Interview with Dennis Carter
April 20, 2004
Mountain View, California

RW: Dennis Carter was Vice President and Director of Marketing at Intel in the '90s, when it grew from a component supplier to one of the world's most widely recognized consumer brands in the world. In this 2004 interview, he recounts his four years as Technical Assistant to Andy Grove, and what he learned about Grove's management style and drive. He then initiated and managed Intel's successful Graffiti Program to encourage upgrades from the 286 to 386 and the 486, both increasing the company's margin and segment share of the microprocessor market. He went on to create and manage the Intel Inside Program that made Intel a household name alongside BMW and Coke.

RW: Okay, Dennis, so tell us about your early days.

DC: I grew up in Louisville, Kentucky. And went to school at Rose-Hulman Institute of Technology, and majored in electrical engineering and physics. Really wanted to major in physics but ended up choosing more pragmatic electrical engineering approach. And then went to grad school at Purdue. And also did electrical engineering there. And then went to work.

RW: Well now, what was your family like? Your - your mom and dad, and so on?

DC: Well, I - I grew up with my mother. My parents were divorced and I never really knew my father. And very - very strong close-knit family. Lots of cousins and grandparents and - so that was kind of it.

RW: So were you the, the only child?

DC: I was the only child. That's right. I didn't have any brothers or sisters.

RW: Mm-mmm. Okay. So what made you go into electrical engineering?

DC: Well, as I said, I - I was really interested in physics. That was what I wanted to, to major in. Actually astrophysics. I was very interested in astronomy and those kinds of things but also pragmatic in looking at the potential of getting a job. It didn't look real promising going that direction so electrical engineering was a nice solid career. So that was why I chose that. Plus it was very interesting at the time, too and sort of related to physics. And turned out to be a really good choice because the whole digital revolution was just starting about the time that I was doing this. And that changed it and made it - made it very interesting.

RW: So what was your first job?

DC: My first job was working at, at Rockwell. Was - had been in Collins Radio. It was located in Cedar Rapids, Iowa, and a - an entrepreneur, Art Collins, had out of World War II, had started a - a radio business. And he was - he was a brilliant designer. And this business had grown into be quite a large company. But like a lot of entrepreneurs it was - the engineering and - and the products were a lot better than the - than the business - than the - than the finances of the business. And so it came to be acquired by Rockwell just about the time that I joined it. And I worked there doing design engineering and designed systems for - for avionic systems for commercial airplanes. Worked on radar altimeters, writing 8080 assembly code, worked on transponders, and collision avoidance systems and things like that. Did a lot of microwave design, too.

RW: What was your opinion at that time of Intel?

DC: Well, Intel was - was - I first became familiar with Intel when I was in - still in college, and one of our professors got a hold of a 4004 data sheet, and we just memorized this. This was like the most exciting thing that ever happened on the planet. Because you could see the potential of what you could do with processors, and never actually got my hands on one of them, but - but just thinking about it was exciting. And then when I went to Rockwell we - we did - the group of new engineers that were hired in, we were all pretty conversant on - on digital technology. In fact at Purdue, I actually stayed around a little bit after I got my master's degree and - and put together the first digital course for the EE Department and taught it. And there was a whole generation of digitally adept engineers coming out of the colleges at that point. And so Rockwell was starting to design in digital technology into the - their avionics equipment, which had historically been all analog. And it really revolutionized what you could do in terms of the fail-safe and all kinds of new features you could put in and - and - and so forth. So it was really very exciting time to be there. And so my impression of Intel was that they were clearly one of the leaders. And - and I - I suppose I had a lot of early experience with Intel architecture and knew the - the 8080 and the 8085, and a little later the 8086, knew those processors and the - the - the instructions stats quite well. And so Intel was this great West Coast Company that I kind of always looked up to.

RW: So then you eventually went to Intel.

DC: Right. I eventually went back to business school. From being an engineer I got more and more involved in the - in the business side of things. And that - that was interesting. And I knew nothing about it. So I decided if I was going to do that it - it would be good to have a - a - some background in it. And so I went to Harvard Business School. And coming out of Harvard, got a job opportunity at Intel and came out here, my wife and I came out here thinking that - that we were kind of East Coast/Midwest people. And we thought we'd be here for a couple of years just to experience Silicon Valley and Intel and see what this was all about. But we never left.

