Why smart companies repeat dumb mistakes

A version of this article originally appeared in Quartz’s Leadership newsletter. Sign up here to get the latest leadership news and insights straight to your inbox.

"Here we go again."

It’s a feeling many of us Cleveland Browns fans experience at some point in the second half of the first game of each new season. We find some way to convince ourselves that we’re due for good things to happen, but at least since returning to the NFL as an expansion team in 1999, good things rarely do. 

From 1999-2021, the Browns’ record after Week 1 of the season is one win, 22 losses, and one tie. Woo-hoo! Then, the team rattled off three consecutive opening game victories from 2022-24, which might lead you to believe the organization had begun a cultural turnaround. The head coach won the league’s coach of the year award twice in his six seasons on the job.

He was fired this past offseason.

The Cleveland Browns are the professional sports poster child for appearing to make the same mistakes over and over again, or in other words, an organization suffering from memory loss.

Replacing the head coach may or may not work out for them. This isn’t a cheap shot toward the team I’ve loved like family (I see you, Jets fans) since I was in kindergarten.

The people running the Cleveland Browns are neither stupid nor lazy. Quite the opposite. But outside of some supernatural explanation involving curses or jinxes, how can the team consistently perform so poorly in a league intentionally designed for parity?

Maybe it’s institutional forgetting. Organizational memory loss.

Organizational memory is the knowledge accumulated from past decisions

Organizational memory lives in three places, said Arthur Favier, founder and CEO of Oppizi, a firm which digitizes offline marketing. Those three places are:

• People
• Process
• Culture

“If one fails, the other two have to pick up the slack,” Favier said. “If you have high turnover, for example, your processes have to be extremely codified. If you have a structure that’s constantly in flux, your culture has to be one that understands why we made certain decisions in the past.”

In high-growth environments, you’re not just losing people with turnover, but the context for which decisions were made during their tenure, he said.

“New people are given tasks, not the rationale for the tasks,” Favier said. “Decisions are repeated, not knowing the trade-offs that went into the decision. This is how organizations begin to implement the same solution twice.”

Turnover is the most obvious and visible cause

Alex Sarellas, managing partner and CEO of PAJ GPS, a maker of GPS tracking devices and security solutions for vehicles and people, agrees with Favier that this problem hits tech companies especially hard.

“When an engineer leaves our team, for example, we aren’t just losing someone who understands code or was filling up a role,” Sarellas said. “We’re also losing all of the reasoning this person put into past decisions, the shortcuts they knew, and the kind of practical knowledge that no amount of writing can teach you. I’ve seen the same problems come up again and again just because this one person who knew the entire backstory and cause has left the team.”

Archie Payne, co-founder and president of CalTek Staffing, sees these same pain points play out due to his focus on IT and engineering, he said.

“Organizational memory is a competitive advantage, particularly in highly technical environments, and when it disappears companies often lose quality, efficiency and morale along with it,” Payne said. “Turnover isn't the only thing that causes this issue, but it is the most obvious culprit. Losing a senior engineer means also losing their context around why systems were built a certain way, as well as any decisions they didn't fully document and their hard-learned lessons from past failures. 

“As a result, future teams repeat the same mistakes as in the past because they don't fully understand the reasoning behind certain decisions. I would say this issue is a particularly high risk in IT environments. Legacy systems and architecture decisions often live in people's heads, meaning companies inherit technical debt they didn't even realize they had when those individuals exit.” 

The first step to counteracting the issue is for leaders to treat knowledge as an asset rather than an afterthought, Payne said.

Sarellas’ company has implemented these very changes.

“Now, we document as much as possible and are trying to build systems that help the company ‘remember,’” Sarellas said. “Clear incident logs, short notes that explain the reasoning behind major decisions, and cross-team reviews where people share the knowledge they have.

“I think the biggest change has been cultural, though, because we’re encouraging teams to write down why we did something instead of just what was done. This makes the information valuable and also adds to the company’s memory, even if a team member leaves.”

The turnover problem is easy enough to spot. Other causes of structural amnesia may be less obvious, Favier said.

Companies are often changing direction faster than they learn. Examples of structural forgetting include:

• Leadership transitions
• Rebranding
• Aggressive pivots
• Frequent restructuring

“When the organization’s priorities change every 12-18 months, learning compounds too slowly,” Favier said.

Another factor that leaders may not account for is the concept of documentary theatre, Favier said.

