was still living it.

The Memory of an Organization

The idea that a company has a memory is not a metaphor. Since the early 1990s, when researchers Walsh and Ungson published their landmark framework on organizational memory, management scholars have treated firms as information processing systems that store their experience the way a brain stores a life. The parallel is not perfect, but it is revealing. A person’s memory is not a single filing cabinet in the head. It is distributed across a web of neural connections, and damage to one region does not erase a memory so much as make it unreachable. The same is true of a company. Its memory lives in several places at once, and the researchers identified five distinct storage bins where experience accumulates.

The first bin is the individuals themselves. Every veteran employee carries an archive of hard-won knowledge: the supplier who quietly delivers late goods despite what the contract says, the machine that fails every third Thursday, the customer who cannot be trusted regardless of what their credit file suggests. This is tacit knowledge, the kind that is never written down because it lives in judgment rather than in documents, and it walks out the door every time an experienced person retires or resigns.

The second bin is the organization’s culture, the set of shared assumptions and stories that tell people what is valued and what is punished. Culture is memory in its most durable form, because it survives individual departures and is passed from one generation of employees to the next through example and ritual rather than through instruction. The third bin is the routines and processes of the organization, the procedures that encode prior learning in the way work is actually done. The fourth is the formal structure, the reporting lines and decision rights that reflect what the organization has learned about who should decide what. And the fifth is the external record, the archives, the files, the databases, and the institutional documents that persist long after their authors are gone.

The critical insight is that most of the knowledge a company needs to survive is spread across all five bins at once. When a piece of knowledge exists only in one bin, it is fragile. When it exists only in the heads of individuals, a single retirement can erase it. When it exists only in the culture, a merger can overwrite it. And when it exists only in the archives, it is inert, present but unreachable, the way a book on a shelf is not knowledge until someone opens it. Forgetting is rarely a single catastrophic event. It is a slow erosion across many small losses, and by the time it becomes visible, the memory is often gone beyond recovery.

The Silent Forms of Forgetting

Understanding why companies forget requires a distinction that researchers de Holan and Phillips drew more than two decades ago. Forgetting can be intentional, a deliberate choice to unlearn obsolete routines in order to adapt, or it can be accidental, an unplanned loss of knowledge that the organization needed and did not mean to discard. The first kind of forgetting is a necessary part of change. The second is a quiet destroyer of value, and it is far more common than most leaders recognize.

The most obvious mechanism of accidental forgetting is turnover. Employees leave, take their tacit knowledge with them, and the organization becomes slightly less competent in ways that are difficult to notice at any single moment. The damage compounds because it is invisible. No metric on a dashboard captures the judgment that left with a retiring plant manager. No report quantifies the caution that departed with the risk officer who remembered the last downturn. Over years, a company can lose the very capability that made it successful while its financial statements continue to look healthy.

The second mechanism is what might be called narrative reconstruction. When a company experiences success, it tends to rewrite its own history, smoothing over the struggles, the near-death experiences, and the luck that carried it through. The stories it tells itself and its investors become cleaner and more heroic than the reality. This is survivorship bias applied to one’s own past, and its cost is that the lessons embedded in the struggle get forgotten along with the struggle itself. The discipline that saved the company in its difficult years, the frugality, the wariness of debt, the respect for competitors, is rationalized away as the product of a simpler time, and the next generation of leaders, raised on the polished version of the story, lacks the instinctive caution their predecessors had earned the hard way.

The third mechanism is the arrival of a new leader with a mandate for change. Every incoming chief executive faces pressure to demonstrate that the past regime was a failure, and the cleanest way to do that is to dismantle its artifacts. New leaders purge the ranks, dissolve the old teams, discard the old playbooks, and declare that the future will be unconstrained by the past. Sometimes this is exactly what a company needs, but more often it is a wholesale disposal of institutional memory dressed up as transformation. The phrase that the company must forget its history to write a new chapter sounds bold in a press release and is catastrophic in practice, because the history being discarded contains not just the mistakes the company wants to escape but the knowledge of how to survive that it cannot afford to lose.

The fourth mechanism is the corporate amnesia loop, a pattern that recurs so reliably across industries that researchers have given it a name. An incident occurs. A product fails, a project misses its deadline, a risk materializes. The organization responds with a temporary correction, restores stability, and conducts a review that produces a document titled lessons learned. Then the document is filed, the participants disperse or retire, and the conditions that produced the incident quietly reassert themselves. Years later, a strikingly similar incident occurs, and the organization responds with genuine surprise, because nobody who remembers the previous lesson is still in the room. The loop is so common that the organizations that break it are not merely better managed. They are operating on a different psychological plane.