RW: So what was your first job at Intel?

DC: At Intel, my first job was working on bubble memories. And I was Product Marketing Manager for Bubble Memories. And in those days marketing at Intel was very technical. The whole idea was that we would - we were marketing from a very technical engineering company to design engineers who were making design decisions, design engineers like I had been at Rockwell. And so in fact, the kind of people we hired for marketing were people like me, people who had a background in - in design engineering and knew what - knew what that was all about. So my first job was working on bubble memories which were - were kind of an exciting technology but came to nothing, and the reason it came to nothing was because so much of the world's energy and - and resources were being invested in silicon that no technology could keep up with that. And even though bubbles, which were based on a garnet-magnetic concept, even though they - they had some really good advantages, they just got washed away by the silicon tide. And so that was an interesting experience.

RW: Yes. So Intel shut it down.

DC: Eventually we did it. It lingered on for a few more years. It had some niche applications. But it wasn't - it didn't turn into the National and Intel and the other companies that were making bubble memories, quickly realized that this is not going to turn into the next great memory. It's going to be surpassed by - by silicon and - and by discs, and - and technologies like that. So from there I moved - from doing bubble memories, I moved to marketing microprocessors. And Intel was involved in the - in the big design-in program around the 286. And so I - I went to work for K.C. Powell and Jim Saxton and - and in doing working on the 286 marketing. And that was my second job at Intel.

RW: Did - was Dave House there at that time?

DC: Yeah, Dave was there. He was - I think he was running the microprocessor group at that point. I'm a little unclear in who was running what at that point, but yes, Dave was certainly there. Yeah. He was very charismatic. Very visible leader, yeah.

RW: But now had - Jack Carsten had moved on into ASIC?

DC: No, that - it was actually pre-ASIC, and Jack was still there and very involved in the - in the design-in program. We - when - I was involved in quite a few things on that - on that program in both the advertising and - in a - in a seminar series that we did actually worldwide. It was a full day, technical seminar talking about the 286 and all the peripheral products that supported it. And Jack and Dave both were very involved in reviewing the materials and reviewing the progress of the - of the - the program. Had to do a lot of presentations in front of Jack. So.

RW: That's always fun.

DC: It was very exciting.

RW: So was the 286 successful?

DC: The 286 did turn out to be quite successful. It was - it made it in the - the second round of personal computers. The first round of personal computers that IBM launched, and then lots of clones jumped in and - and replicated was based off the 8088. And the second round was based off the 286. And so the 286 became a - a - it was called the workhorse of the - of the PC industry in those days, and stayed around for a long, long time as sort of the main processor because they had a lot of advantages over the - the 8088 and the 8086.

RW: And it was multiply sourced, too.

DC: It was multiply sourced. There were - there were - there were quite a few companies that made the 286. Yeah.

RW: So what happened next?

DC: Well, from there, after - after that design in program was kind of completed and - and the 286 was successfully in the market, we - I moved to the Development Systems Operation. Development Systems was - I had used Intel Development Systems at Rockwell. And Development Systems were - were a complete turnkey package that Intel offered to help design engineers design in microprocessors. It allowed you to - to do in-circuit debug of - of the actual hardware and it allowed you to do the software development as well. We - so we had compilers and real-time operating systems and all kinds of different pieces that allowed engineers to design the processor and made it easy. And so I moved into - into that organization and - and ran the - the - the software marketing of that. And that was a very interesting period, too, because during that period, the - the PC that - that - that we were integral in with the 8088 and the 286, the PC itself was starting to - to take away quite a bit of our development systems business because our compilers and our software would run on PCs just as well as they would run in the - on - on the blue box, on the development system Intel sold. But the business was structured such that the blue box was where most of the - the - the margin came from that allowed us to continue to invest in the - in the software. So it was an interesting business challenge as that - as that whole area changed and modified and morphed. And so that was my second - or third job I guess at that point at Intel.

RW: Yes. The intellect systems used to be hugely profitable. And that gradually went away as you could buy a simple PC and - and do the same thing.

