At a growth hacker conference, one slide made the entire audience fall silent for three seconds.

It was a digital transformation stage diagram. Five stages, from “sensing threat” to “successfully establishing a digital experience platform,” drawn clean and neat, with every arrow pointing forward, as if success was guaranteed by simply following the steps.

Looking at that diagram, I didn’t think of success stories—I thought of corpses.

In over a decade of consulting and entrepreneurship, I’ve witnessed at least a dozen companies launch digital transformations. How many completed the journey? Two. The rest mostly died before reaching stage two—not falling in battle, but bleeding out internally, killing themselves from within.

If we compare transformation to labor, those five stages on the slide represent full dilation. The problem is: until you reach full dilation and the baby is born crying, nothing counts as success. Most companies I’ve observed called it quits by the time they reached two centimeters of dilation—the pain was too much to bear.

The First Cut: Mindset Transformation Equals Blood Transfusion

The biggest lie about digital transformation is: “Let’s implement some new tools.”

It’s not about tools. It’s about minds.

Reading inspiring transformation articles is completely different from actually driving change within an organization. It’s like watching medical documentaries versus performing surgery.

I witnessed one case. The owner of a traditional manufacturing company got hit by the phrase “digital transformation” at a forum and returned with grand ambitions, declaring: “We’re going to become a digital enterprise!” Then what? He had the IT department purchase an ERP system, spent six months implementing it, only to have the veteran workshop masters refuse to use it. These masters had managed the production line their own way for twenty years, with every parameter stored in their heads. Now you want them to input all this into an interface they don’t understand? Their response was: “I’ve used my method for twenty years—have there been quality issues?”

No. Their methods were indeed effective. The problem was that their methods couldn’t be scaled, couldn’t be passed down, couldn’t be AI-optimized. But how do you tell someone who’s performed well for twenty years: “Your method needs to change”?

This is the pain of mindset transformation. It’s not as simple as switching systems—it’s like a blood transfusion. Changing type A blood to type B can be fatal. What’s even more terrifying is that not transforming quickly won’t cause your company to collapse immediately. It’s death by a thousand cuts—a slow process of losing competitiveness that most people don’t feel. By the time they notice, it’s usually too late.

The Second Cut: The Desert of Digital Talent

Transformation requires talent. But what kind of talent?

Not programmers who can code—there are plenty of those in the market. What’s needed is a rarer species: people who understand both industry and digital, while also having sufficient management capabilities and political acumen to drive change within organizations.

How scarce is this talent? Among the people I know, those who possess all three capabilities number no more than twenty.

Where’s the problem? Traditional education.

Our education system produces specialists—you understand technology but not business, or you understand business but not technology. To understand both sides well enough to make decisions at the executive level requires “continuous learning capacity.” But traditional education makes many people think that getting a diploma is the end of learning. The result: many people’s knowledge frameworks stop updating after graduation, but the world doesn’t wait for them.

I discussed this issue in “Super Learners: The Learning Revolution in the AI Era.” Learning isn’t a one-time event—it’s a continuously operating system. But in Taiwan’s corporate culture, “learning” is often equated with “attending training courses”—as if completing a course means mastery, like taking medicine means getting well.

Digital transformation doesn’t need people who’ve taken courses—it needs people who live in digital thinking. The gap between these two is like the difference between someone who’s watched swimming tutorials and someone who’s in the water every day.

The Third Cut: Assassination by Power Structures

This is the least publicly discussed cut, but often the fatal one.

Every digital transformation means power redistribution.

New digital systems mean more transparent information—data that only a department head knew before is now visible to everyone. This is good for the organization, but for that department head, it’s a loss of power. Much of his value was built on the foundation of “only I know this information.” Now you want to tear down this foundation?

I witnessed a real case. A company established a digital innovation group reporting directly to the general manager. The group’s mission was to promote data-driven decision processes. Sounds wonderful, right?

The result: the sales department refused to share customer data, citing “customer relationships are sensitive information.” The finance department questioned the new system’s security, demanding six months of evaluation. The IT department said the new system was incompatible with existing architecture, requiring major modifications. HR worried that new processes would affect the fairness of employee performance evaluations.

Each objection sounded reasonable. But taken together, you realize this isn’t a technical problem—it’s a war to defend power. Vested interests won’t openly oppose transformation (that would be too obvious); they’ll use various “reasonable concerns” to delay, dilute, and ultimately eliminate any change that threatens their position.

This is what I call “assassination by power structures.” It’s not a killing blow—it’s chronic poison.

The AI Era: Same Pit, Deeper Hole

When I first organized these thoughts in 2019, “digital transformation” was the hottest keyword. Six years later today, the keyword has changed to “AI transformation.”

But the underlying problems are exactly the same, even worse.

Mindset transformation? AI demands not just “learning to use new tools,” but reconceptualizing “what work should be done by humans.” This touches fundamentals more than previous transformations—you’re not just changing how work is done, you’re changing the definition of work itself.

Talent scarcity? People who can simultaneously understand AI capability boundaries, industry needs, and organizational politics are even rarer than during the digital transformation era. Because AI capabilities are rapidly changing, judgments made six months ago might be completely obsolete six months later. You don’t need experts—you need generalists who learn continuously.

Power assassination? AI makes information more transparent, decisions more traceable, and performance more quantifiable. This means those who maintain their position through “information asymmetry” face greater threats than before. Their resistance will be more intense.

In pushing AI adoption at my own company, I’ve deeply experienced this. The biggest resistance was never technical—technology can always be solved. The biggest resistance was people. Those who feel “AI will make my experience worthless,” those who worry “if all data becomes transparent, how do I maintain my authority.”

Won’t Die Suddenly, But Will Slowly Disappear

This is the most brutal part.

Not doing digital transformation won’t make your company collapse tomorrow. It will make your company irrelevant in five years. Like a person who doesn’t exercise won’t get sick tomorrow, but ten years later at a health checkup, all the numbers will be in the red zone.

This characteristic of “chronic death” is transformation’s greatest enemy. Because humans have extremely poor perception of gradual change. The frog doesn’t not know the water is heating up—the heating rate is just below its perception threshold.

I’m not sure we’re living in the best of times. But I can be certain this is an era of change and brutality. Not just traditional industries—even Web 1.0 internet companies that were once disruptors will be abandoned if they refuse to continuously evolve.

Environmental progress is ruthless and rapid. But people in organizations, when facing change, have the same fears and resistance as our ancestors thousands of years ago.

This contradiction is the eternal labor pain of digital transformation. It won’t disappear because of technological progress. Because the problem was never technology. The problem is people. And people are the hardest system in the world to upgrade.