I once paid almost a hundred thousand pounds for an organizational transformation strategy that was so irrelevant it moved past useless into harmful territory.
During a post-merger integration, we needed to meld two recently merged companies with very different organizational cultures to deliver a unified digital transformation. The transformation required these teams to work together for the first time, even though they had, until just weeks ago, been competitors.
One company’s culture inspired me to nickname them the ‘Cowboys’. It was the Wild West of IT implementation. Whoever was fastest to draw a weapon won. Surviving to fight another day was more important than playing fair.
The other organization I described is the Bureaucrats. These were people who fundamentally wanted to do a good job, remain apolitical, and deliver exactly what they had promised. The Cowboys were reckless but incredibly innovative. The Bureaucrats were slow and often risk-averse, but achieved every target they set.
What Institutional Memory Actually Is
By ‘institutional memory’, I mean the stories that internal employees collectively tell themselves about a situation, a person, or an event. Cultural institutional memory is the shared lived experience of your values, strengths, and mistakes.
The more people get things done by phoning a friend, asking a favor, or working with others to find a workaround in the actual system, the more important institutional memory becomes. The new processes, however well-documented, are no longer an accurate representation of how your business runs.
Why Starting From Scratch Fails
Bringing these two cultures together was never going to be trivial, so the expensive culture consultants suggested we abandon both cultures and build something completely new. They urged us to start from scratch with new values and behaviors.
While this seems like a good idea if you don’t think about it too much, there are obvious problems. When you try to start from scratch, you ignore all of the mistakes you’ve already made and learned from. People are often very attached to their culture. They joined their respective organizations in part because of that culture, and nothing could fully erase its memory from their ways of working. Ignoring that lived memory often comes across as erasing and invalidating what made the company successful enough to acquire to begin with.
Why You Can’t Skip the Mishaps
Often, when people talk about their lives, even when they’ve made terrible errors, they can acknowledge that without those mistakes, they could not have achieved their successes. They couldn’t have gotten to point C without the mishaps of point B. But what if the crisis of B led to the solution at C? When long-term employees resist transformation approaches, they’re often protecting successful practices that aren’t obvious to newcomers.
Understanding which situations were bad enough to trigger change, or which language convinced managers in the past, becomes a shortcut to creating the conditions required to repeat that success.
Institutional memory is the muscle memory that allows bad ideas to fail faster and good teams to bounce back faster.
That’s why losing employees is so expensive and why recruiting is more expensive than retention. The productivity and cultural gains you lose when someone already well acclimatized to your organization leaves cannot be recreated simply by maintaining an accurate record. This is particularly true for longer-tenured employees who carry the most institutional memory.
The Value of Tenure
In an environment where we tolerate accidental ageism that says if you are over a certain age, you must not understand how AI works, institutional memory gives us a way to value employees with longer tenures.
While they may not fully understand AI, they have the muscle memory of having implemented new things in this organization before. They understand the challenges, whether structural, systemic, or cultural, that people run into.
Ignoring them means you prioritize the mechanics of transformation over its substance. Understanding the people involved, the company, how shareholders perceive certain activities, and how the market has moved over time – all of that is valuable data that isn’t often documented and that you can’t just Google.
What Data Cannot Tell You
If your organization is trying to use AI to train a recruiting bot, the bot will use historical data to understand how you’ve recruited in the past. If you’ve been biased in your recruiting, as many organizations have, the bot will recreate that bias.
Your data will show that men of a certain demographic work here longer, get promoted faster, and earn more. Women may take a year out every couple of years. Minority groups may leave very quickly, so they become less worthy investments.
Institutional memory, however, will tell you what the data cannot. That you created environments where women and minority groups could not be successful, and that was entirely independent from the talent or potential of those individuals. It might tell you people left because of a toxic manager, not because they were unable to fulfill the job.
The data will never by itself give you the richness of that understanding. As a result, the AI will give you the wrong answer because it does not have the required data from documented quantitative data alone to make the right inferences.
It is the classic rubbish in, rubbish out. Inevitably, this results in a game of broken telephone, whether between two humans, a human and an AI, or an AI and your database.
That’s why institutional memory, even in the age of technology and automation, boosts productivity rather than dragging it down. It is often the missing ingredient that organizations seeking to innovate overlook, and therefore do not have the full quantity of data they need to be truly data-driven in their decision-making.
In our rush to get employees excited about the future, we miss the power of institutional memory as a driver of culture and change. Only people who have been in your organization for some period of time will understand the journey you’ve already been on and the limitations you have, both spoken and unspoken. The question to ask is: what about the past do we need to understand to prepare for the future? Without an answer, you risk making expensive choices with the ‘start from scratch’ strategy for your next transformation. Better to treat institutional memory as the contextual and cultural data it is, and make sure you’re not throwing away data you’ll desperately need six months later.