DC: Absolutely. Right. And so - and as well as that going away, then all - also lots of companies started offering compilers, too. One of the reasons Intel needed to do that in the - in the early days was because there wasn't a lot of sources of compilers for those early microprocessors. And so Intel invented its own compiler, PLM, which was a great compiler. Dan Lau invented that and it was really - really good technology. And - but it was necessary to do that. By the time that it – that we're talking about here, not only was - was the PC making a really low-cost platform so that development systems weren't really needed. But on top of that, companies like Borland and Microsoft were starting to take seriously the compiler business and offering, you know, serious compilers for the - for the microprocessor. So there was - there was a lot of changes in that business as well. Yeah.

RW: So - so that was winding down then.

DC: Well, it wasn't quite yet winding down. It was still at its peak, actually. But all of those changes were - they were there. They were evident. It was - it was - it was coming. The - the winding down was coming. Yeah.

RW: So what's next?

DC: Well, about that time Intel was heavily involved with IBM. And IBM had actually made an investment in Intel and - and because of our strategic position as the developers of the - the processors that IBM was using, IBM showed a lot of interest in Intel. And Andy Grove, who was President of Intel at the time, saw - spent a lot of time with IBM and saw that they had this program which was called a Technical Assistant Program, where executives at IBM would have technical assistants. And these technical assistants would be kind of gophers for the executive, but also they would be - it would be a - a training job where the - the lower-level person would get fairly high-level exposure and understanding of what's going on in the company. So it was - it was good both for the executive and the - the - the technical assistant. Well Andy decided he wanted to try that at Intel. And so he - he interviewed a few people, and he knew me from - from the - the processor designing days and presentations I had done and ended up hiring me for that job. And so that was my next assignment. And I - I did that - I worked as his technical assistant for - for about four years. And that was a - that was a wonderful experience and a - a really interesting time. Andy Grove is an amazing man. And the opportunity to work that closely with him and see - see him day to day and see his - his thought processes and watch him hands-on manage the company. That was a terrific experience for - for me at the time. Gave me great exposure to the industry and to the strategic problems into what was going on. So that was my next job at Intel.

RW: Well, what were the - what's - what's he like?

DC: As I said, Andy's an amazing man. He's a brilliant man. And that sort of - that's sort of his reputation, as being brilliant. And I - I found working for him really challenging and really rewarding. He has a unique ability to listen. He's - he's - he's very driven and very - has - has very strong opinions. And that also is very well known about him. But he also has an uncanny ability to actually listen. And that's a very rare skill, particularly among senior managers. And if you have a point or an idea and you manage to get it across to him, you manage to communicate it to him, you can - you can see him comprehend it and absorb it. And I think that's one of the reasons that - that made people so loyal to him, is that ability both to listen and - and take information that - that you're offering. You - you feel like you're being paid attention to. And so it was an amazing experience working for him.

RW: Well, he also is unique in that he's not big on perks. On the jet aircraft, on the limo, the - the big house, the…

DC: Right.

RW: …whatever.

DC: That's correct. There - there were no perks. In fact, one of my - an ironic event, when I was working in his - in his office as - as his TA, PCs were not yet ubiquitous inside companies. They were still kind of novelties. And I wanted a PC because I - I knew that there were things I could do with a personal computer on my desk that - that I couldn't without it. And to him that was perk. So it - it took a long time to get PCs into his office. When - when they finally came in, he was - he became a totally - addicted and drove e-mail through the company almost immediately, and - and later became addicted to the Internet. He was absolutely just - he's - he's a information hound and loves information. So when he adopted, he adopted very fast. But in those - in those days that was viewed as a perk. We didn't have parking places. We had very small offices. You know - there were no perks at Intel.

RW: Well, except stock options.

DC: Stock options - yes, Intel was very generous with stock options throughout the whole company. And so that - that was a good perk, and it was a perk where, you know, it was - it was positive for the company and positive for the individual because since everyone felt a vested interest in the company, it made everyone want to - want to work hard and - and focus on the - on the same results. So, yes, stock options I guess were a perk but…

RW: So, made a few millionaires?

DC: Yeah, I think Intel probably did make a few millionaires. Yeah, because the Silicon Valley was a - it was a good time to be in Silicon Valley from that, from a stock option point of view, I guess, during the - the '90s.

RW: Well, so you were working for Andy Grove. So what did you do there?