“Organizations produce reports and dashboards, but not the infrastructure for knowledge transfer,” he said. “Knowledge exists, but it’s not being leveraged. This is not memory, this is storage.”

Acquisitions disrupt institutional memory

“The idea that companies fail because they lack memory rather than intelligence is especially visible in mergers and acquisitions,” said Klint Kendrick, adjunct instructor of human capital management at the NYU School of Professional Studies.

“Most acquirers are smart and sophisticated,” Kendrick said. “They model financial synergies, assess market positioning, and run robust due diligence. However, they struggle to preserve the operating memory embedded in the target company.”

MIT Sloan compared acquired employees to similar new hires at the same firm and found that acquired workers are 22% more likely to leave in the first year, Kendrick said.

“Retention rates improve when the founders stay onboard and when the acquired firm remains standalone for some time rather than being integrated quickly,” Kendrick said. “This second pattern suggests the issue goes beyond turnover.

“When we evaluate this through the lens of organizational memory, an acquired firm’s success often lives in experiential learning rather than documentation. This means the decision-making process, team coordination styles, customer relationships, and tacit technical knowledge are critical success drivers. During integration, those elements are altered before buyers understand how they produced results.”

Over time, that disruption appears as slower growth, stalled synergies, and missed financial targets and milestones, Kendrick said. 

“New leadership teams treat recurring issues as novel because the people and structures that once carried the solution are gone,” he said.

Approaching the retirement cliff

More than half of frontline workers over 55 plan to retire within five years, said Lynda Braksiek, the principal research lead for knowledge management at the American Productivity & Quality Center (APQC), a best practice and benchmarking firm.

The largest knowledge transfer event in modern workplaces may already be underway, and most organizations aren’t ready, Braksiek said.

According to APQC:

• Only 8% of organizations consistently capture knowledge
• Only 35% document critical knowledge

“Organizational dynamics, not just turnover, accelerate knowledge erosion,” Braksiek said. “Knowledge transfer efforts often stall due to lack of leadership buy-in, time, resources, supportive culture, and technology. Silos further keep expertise in isolated pockets, leading to duplicated work, slower decisions, and lost insight across functions.”

Mark Gillilan, founder and chief operator of Kyoto Botanicals, a wellness brand, said that outsourcing creative leadership erodes institutional brand memory.

Companies hire new CMOs. Those CMOs hire new agencies of record. A new brand narrative is formed. 

“The agency of record model is the primary driver of institutional memory loss in modern brands. I have seen it from the inside at too many companies: they struggle and fail because they outsource their core brand (branding and creative) to external agencies and lose the soul of the brand in pursuit of shiny and new marketing whenever a new CMO takes over and switches AORs,” Gillilan said.

“The company's history and DNA and ‘why’ vanish as the new CMO and agency try to prove their worth, and the best way they know how to do that is through change. Over enough time this leads to the dilapidation of the core brand itself and the connection to its history and its customers,” he said.

This cycle leads to brands getting reinvented every few years instead of brand equity being compounded.

This is the argument for keeping creative branding efforts 100% in-house, Gillilan said.

How to fight organizational memory loss

Payne recommends enhanced documentation with decision logs, and meaningful exit interviews where key contributors leave behind insightful, instructive manuals on how and why they did things.

Sarellas is especially focused on documenting not just what happened, but why. 

Braksiek recommends knowledge mapping and what she referred to as communities of practice.

“Leaders can combat memory loss through intentional knowledge transfer and communities,” Braksiek said. “Structured approaches including risk assessments, expert interviews, knowledge mapping, and customized transfer plans can help preserve critical tacit knowledge before it disappears. Communities of practice break down silos, accelerate learning, and connect employees to expertise across the enterprise.”

AI can be used to help scale institutional memory but only with strong governance and content management, Braksiek said.

“AI is already being used to record, summarize, and convert expert insight into reusable knowledge assets,” she said. “However, barriers to adoption are largely rooted in governance and content management gaps, not necessarily cultural resistance. Without structure and oversight, AI can amplify outdated or inaccurate knowledge instead of preserving institutional wisdom.”

Organizations don’t only suffer from bad decisions. They suffer from forgotten ones.

Whether it be a football team, a startup, or a global behemoth, institutional learning only compounds when memory survives leadership changes, turnover, and time.

Otherwise, new seasons often begin the same way — with rebuilds and reset buttons.

"Here we go again!"