The Price of Forgetting

The cost of this forgetting is not abstract. It is visible in the fate of companies that lost their memory at the exact moment they needed it most. Consider Boeing, which for much of the twentieth century was the embodiment of engineering memory. Its culture, built over decades, encoded a respect for the incremental, the testable, and the safe, the accumulated wisdom of generations of engineers who knew that in aviation, caution is not a cost but a survival trait. In 1997, Boeing merged with McDonnell Douglas, and the merger is widely regarded by analysts as the moment the company’s memory began to erode. The engineering-led culture of the old Boeing collided with the finance-led culture of McDonnell Douglas, and the finance-led culture won. Experienced engineers who embodied the old judgment were pushed out or retired early. Decision-making shifted from the hangar to the spreadsheet. The organizational memory that had kept Boeing safe was dispersed across the five storage bins, and then, bin by bin, it was drained.

Two decades later, two Boeing 737 MAX aircraft crashed within five months of each other, killing 346 people. The investigations that followed described a certification process in which safety assessment had been subordinated to schedule and cost, a chain of decisions that a company still in possession of its old engineering memory might have resisted. The financial cost was staggering: grounding, canceled orders, tens of billions in market value destroyed, and a reputational and regulatory reckoning that continues to this day. The tragedy is usually analyzed as a governance failure or a regulatory failure. It was also a memory failure, a company that had forgotten what it once knew about how to stay alive.

The same pattern appears at the level of entire institutions. When the Apollo program ended, NASA lost not just its budget but its workforce. The engineers who had built the Saturn V, and who carried in their heads the knowledge of how such a program was managed, dispersed to other industries or retired. When the space shuttle era demanded capabilities similar in kind if not in scale, the agency had to relearn lessons its predecessors had already mastered. The Columbia Accident Investigation Board, reporting in 2003, explicitly identified the loss of institutional memory and the erosion of the organizational culture that had made Apollo possible as contributing causes of the disaster that killed seven astronauts. The report is, in large part, a document about forgetting.

Financial crises follow the same logic. Every generation of bankers believes it is more sophisticated than the last, and every generation discovers that the lessons of its predecessors had not been preserved but merely stored in archives that nobody read. The executives who lived through the savings and loan crisis of the late 1980s carried a visceral sense of how quickly asset quality can turn. When their retirement and their replacement by a generation that knew only prosperity, the institutional memory of that earlier trauma faded, and the conditions that produced the 2008 collapse reasserted themselves. Analysts who study the recurrence of financial crises increasingly describe them not as failures of regulation or greed but as failures of memory, the predictable return of risks that were identified, catalogued, and then forgotten.

Where Forgetting Is Not Tolerated

The most instructive evidence that corporate memory is a choice rather than a fate comes from the industries that refuse to forget. Commercial aviation is among the safest industrial activities in human history, and its safety record is not an accident of technology. It is a product of a system designed to remember. Every incident, no matter how minor, is investigated. Findings are shared across the entire industry, not buried in the files of a single carrier. Lessons are codified into checklists, training programs, and regulations that persist independent of any individual’s memory. When a pilot dies or a plane is damaged, the knowledge extracted from that event becomes part of the permanent operating memory of the entire field. The system does not rely on any one person to remember, because it embeds memory into structure.

Medicine operates on the same principle. Hospitals hold morbidity and mortality conferences at which failures are examined openly, with the explicit purpose of learning rather than assigning blame. Guidelines evolve from these discussions and become part of the institutional record. The practice is imperfect, but its existence signals something profound: an understanding that failure, studied and preserved, is a form of knowledge that no amount of training can replace.

The most extreme demonstration comes from the oldest companies in the world. Japan’s shinise, businesses that have operated for centuries, treat continuity as a sacred obligation. Kongo Gumi, the construction company founded in 578 and run by the same family for over fourteen hundred years, and Hoshi Ryokan, the inn that has been in continuous operation since 718, built their longevity not on brilliant strategy but on a disciplined culture of apprenticeship, mentorship, and deliberate succession. Knowledge was transferred from generation to generation through direct teaching rather than through documents, and memory was treated as the most valuable asset in the enterprise. Such firms are rare, but their existence proves that organizational memory can be sustained across spans of time that dwarf the life of most modern corporations.

The financial industry offers a less heroic but equally instructive example. Auditors reconcile forecasts against actual outcomes, and the discipline of measurement creates a form of memory that even institutional forgetting cannot easily erase. When a company’s projections are checked against its results year after year, the organization cannot fully rewrite its own history. The record persists, and the next forecast is implicitly benchmarked against the reality of the last one. This is memory imposed by process, and it is the closest thing most organizations have to a guarantee that their past will not be silently erased.

The Anatomy of Remembering

What, then, do the companies that remember actually do differently? The answer, in every case, is that they treat memory as a deliberate project rather than a happy byproduct. They build mechanisms that convert individual experience into institutional knowledge, and they build them on purpose.

The first mechanism is the institutionalized postmortem, and the distinction between a real one and a performative one is stark. A performative postmortem produces a slide deck, presents it once, and files it where no one will ever retrieve it. A real postmortem produces a change in behavior. It captures not just what happened but why it happened, the reasoning, the assumptions, and the warning signs that were ignored. Bridgewater Associates, the hedge fund founded by Ray Dalio, institutionalized this practice to an unusual degree. Dalio and his colleagues maintained a systematic log of mistakes, each one documented with its root cause and the lesson to be drawn from it. The log was not a private record of past errors. It was a working memory that shaped future decisions, and it allowed the firm to avoid repeating errors that other organizations make again and again.