DC: Well, I did a lot of things. One of my tasks was to read his - to read all the mail coming into the Office of the President to try to reduce his load because he had a terrific flow of mail. And so I - I think the logic was that I would read the mail and maybe triage it and summarize it. I was supposed to make a summary of what the key points of this piece of mail was and put it on a little yellow sticker and send it to him, and then he would be able to assimilate the mail more quickly because he wouldn't have to necessarily read it all. Of course, being the kind of person that he is, being a stickler for detail and an information person, he still read all of the mail and he would read my little notes, and he would start dialogues back with me. He'd write little notes, write on my notes and we'd start exchanges. So it actually I think gave him more work rather than less work. But that was one of my jobs. Another job was working on speeches for him. Andy - Andy was very involved in his speeches. It's not like he would blindly write a speech and he would give it. It was a very interactive process. But we collaborated on - on his speeches. And I ended up picking up a - a lot of tasks during the years in that job, and - and by the end of that tenure was one of - one of the main things I was doing was working on the strategic planning process for the company at a - at a executive staff level, at least, and helping drive that process and come up with the questions that - that we needed to be asking about what was going on in the - in the environment and in the industry. And - so it turned out to be a very broad job. And when I was in business school, professor once gave the advice that you should always seek out a job where no one knows exactly what it is you're supposed to do, because then they don't know what you're not supposed to do either, and you can pretty much do, you know, whatever you find interesting and - and rewarding. And so that was that job. I was able to kind of - had a lot of tasks that I had to do, but beyond those tasks I could kind of go play around almost anywhere. I had carte blanche to play around almost anywhere. And so it was a very interesting, fun job. But eventually that - that time came to an end because strategically Intel was faced with a lot of issues in the late '80s. And working on the strategic planning process, I became really passionate about the marketing change that was going on in the world. The PC had changed everything, and we talked about some of those changes that had come along to Intel's business structure on compilers and development systems, and in the processor itself. But by the late '80s it was very clear that our market was changing, too. We - our whole marketing system was very design engineering, very technically focused engineer to engineer. And an example of this was the 386. When we launched the 386 microprocessor in the early '80s, we did it at a museum in San Francisco, and the audience was OEM's, representat - executives from different PC manufacturers, Compaq and IBM and companies like that. And they came to the launch and they heard about the new product we were introducing, and it was an exciting time. But by the time we got around to launching the 486, which happened in '89, those people already knew long before we would actually launch the product lots about the product. They had early design information and were already designing it into the next generations of computers. So the launch clearly wasn't targeted at them. And moreover, they're decisions weren't really at that point so - they weren't based on technical timing issues as they had been based in the past. They were based much more on what their market was demanding. Did they use a 486? Do they use a 386? Do they use a 286? It really was their marketplace that was telling them what they wanted, not the design engineer at those companies. So we're getting ready to launch the 486, and we were confused about who we were launching it to. So that was a strategic problem. And I sort of recognized that the reason that was a problem was because our marketplace had changed. And now who we needed to be marketing to were people who were buying PCs. We needed to be telling them about the microprocessor, what was important about the processor, why they would want a 486 instead of a 286 or 386. And we weren't geared up to do that at all. Our entire structure was oriented around design engineer to design engineer. So I became a - a strong advocate for that. And finally Andy sort of threw me out of his office, this technical assistant job, and said, yeah, go - go do this. Somebody needs to do it. We're not structured to do it. You believe in it. You go do it. And the first challenge that we had was actually related to the 386 itself. The 386 had been in the market for several years, and it was very successful but had more or less stalled in the marketplace vis-à-vis the 286. In market segment share, both were growing but both stayed relatively con - constant with one another, which was diff - excuse me - different than what had happened in the past, where a new processor would eventually push the old processor out. The 286 was rock solid in 1988. And so Intel introduced a low-cost version of the 386 call the 386 SX. And we thought that that would - would address that issue. But it didn't really address the issue at all. All it did was - was cannibalize the 386 DX, the more expensive, more robust 386. It cannibalized that. But it didn't really make inroads into the 286. From an engineering perspective, this made no sense to us because it was much more cost effective, or equally cost effective as the 286 and more powerful. We didn't understand why it wasn't moving into the market. But it was the strategic problem that the design decisions weren't being made by the OEMs, by the computer manufacturers, but rather were being made by the IT managers, and by individual power users who were making PC purchase decisions. So the first thing I did when - when I took this responsibility of marketing to these people or figuring out if we should, Larry Kerr, who ran MarCom at the time, and I went out and we did market research to find out why the 386 was sort of stalled and the 286 was so popular. And we found out that the general perception in the marketplace at that point and time was that the 286 was seen as the workhorse of the industry, perfectly