On March 2, Taiwan Semiconductor Manufacturing Co. chairman and CEO Morris Chang called an extraordinary board meeting to announce a plan to find a successor – or successors – for his CEO post.

The move came two years and nine months after Chang reassumed his position as TSMC CEO during the global financial crisis, replacing Rick Tsai, who had been named CEO in 2005.

This time, Chang did not tap a single individual as his successor. Instead, he handpicked three "co-chief operating officers": Shang-yi Chiang, executive vice president for Research and Development; Mark Liu, executive vice president for Operations; and C.C. Wei, executive vice president for Business Development from the company's research and development, operations, and business development divisions.

Their responsibilities will be rotated to expand their range of capabilities, and ultimately, Chang told the board, one or two of them will take over as chief executive.

Members of the board were taken aback when they heard Chang's plan. "Some foreigners were shocked," said one board member in describing the scene.

But Chang stressed that the move was actually part of a long-term talent cultivation plan, saying that when he returned as CEO in 2009, he set a timetable for himself to hand over the reins as CEO within three to five years.

"I have two batons in my hands," he said at a press conference on March 5, held to publicly explain the planned transition. "One is as chairman, the other as CEO. I will not hand over my baton as chairman. I will continue in that role," he said categorically.

Chang also made it clear that this phased-in succession approach was his "innovation."

How does he know this "innovation" will succeed? "Because I know this company, I know this industry, I know these people, and I know this idea will succeed and this approach will be effective," Chang said.

The same day as the press conference, Chang sat down for an interview with CommonWealth Magazine and talked about his innovative succession model in more detail.

---

Q: Why make the announcement now? Was there any particular reason why you gave yourself three to five years to develop a better succession plan?

A: Why was it five years? That's all I wanted to give this company. Honestly, I was already thinking about pulling out in 2005, of gradually moving away from the company. But in 2009, I felt I should come back for "one last hurrah." This will be the title of one of the chapters in the next volume of my autobiography.

At that time (2009) I was 78 years old. When I came back to become CEO again and make my last hurrah, I had thought it through and decided it would be for five years. I have other interests, including writing the next volume of my autobiography. I don't have as much time to read as I did when I wasn't CEO, and now I almost never play bridge, not to mention travel. Those are all things I still want to do in my life.

Q: Why did you choose these three people? I recall that in a CommonWealth Magazine article last year about TSMC's five big stars, you said that TSMC had even more heavyweights. So why was it those three? What kind of qualifications do you think a successor needs to have?

A: The real answer is complicated. First, they don't just have engineering expertise. They also have the strength to lead in their respective areas. It's not just a matter of specialized knowledge and experience, but rather having leadership strength in this company and in this circle. The biggest difference between being CEO and COO is that you have to devise strategies, strategies that are for the company as a whole rather than for more specific functions. A company can only have one strategy, and these three people have to cultivate that skill. 

Why these three? These three have this kind of potential, and they have a bit more potential than others. I pretty much tied the three of them together to strategically plan for the company as a whole rather than for any specific area. Of course, the three have to have good relationships with the chairman, based on trust and, even more importantly, loyalty, because this still involves interpersonal relationships.

Furthermore, some people are bolder when it comes to talking to the boss. I believe this quality is very important. I wanted to choose people who can closely interact with me to serve as co-COO, people who can work well with me day in and day out.

Q: You talk of "tying them together," to the point of working in the same office. How will that work?

A: For example, I want to talk to them now about something, and things like this happen every day. If I receive an e-mail or a phone call, and I need to see them, two of them are on the fourth floor and Shang-yi Chiang, the executive vice president of research and development, is a 10-minute walk away. There are many things that are much better discussed face-to-face.

Quite often, I want to say something to more than one person. Sometimes, I also want to tell Lora (Chief Financial Officer Lora Ho) or Dick (General Counsel Richard Thurston), and I have to walk 20 or 30 steps to their offices. To make sure that they're in, I also need to call first.

That's why I have an open door policy. Unless you have something personal to discuss, I don't want to see one person alone with the door closed. I don't like that.

I laid down an open door policy. The three COOs are bound together and each one has to respect the open door policy. Among ourselves, we all have immediate access. Even if they are talking with a customer, if I want to speak with them, they have to take a minute from their meeting to talk to me.

Q: Is that because you're hoping to react to things more quickly?

A: Our reactions are faster, and we work more closely together. In the past, there would often be a particular situation where if Shang-yi Chiang, who's responsible for R&D, and Mark Liu, who's responsible for operations, needed to discuss something, Liu might set it aside and wait until he ran into Chiang before bringing it up. Now, he will bring it up immediately.

Q: Does that also help cultivate the ability to make decisions independently in second-tier people or react more quickly to external factors? 

A: You have to foster more independence among your second-tier managers. If they ask a co-COO about everything, that's too much and goes beyond oversight.

That not only helps us react more quickly, but allows us to do more. It used to be that everything had to come from the top. I hope that the 20-plus managers directly under the CEO can take more initiative.

Q: While you still retain control over legal and financial affairs, you have deliberately ceded authority over personnel decisions to the new succession team. Is there a special purpose in this arrangement?

A: In most cases, the legal and financial divisions report to the CEO. When the three take over as CEO, those divisions will report to them too.

Also, what people on the outside haven't yet considered is that if the plan fails, I will still go outside the company to hire a CEO.

Q: How will you define whether the succession is a success or a failure?

A: If it fails, you'll know it. I will know success when I see it. I will know failure when I see it, too. There is an example in the United States where someone asked a Supreme Court justice, "How do you define pornography?" He answered, "I will know it when I see it." The definition of success and failure cannot be too rigid.

Q: What would you consider to be a relatively good succession model?

A: There is no absolute answer. The succession process at GE is not a model. It has a 10 to 20 person list of candidates that then gets narrowed to a 5-6 person shortlist and is announced.

Those five people then participate in the succession process until ultimately only one person is selected. But what actually happens? The other four or five people all leave GE in the following year or two. That's a huge waste of talent. For the sake of being able to choose only one final candidate, four or five really top people have left the company. That's not a good succession.

Q: What is the difference between you choosing three people and the GE model? Won't the other people want to leave?

A: These three are respected by everybody. Of course, at the end, if only one person is chosen, there could be a bit of a problem. If two people are chosen, there won't be any problems, and I can give the third one a seat on the board of directors. 

Another problem is if something suddenly happens to me, then what? My vice chairman (F.C. Tseng) would become the acting chairman and hold a board meeting. The first order of business would be to select a CEO. Next would be to choose a chairman. Maybe it would be him, but that would be up to the board to decide.

Q: The co-CEO model adopted by RIM has recently been proved a failure. How do you see it?

A: That was because they didn't have a chairman.

Q: The three successors need to form a trinity. What indicators will you use to assess it? 

A: P&L and market share are indicators. Even more importantly, we have quantitative indicators and there are also technology-related strategic plans that we'll see if they can meet. Then, there are client satisfaction surveys done through a third party consultant that will tell us whether these three are providing very good service. The stock price is also an indicator.

Translated from the Chinese by Luke Sabatier