And the open innovation platform takes all of this intellectual property, both TSMC’s and the third parties’ IP, and lets them test it and validate it. When customers use these processes, TSMC will guarantee that they’ll get their chip through production the first time without any mistakes. Errors can add significant costs, and can slow production by weeks or months. So TSMC really provides some peace of mind and assurance, plus a faster time-to-market. The whole system is a way of aggregating third-party and internal IP alongside the manufacturing services.
There are quite a few other companies providing foundry services now, but TSMC still has about half of the market by market share. When you’ve got a great new process, tool, or reference design, usually the first place you would go is the biggest guy in the market. So this is a case where the big do get bigger, and it becomes a very powerful advantage.
S+B: In other words, being first with their open innovation model gave TSMC a hard-to-beat competitive advantage.
CHESBROUGH: That’s right. They brought the foundry model into the market I think in 1987 or thereabouts. But since then, a lot of other companies have come into that market as well, so they have to keep that advantage by continuing to move forward. TSMC is now raising the ante by offering to test and validate the assortment of tools and processes, many of which come from outside companies, that go into the chips’ designs. But because of their testing and validation, you the customer don’t have to worry about how to make all this work for you; they’ve done that work for you.
It’s an example of how you really can build a sustainable advantage even in a services business. The normal knock on services is that these things don’t have a lot of intellectual property protection. Patents are less commonly used in services, and services are not always standardized — there’s always subjectivity. And any of us can undergo the experience and then reproduce it, so services are easy to copy. TSMC is an example of how IP can create and capture value, even in a services context. By developing its own IP for designing and verifying chips for its customers, and creating published interfaces and validation for third-party designs, TSMC has been able to build an enduring advantage.
S+B: Another example you give of excellence in service innovation is Xerox. This might be surprising to some people, since Xerox is most often cited as a company that lost its way as an innovator.
CHESBROUGH: I’ve been studying Xerox for more than 12 years, and I have documented in some detail some of their challenges, particularly those involving the Palo Alto Research Center. Their technology spin-offs, like the graphic user interface, [created] a lot of value for other companies like Microsoft and Apple, but not much for Xerox itself. Those are the stories you’re referring to. But one thing that I’ve learned from Xerox is that a big, established company can generate a fundamental business model innovation. For the first 40 years of its life, Xerox sold its copiers and printers as products. Sure, it charged for toner, paper, and service, and also for financing the purchase of its products, but those items were secondary to its business model (except eventually for toner, which became a key source of profits).
Today, however, it has an entirely different model: Instead of selling copiers and printers, it is offering a service it calls Managed Print Services, in which you pay only for what you use, by the copy or by the page. Xerox takes over all the management and maintenance of the equipment and keeps all the toner stocked, the paper in the machines, and all the rest. It’s a very different way for its customers to obtain copies. It’s a deceptively simple idea, but it’s devilishly difficult to execute.