adequate processor. The 386 was seen as incompatible with the 286, which it was not. And it was seen as something you would only buy for power users. And it was seen as quite expensive, even though at a processor level, or the system design level, that was no longer true. And so that was sort of a revelation, that the marketplace had not the same perspective on this that we did, and had actually misinformation, but - but it was really imbedded and nothing was going to happen to change that unless we did. And so I went back to Andy and I proposed that we take - undertake a marketing campaign to address that audience. And this would be an end user marketing campaign, which were not positive words inside Intel. And that goes back to earlier days at Intel when Intel briefly got into the digital watch business with Microma. And - we thought it was - at that time the management team thought it was the semiconductor business. And it wasn't, it was a - it was a design business, and a distribution business. But we acquired Microma which was a - a watch making - or a company that made a semiconductor for watches, and we thought we would get into the business and it didn't work. But during that process they made a television ad which never really got aired. It starred Arty - Arty Johnson from Laugh-In and - and it was a - quite a humorous ad. But it was really expensive and just didn't work, and it - it never made it on air because it was too expensive. And all - that whole process sort of left a bad taste in Intel management's mouth about end users. So it - was not something that was very supported. It wasn't a good idea to go talk to the end users. We talked to design engineers. So I went back and proposed that we go to the end users. And Andy said, "Well, you know, yeah, I buy your logic. It's good logic, and yes, you should go do that. We'll put aside an amount of money for a marketing campaign. We'll put aside five million dollars. But you can't spend that five million dollars because that's just too much to go spend on this because it's kind of not what we do. You go spend a tenth of that and prove that you can have an effect on the marketplace, and then come back and we'll talk about the rest of the money." So I put together a really small team. And the first thing was Larry and I went out the - and looked for an ad agency that could come up with a concept that addressed those issues that we had found about the 286 and 386. And we used a designer, Chip Shafer, who - brilliant man, who came up with a - a great ad concept. And Larry and I are sitting in a conference room, and he comes in and he - he puts up a - a big, bold 286. That's all that's on the page. Just black-type 286. And he takes out a spray paint - spray paint can, shakes it, puts a big X across it. He says, "This is what the ad's going to look like. 286 crossed out with a red X, grafittied out. We're going to show that the 286 is not the state of the art anymore. We're moving on." And by this time the room is filled with spray paint fumes. And we're, you know, like suffocating in the spray paint fumes. But then he puts up another page. And so what he - what he had done looks like this. And he puts up another page and it has 386 in smaller letters, and he spray paints across it SX, like this. And he says, "This is the ad. We'll do this on two facing pages. We'll cross out the 286 and we'll - we'll sell the 386 SX." And so we went with this concept. And so now the situation is it's approaching summer of 1989. And we have a very limited time in the market window to make the 386 SX successful, because the 486 is right behind it. And so we decide that we have to do a campaign for five hundred thousand dollars and we have to be able to demonstrate it had an impact. So I go recruit a team of just a handful of people, and I picked Denver because we could buy some media in Denver. And we go to Denver and we - because of the - because of the speed with which we needed to do this, we couldn't put the ads in magazines. We had to do them in very immediate media. And so we chose to do billboards. I don't know if this shows up on the camera - but we bought a series of billboards in the Denver area. And we visited IT managers, and we visited Business Lands which were the distribution channel for PCs at the time, and we put collateral material in and we did training. And we did a blitz campaign for six weeks. And we ran the ad in local newspapers. And we did market research in test markets - other places where there was no campaign. We did market research in Denver before we started, and we measured perceptions of 286 and 386. We measured buying plans. We looked at all kinds of very detailed data. We ran the campaign for six weeks, and then we did the follow-up research at the end of it. And the results were profound. The whole Denver market just dramatically shifted from planning on buying 286-based PCs to planning on buying 386-based PCs. And the - the event did not happen in the control markets where there was no campaign. So I went back with that data and showed it, and was told, yes, that looks like it was successful. Go spend the whole five million and make it happen across the country. Well, being an engineer you want to replicate an event that happened that you - that you knew worked, and so instead of rolling the campaign out nationally like you normally would with magazine ads, I decided to replicate as much as I could with the four and a half million we had left, to replicate the exact thing we had done in Denver. So we did billboards across the - the country. We picked nine cities - nine other cities, the largest PC consumers of the time, except New York City which we couldn't really afford to do. And we picked nine other cities, we ran billboards, we ran local print ads, we did training in the stores. And I had to use our field engineers to actually do that training that was very controversial inside the company because it was viewed as why are you wasting our resources on this? They need to be out with design engineers. But at the end, it turned out to be a very effective campaign, and the market toggled from 286 to 386 all across the country by the end of the year. And so we were off and running with marketing to this new organization of - this new audience rather, of IT managers and power users rather than design engineers. And so that was the beginning of our end user marketing.

