by Phil LoPiccolo, Editor-in-Chief
In a revealing keynote address at SEMICON West last week, Micron Technology chairman and CEO Steve Appleton offered a behind-the-scenes look at his company’s experience in forming business partnerships, and offered some hard-won insight into why partnerships fail and what can be done to ensure their success.
According to Appleton, business relationships are a lot like marriages, in that the majority of them do not succeed. To the many laws governing the science and business of semiconductor manufacturing, we must now add one more, he suggested — “the Law of Partnerships,'” which states that 40%-50% of all business partnerships, and 70% of all acquisitions, will fail. Appleton citied the extensive experience Micron has had over the years in building partnerships, most notably with Intel and Toshiba, as well as with lithography, etch, and materials suppliers and customers.
There are a number of reasons why partnerships fall short, according to Appleton, who offered advice on how to avoid common pitfalls. One big mistake companies make is going into partnerships thinking they are going to change their partner, and as in marriages, that rarely happens, he said. While it’s true that companies do change over time, such as when transitioning from a start up to a mature company, these changes need to be allowed to happen internally, he explained.
Another reason partnerships fail is because expectations are not set correctly, Appleton said. As an example, he pointed to Micron’s operation in Lehi, Utah, which after completion was scaled back to a test facility because of collapsing market conditions. “Despite the fact that we spent a billion dollars on the facility and hired hundreds of people for the test operation, we were very heavily criticized by the people of Utah and the media,” he said.
To avoid such issues, expectations have to be well designed and understood, Appleton asserted. For Micron’s new joint venture with Intel, IM Flash Technologies (IMFT), the two companies spent months negotiating the details of the project, which as a result is governed by 43 contracts, each of which totals some 200-300 pages. “Part of this is a case of attorneys gone wild,” he said, “but the contracts describe the expectations and goals all the way from the business plans to the investments to the timing, so that we have a good understanding of what the expectations are on their part and on our part. And that’s critical to making sure you have a good partnership.”
Partnerships also fail when the goals of the participants are not aligned. Appleton cited an agreement from the 1980s in which Intel invested about a quarter of a billion dollars in Micron, and ended up making a 4X return on investment — but still was disappointed because only six or seven of its 10 goals were accomplished. Fast-forward to a few years ago, when Intel made a similar investment in Micron, and the goals of both companies were completely aligned, related to developing next-generation technology, bringing it to market, investing in advanced capacity, etc. Intel didn’t make nearly as much money, but they were a lot happier about the partnership, and it ultimately led to the duo’s new all-out IMFT joint venture announced in January to manufacture NAND flash memory products, Appleton said.
Speaking about the new JV, Appleton noted the tool-installation schedule calls for the implementation of some 100 tools/month, which means that some 4000 training events must occur prior to full production. “We have to work together, or we can’t possibly get this accomplished,” he added.
Forming partnerships wasn’t always a priority for Micron, Appleton admitted, but he said it’s now one of the key objectives for improving the value of the company. He noted that Micron is now working to implement 35nm technology, and that partnerships have been the key to enabling them to “to get this far this fast.” Moreover, Micron is extending its partnerships further. During SEMICON West, the company announced that it has joined IMEC to conduct research on sub-32nm CMOS and advanced flash memory.
“Partnerships are difficult, to say the least,” concluded Appleton, “but they are necessary, and if done correctly for the right reasons, they can lead to a tremendous benefit. We’ve proven that they can work — and we’ve proven that they can’t work if you don’t have the right approach.” — P.L.