Trending
Written by: Peter Laughter
The efficiency of business is a myth. Leaders have made peace with projects failing and have resigned themselves to unproductive strategies.
70% failure of corporate transformations; business is not a system that can easily adapt.
And as we are in the Age of Acceleration, adaptation is a vital component of success, but that doesn’t mean we’ve spontaneously developed the ability to make that happen.
AI has arrived, promising to fix all of our problems, but the rate of stagnation has only gotten worse. According to MIT, 95% of GenAI pilots deliver no measurable return.
The truth is, business is not efficient and technology isn’t magic. As my friend Ed Hansen says: “It’s people, process, and technology, in that order. Always.”
A 70% failure rate on corporate transformations. That statistic has existed for over twenty years, its been equally depressing the entire time. But why?
This is a problem intrinsic to our very idea of leadership. Our traditional business structure corrodes communication and stifles information, making it almost impossible for a corporate transformation to actually reflect the needs of the organization.
Leaders make assumptions without full knowledge of the situation; the leader starts with their own picture instead of co-creating it with the people doing the work; silos optimize for themselves, not the greater organization.
What’s more, people with higher status are the ones who are listened to. Information flows down, not up, so the observations made by those with lesser status are dismissed. Often times, those lower status employees are the ones coming face to face with the problem, but their wisdom and hands own experience has no where to go. So there are all of these problems, that never move up the ladder and are never heard, instead they get stuck, they linger. And this process repeats over and over, stifling communication and productivity.
Sidney Yoshida referred to this as the Iceberg of Ignorance. Leaders only see a sliver of the problems their organizations face.
So when we are thinking about that 70% failure rate of corporate transformations, we have to look at that with the awareness that businesses are not set up to support efficient communication.
Corporate transformations are intricate — mergers, digital transformations, re-organizations, to large scale software upgrades — if the leader spearheading the transformation doesn’t understand the full scope of the problem they are trying to solve, how on earth could they be successful? So problems don’t get solved and most of the time they are hardly even addressed. Often, the problems that are solved are merely symptoms of something larger.
Furthermore, we’ve optimized for rapid decision making and the result is often a misunderstanding of what needs to be done. Leaders who think they know best when they really don’t, and so decisions get made without critical pieces of information.
Do we think these problems will go away when things grow more complex? Absolutely not
While 70% corporate initiatives fail, 95% of AI implementations go nowhere.
And no, these numbers are not directly measuring the same exact thing, but this tells us that both corporate transformations and AI implementations are struggling. What is an AI implementation if not another kind of corporate transformation? If anything, it’s expensive; everyone is pouring money into AI right now, we’ve bound ourselves to it, and yet we are not seeing results.
So what’s going on? why isn’t AI working for us?
The problem lies in a fundamental misunderstanding what AI is; leaders treat it as a cost reduction tool, when in fact it is a tool to enhance human capabilities. Dana Grundy at @PredictAP says it well, describing that AI is “an amplifier. And amplifiers are only as effective as what they are amplifying. If you apply AI to a chaotic process, you get automated chaos. If you apply it to an ill-defined goal, you get very sophisticated confusion” (link).
When AI is being added to a situation without a clear understanding of what it is enhancing, failure is no surprise.
Meta for example, said they weren’t trying to replace people. Then they did replace people with a bot and it failed miserably. Meta created an AI bot for “low level tech support issues” and hackers went to town using it to pose as people locked out of their accounts. They then hacked into accounts like Sephora and the defunct Obama White House Instagram page, and more.
Don't you think that if Meta had bothered to incorporate the people they were trying to replace into their process, that they would have uncovered this ridiculously stupid oversight?
That idea of AI as a cost cutter is pervasive, and frankly dangerous. Employees know when AI is aimed at cutting their jobs, so they quietly undermine it. About 29% of employees admit to sabotaging AI strategy (44% of them being Gen Z) and 60% of execs plan to lay off non-adopters. (WRITER.com - April 7, 2026)
This misguided interpretation of AI causes massive friction that reverberates throughout the entire company. It’s a significant sense of conflict — people can tell that their companies want to replace them, and they don’t like it, so they fight it.
Command and control is about holding onto power, AI is being used to enhance that sense of control. But what if it wasn’t?
We’re using AI for the wrong things; we are just creating more stuff, not better stuff.
We’ve all experienced the flood of AI slop that sales people are throwing at us, its unbearable. And compare that to sales people who are using AI to understand us better, to find better ways to communicate and connect with us. Those enlightened sales people are using AI to expand their human connection, instead of having one person do the work of five (badly).
What if Meta had looked to make their tech support workers more efficient with AI instead of just replacing them? What if the process to make the work more efficient was driven by the workers themselves?
With AI implementation, intent is critical for people to buy in. Everyone in the organization knows when the intent is to save money, instead of enhancing productivity.
Imagine the possibilities if the intent was to use AI to enable people to be more effective at fulfilling the purpose of the organization.
People need to see that they have agency in this transformation.
Leadership does not know what is going to be most effective, they don’t see everything and they don’t need to, they need to create structures for others to get work done well.
We don’t know what is going to come out of AI, leaders don’t know and honestly none of us do. However, the people who are seeing the impacts of these emergent technologies are the ppl on the frontlines. And if decisions making remains relegated to the top few, you lose out on emergence.
It is critical when looking at AI implementations that we turn to distributed leadership, the process of pushing power, authority, and decision-making to the front lines of the organization
Yet, when ppl hear distributed leadership they hear chaos, but that’s not what I’m talking about.
It’s an understandable confusion. There are processes that require order and focus: a nuclear power plant, a complex factory floor, a civic emergency, a military campaign.
But distributed leadership isn’t antithetical to structure: plenty of organizations have thriving distributive leadership systems in a hierarchical organization. It's not about the structure. It's about what we bring to the structure. It's about our concept and application of power, authority, and control
Many distributed leadership organizations are secretly failing to escape from a command and control mindset. Knowingly or unknowingly, their intent is coming from a place of control.
If the leader has the intent of being the one with the all of answers, they are not listening or seeing the realities around them.
Within the very concept of leadership that we all grew up in, is the belief that the leader is the one with the all of answers. Once you are inculcated in this paradigm, it’s tough to get out. It is immensely difficult for leaders who are immersed in the command and control orthodoxy to listen and see the realities around them.
I’ve seen many adaptive leaders fall into this trap. I have fallen into it myself.
The solution for this, is allowing leaders to see what their teams are capable of creating on their own.
The solution is to evolve our intent toward supporting your team and waiting for answers to emerge. Many leaders have been successful because of this exact idea that they can see solutions, and they probably have. This mindset has got them very far and brought them success, but our world is too complex to think like that anymore.
This is emergent technology — no one has the answers.
We need to shift our intent to distributing agency, so that when a problem or rare opportunity emerges, everyone who witnesses it has the freedom to act. The hard fact is, leadership is not a rare commodity at the top. It is effervescent and all around us. It is the job of modern leaders to create the conditions for that leadership to freely flow and create value in the moment that it's needed
We have to start by having conversations.
And we need to include everyone, especially those closest to the problem.
I think we should approach AI implementation with an appreciative inquiry, in the style of Cynefin and Strategic Doing. We need to ask questions big enough to inspire the entire team.
"What would be possible if we could solve our customers' problems faster and improve their lives using generative AI?"
"How might we strip the obstacles and frustration out of our work so we can have more energy to take care of, ourselves, our colleagues and all the other people we serve?"
Then we experiments.
Because leadership was never a seat at the top. Everyone leads something.
The future is moving too fast; if you have people who are waiting to be told, everyone will be left behind.
So go forth and do great things… together.