RW: Well, it also have had the effect of shutting out the second sources who were in the 286. You only had, what, one 386 second source?

DC: Yeah. There were - I actually don't remember the details of who. IBM I think made 386s, and Harris I believe made a - a CMOS version of the 386. I'm not really sure who else was there. The interesting - the interesting strategy - one of the interesting business problems that Intel faced, and so by the end of this era we have several interesting business problems - one interesting business problem is that our strategy had always been to design the next generation of microprocessor, be first out with it, take advantage of huge technical investments and very costly fab investments to create the state of the art and get it to the market first. And then through our efforts of design engineer to - to the marketplace - sorry - design engineer to design engineer, do all of the necessary explanations of the new product, create the development systems so that the engineers had the design media to design that product in. And - and so that was the business model. We're going to create the state of the art processor and we're going help the design engineers design it in quickly, so that there's a marketplace for it. And then later the market would follow - the - the general semiconductor industry would follow with comparable product, with clone products. Actually, in the early days I guess the - the mechanism was that whoever designed the product would give their masks and everything else to - to their competitors so that they can make exactly the same product. And that was a model that - that seemed to make sense I guess. That was before my time. But it seemed to make sense in the early days of the semiconductor business when the marketplace itself was relatively small, and the semiconductor companies were very small. And maybe large companies would look at them and - and worry about whether this small company - you know, whether I design in this product provided by this small company because they may not be around. So if there were - if - if the product itself became a standard, then that would give them the comfort to design it in and we'll - by the year we're talking about the semiconductor companies were quite large, and had been in business for long enough that that - they were solid and people believed they would continue to be in business. So the market and image changed that way. And the other thing that had changed from Intel's perspective was that the - the design of a new generation had become really prohibitively expensive. And I forget the numbers, but if you look at that - that architectural design for a processor of the complexity of a 386 and the - the fab development and so on, you're looking at hundreds of millions - billions of dollars of investment. So business-wise it no longer made sense for - for companies just to spend billions of dollars or hundreds of millions of dollars in investment and then hand it off to another company and not have an ability to recoup that investment. It would - people would quit making investments. It - it just wasn't strategically sound. So it - so Intel had made the decision not to second source the 386, or at least not to give it away. I think offering, you know, a sharing of the - of the development costs or something might have worked. It may have been done. I don't know. But - but giving it away was just not a - a palatable choice anymore. And so at that point we were the only providers of the 386. As it turned out later there were - well, except for Harris and IBM for - for - Harris had been. And it - and there may have been a - a fourth one. I'm not sure. But in any case, there were - there weren't the dozen that there had been in previous processors. So that was a change in the marketplace.

RW: So that was successful.

