Exiting the semiconductor diffusion furnace and writing off about ¥5 billion (1992)

What it means to shut down a product that works

It is not enough to read this withdrawal simply as a failed diversification outside the core. The furnace was technically well regarded, and Sekiya Kenichi himself accepted that as a product it had succeeded. With 95% of the domestic back-end process already taken, there was little headroom left there, and the front end was one of the few next markets available. This was a decision to fold not an unsellable product but a working one. What made it possible, it appears, was that the son who had inherited the company sat in the one position from which he could name the person who had decided to enter — the founder, his father Sekiya Mitsuo.

That said, the judgement was not fast. Development ran ten years, and the roughly ¥5 billion spent was money already gone by the time of the exit. The clincher, too, lay less in an argument about base technology than on the cash-flow side — that carrying on would leave the company exposed. He spoke of having strayed from that base only in 1995, two years after the exit, and the business domain was put into words as “cut, grind, polish” in 1997. A lesson, one could say, begins to work not right after the failure but at the moment it is restated.