The second mechanism is cultural. Toyota, whose production system became the most studied manufacturing approach of the twentieth century, built memory into its routine operations through the andon system, the cord that any worker could pull to stop the production line when a problem appeared. Stopping the line costs money, and the willingness to accept that cost reveals the deeper logic: a problem surfaced and studied is an investment, while a problem hidden is a liability. Toyota’s culture treated every mistake as information to be preserved, and the cumulative effect of that attitude, applied over decades, was a compounding knowledge advantage that competitors could not replicate by copying Toyota’s processes, because they had not copied Toyota’s memory.

The third mechanism is the deliberate cultivation of veterans and the deliberate transfer of their knowledge. Companies that remember do not treat succession as a handover of authority. They treat it as a transfer of memory, and they structure it accordingly. Apprenticeships, mentorship programs, cross-functional rotations, and communities of practice all exist for this purpose, even when they are not described that way. The most successful serial acquirers, companies like Danaher and Constellation Software, approach their acquired businesses with the discipline of memory preservation, keeping the people who hold the knowledge and integrating the lessons rather than the labels.

The fourth mechanism is a particular relationship with the past, one that is best exemplified by the investors who have made the study of other people’s mistakes a competitive strategy. Warren Buffett and Charlie Munger built one of the great fortunes of modern capitalism on an insistence that history be read, not forgotten. Munger in particular was famous for urging people to study the errors of others so that they could be avoided without being experienced. This is the individual investor’s version of institutional memory, and its logic is identical: knowledge does not have to be lived to be useful, provided it is preserved and retrieved at the moment of decision.

Reading a Company’s Memory

For the investor, the psychology of corporate memory offers a lens that is not captured anywhere in the financial statements. The balance sheet records assets that can be valued. Memory is an asset that cannot be valued, but it can be observed, and the observation is often more predictive of long-term outcomes than the numbers.

The first thing to observe is how a company treats its own failures. Does it conduct genuine postmortems, or does it produce performative documents? Does it acknowledge its mistakes in public, or does it rewrite its history to erase them? The leader who can name the company’s past failures, and describe what was learned from them, is a leader with access to the organization’s memory. The leader who describes the company’s past as an unbroken string of successes is either lying or suffering from amnesia, and both conditions are dangerous.

The second observation is turnover, not in the aggregate but in the distribution. A company with high turnover across the board is losing memory everywhere, but a company that loses its veterans while keeping its juniors is losing exactly the knowledge that cannot be replaced. The departure of the generation that lived through the last crisis, the last failed launch, the last near-death experience, is a signal worth taking seriously. Institutional memory has a half-life, and when the people who embody it leave, the half-life accelerates.

The third observation is how the company navigates change. Every organization must unlearn obsolete routines to survive, and the skillful ones distinguish between intentional forgetting, the deliberate release of what is no longer useful, and accidental forgetting, the careless loss of what is still essential. When a company undergoes a transformation, watch what it preserves. A transformation that throws out the wisdom along with the processes is not a transformation. It is a memory wipe.

The fourth observation is succession, and here the corporate memory lens reveals something that conventional analysis misses. The companies that plan succession as a transfer of knowledge, with overlapping tenures, formal mentorship, and the deliberate cultivation of the next generation, are treating memory as an asset. The companies that treat succession as a sudden handover of power are treating it as a reset, and the difference shows up years later, when the knowledge that should have been transferred is discovered to have vanished.

The Moat That Does Not Appear on the Balance Sheet

There is a reason so much of the advice given to investors sounds like the advice given to students of history. The study of markets is the study of what has been tried, what has failed, and what has been forgotten. The companies that succeed over the long term are rarely the ones with the best strategy in any given year. They are the ones that refuse to forget, that preserve the accumulated experience of their people, their culture, their routines, and their archives, and that convert that experience into the kind of judgment that cannot be encoded in a spreadsheet.

Memory is the moat that does not appear on any balance sheet. It is built slowly, through decades of accumulated experience, and it is destroyed quickly, through a merger, a purge, a generation’s retirement, or a leader’s conviction that the past has nothing to teach. When memory is preserved, the organization compounds experience, and each mistake becomes tuition for the next success. When memory is lost, the organization compounds mistakes, and each crisis becomes a rehearsal for the next one, indistinguishable except for the date on the calendar.

Kodak had the future in a drawer in 1975, and it forgot that it was holding it. The tragedy of corporate amnesia is that it is rarely visible at the moment it matters. The balance sheet looks healthy. The products are selling. The strategy is coherent. And inside the organization, quietly, bin by bin, the knowledge that would save it is evaporating. The investor who learns to see memory, and to value the companies that keep it, is looking at something most of the market does not even know exists. That, in an age that celebrates disruption and the eternal new, may be the most durable edge there is. The past, properly remembered, is not a weight. It is the only reliable map of the future that any organization will ever possess, and the companies that keep reading it are the ones that will still be here to be read about.