DC: That campaign was incredibly successful. And we - we - we demonstrated that. In fact this new audience that we should talk to was an important audience and one that we should be talking to because they were making the decisions, not the OEMs anymore. Not really. And we concluded that we wanted to start marketing to this long-term. And so my little design team, my little marketing team I had put together started expanding and it became a permanent job. It was very gusty of the people to do this because they all left other positions inside the company and came to do this task even though they knew it was very - it might not last very long. We might - we might do this for six months or six weeks, and they might need to go back and find another job again. But - but they took the risk and - and it was successful, and so we moved on, and we started broadening our end user communications and our marketing to end users, and using more traditional media of advertising in PC magazines and electrical engineering magazines, things that PC users might be reading at the time. And at this point we felt we needed a - an ad agency, a new ad agency. Chip - the - Chip Shafer, who had done the 286 campaign, was basically a - an individual - very small shop and not really geared up to do a plethora of ads from - from a - for a whole company. So we - we went out and put our ad - put our ad campaign up for - for bid. And - quite a few companies - quite a few ad agencies bid on - on the job, tried to - to win the account. But - emerging was a - a sort of a - an unexpected candidate from Salt Lake City - Dahlin, Smith & White - brilliant, creative guy, John White, was one of the principals of that company, and he had some really - he really got it. He got what we were trying to do. He got the strategy of what we were trying to do. And we were very impressed with their creatives, which surprised us because that was a small company, too, though they did have some hi-tech accounts at the time. Novell and Word Prefect and other accounts. So they had some experience in this industry, and they were - they were large enough to handle our account. But they came out of nowhere and we were surprised. Nonetheless, we chose them. And they came up with a brilliant tag line for us that really captured our positioning. And this is an example of the kind of ads that we were doing in that period. 486 was now introduced, and so it wasn't just the 386 we were trying to up - upgrade people to, but now we wanted to upgrade people to the 486 as well. So were telling a family message. And we were still using the graffiti - graffiti imagery because it was very powerful and set us aside. It made us more friendly. It wasn't the traditional corporate Intel look. It was a - it was a more friendly look and appealed - seemed to appeal more to the end user. So in any case, in this era we used this. And the very brilliant tag line that - that DSW had come up, which was "Intel. The computer inside." And this was the positioning that we started trying to drive into the market. So this was about 1990. And so the strategic problems that Intel faced in the period were several. We still needed to move the market to the latest microprocessor for the reasons I had explained before. Or - or at least communicate to the market the advantages of that processor because it's sort of esoteric what the differences between processors are to people who are buying PCs. They're not design engineers. They probably have never written assembly code. They don't understand architectures. And so explaining the benefits of what a new microprocessor is, that was a challenge. A second challenge was that the microprocessor itself is almost invisible in the computer. It's - you - you literally have to open the box to see it. And you're-it's just, you know, some component buried inside the computer. And even though it's fundamental and defines what the whole computer can do, that's not at all obvious to the user. So these were some of the strategic dilemmas that we faced. And going back to the 386 campaign, one thing I learned in the post-research about this campaign is that we had created a brand, and didn't really understand what that was, but we created a brand. And 386 SX, SX in particular because of that imagery had become a brand. And so I needed to strategically figure out how to deal with that brand that we had created, but 386 brand didn't work for us anymore because we were moving to the 486. So how to deal with that evolution and how to deal with the invisibility of the processor. And so throughout - oh, and - and the final thing, the really important thing is our customers, the OEMs, the Compaqs and IBMs and Dell Computers and Gateway computers of the world, those were the guys who actually were selling the product to the user. They were the people who were - who were - you know, who were actually selling the product. We were merely an ingredient inside their systems. So we had to figure out a way to deal with that as well, that we're not - not actually marketing our product to the users, we're marketing an ingredient that's inside other people's product. So we needed to wrestle with all of those different elements. And 1990 was an experimental year where I kind of played around with different programs and efforts, experimenting with all those different elements to figure out how to do this marketing right. And we come into 1991, and at that point we thought - or I thought that we - that 386 was a trademark of Intel. You had mentioned that we were one of the few people shipping it, and we had invented it, and we had been treating 386 as though it was a trademark. And in early 1991 a court ruled - local San Jose court ruled, no, that's not a trademark. It's just a number. It's an - an identifier for the technology, and there is no trademark rights that Intel holds in that. And therefore, anybody can call anything a 386. Even something that Intel's not calling a 386, they can call a - a 386. So at that point our marketing, which we had been doing with the numbers, became not good because we were no longer marketing unique products. And I remember the trademark ruling came down on a Friday, and Andy stopped by my office and was pretty dejected and said, "You know, look. We - this isn't a trademark anymore. We need - you need to figure out something to do very fast because you're spending a fair amount of money on marketing, and you're marketing generic products at this point. Figure out a way to - to turn this around." And so I went home that weekend and spent the weekend thinking through all the things we had learned, and thinking about all the strategic challenges, and came back earlier the next week and presented to executive staff a program to - the Intel Inside Program. And the basic elements of that Intel Inside Program was we're an ingredient, so this needs to be a program that we work with the OEMs. And instead of us doing the marketing, we need to facilitate them doing the marketing. And we'll do that through a co-op program. We'll set aside a certain amount of processor sales that - that they purchase, that they'll be able to accrue and apply back to advertising. We'll focus it on advertising. And ultimately we'd - we'd like to get an Intel Inside logo on - on their box, on - on the actual PC, on the - the bezel of the personal computer. But we'll start just with this advertising program. And the program was - the - the idea was sort of skeptically received, but Andy was very bullish about it. He thought it was a good idea. And sent me off to put the program together, and we did that very quickly, a - a group of people and I worked together to hammer out all the details of that which turned out to be logistically very complicated. And I went on a road show with another person. And we went to - to OEMs, and several of them and tested the idea, and the general reaction was positive. And so we decided to go ahead with it. And in fact IBM liked the idea so much that we - we had said we would roll the program out in July, and IBM actually asked permission to use it even without the co-op program, just the - the logo port - portion of the program in April. And they actually ran the first Intel Inside ad in April with the logo in it. And so the - the Intel Inside Program at that point in early 1991 was off and running.

