During the week of SEMICON Taiwan, I was frantically working on the latest edition of the print version of Semiconductor Fabtech, when a simple chart appeared in an email from VLSI Research, the market research firm. Little covering text was included, other than highlighting customer satisfaction poll for equipment suppliers to fabs in Taiwan. How helpful?

(Update: what wasn't helpful was a posting error, full text now available).

At first I was just too busy to do anything much with it but on returning to the chart, I noticed that a company name I would have expected to be on that list, was missing!

That name was Lam Research. As Lam is a major etch supplier and in particular does a lot of business with memory fabs, as well as the fact that in Taiwan, there are a lot of memory manufacturers, the missing Lam must surely have been a mistake?

On calling VLSI Research to check the ‘obvious mistake' and then them coming back after checking the story, it got a bit more curious!

 Basically, VLSI Research didn't make a mistake and the reason why Lam is not in the TOP 20 customer satisfaction survey is because they came in at number 28!

That's pretty awful right?

However, it should be clearly noted here that the VLSI Research annual survey has caused problems for Applied Materials in the past as its rankings were going down year-on-year. Financial analysts picked up on this for a while, forcing executives to make it clear that the company did not support the independent surveys and that it undertook its own research, which was considered more accurate. Maybe I should add in that it also gave a higher level of satisfaction than the VLSI Research survey, as I remember past Applied executive David Wang explaining this to the analysts in conference calls. On contacting Lam directly, they also do not directly participate in the VLSI Research survey.

Without getting bogged down in an explanation of the variabilities involved in this type of survey, a few qualifications are in order.

Firstly, it doesn't do an equipment supplier any harm, should they actually participate in the survey, to get in touch with as many of its customers as possible and urge them to fill in the boxes. This is very true of small equipment companies with much fewer customers than a Lam or Applied. Though the comments sections and I would hope the rankings aspect can't be influenced that much, it could be expected that with many more responses than a non-participating company, the rankings may come in a bit better.

Indeed, a small company actually got the highest rankings for Taiwan if you look at the chart below. With large companies and a larger installed tool base, you could debate the fact that they may well do worse than a company with only a handful of tools in the field!

So this and many more types of ‘skewing' can be a play. However, in the past I have had access to the survey findings from companies that did participate, when I have written more detailed articles on the findings and associated trends. The summarized comments from chip manufacturers can be interesting reading and there can be a good level of honesty in the responses. Companies that have regularly participated in the VLSI Research survey have found this to be very helpful and take involvement very seriously.

So why is Lam ranked so low?

Obviously some customers of Lam did fill in the forms and made comments. Considering the research firm posts out over 40,000 of them, Lam isn't going to be missed out. VLSI Research was emphatic that the rankings for Lam are what came back and are accurate from that perspective.

Lam didn't want to discuss the survey findings with me as they felt that was improper, as they do not participate in it. That is fair enough, but I was hoping they would have given some alternative information that showed different customer satisfaction data that they must be doing in-house.

With Lam doing exceedingly well in the current up-turn and very much a focused company that tends to innovate around its core technology, I am sure that customers are very happy with them. After all, they wouldn't still be buying from them otherwise, right?

That point should also be seen in the context that Lam competes in the etch market against Applied and TEL, the two largest equipment suppliers, so you can't gain market share against such rivals if your customers don't think your tools are good!

With that stated, it is interesting that Lam is missing from the TOP 20 in a location in which it sells a lot of tools. It is also not in the TOP 10 rankings overall from VLSI Research. Any help in understanding this further would be greatly appreciated!

Strangely, while doing research on a new product from Straatum for the latest edition, I stumbled across a brief ‘case study' (number 3) at Straatum's website (http://www.straatum.com/hdocs/case.htm) that describes a problem at a leading foundry with a Lam etcher.

Now, Lady Luck would have it that this could have been the only customer to fill in the VLSI Research survey on Lam!!