RW: That - that seems strange that - that IBM would want to advertise a foreign processor inside their computer. What was their rationale?

DC: Well, I don't - you know, that's - that's an interesting question. And - lot's of people have asked a lot of questions about IBM strategy at the time they chose to go with - with standard industry technology, the Intel processor in the Microsoft operating system, and other standard technologies in the industry rather than developing their own, which they certainly had the in-house capability of doing. And - so, you know, I - I'm certainly would have to ask - you would have to ask IBM what - what their rationale was. But I think for the use of the logo itself they were looking for - for that cache that Intel had at the time. We were - we were definitely viewed as a - as an exemplary engineering company with great technology. And I think they wanted the cache that came with that technology image with their product. And it - interesting - interesting thing happened through that process and other processes that - that I learned which really wasn't very surprising. But the - the marketing managers inside the companies welcomed our logo. They liked it. They wanted us then to go out and market our logo and explain to people why it was so good, to the broader audience, not just the - the technical community. What was good about the Intel product and why - why that logo meant something valuable. The people who actually did the ads for those companies though weren't such great supporters because they didn't - you know, you - you - anybody who's ever done advertising knows you create an ad and it's just the perfect visual representation of what you want to convey and it's artistically wonderful. And now someone's coming along and saying put this little logo in there just somewhere, and it's artistically insulting to you. And so the ad agencies ended up being not very positive about this. But the companies - the computer companies were - those first several years, were very positive about the program.

RW: Well, well the co-op people had to be positive because you paid for what percentage of their - of their ad.

DC: Well, we - as I said we - we accrued a certain percent of the - of the microprocessor purchase, and then the OEM could use that to pay up to fifty percent of any advertising, any print ad that they did. But, you know, putting that in perspective, it's a very small percent of their ad budget at the time. So that probably would - and it was - it was a positive, and it certainly - I mean it was - it was a positive and that's why - why I had proposed we do that. It was a - it was an element that - that helped popularize the program. But it - it was a relatively small element. It was - in those days it was more the cache that came with the technology image, because everybody was looking for a way to explain their quality to differentiate themselves, all the - the computer manufacturers. So it was incumbent on us then to broaden the knowledge of who Intel was and do our own advertising that explained that bigger picture of Intel and Intel Inside, and what this meant. So we did a - we did a print ad which was how to spot the very best PCs. This came - became known as the measles ad because for obvious reasons. And we just replicated the Intel Inside logo and basically said when you see this on somebody's computer it means that it has the Intel processor in it, and this is why this is good. But this wasn't very - in and of itself it was interesting, but it wasn't very powerful and it wasn't very broad. And at this point, looking at the market I decided that we wanted to do television, which Intel's only experience with doing TV ads had been the Microma ad I described earlier. So again another unpopular decision. But Andy supported this and worked with Dahlin, Smith & Wright, and we went to Industrial Light & Magic, a Lucas outfit, and - up north of the city, and had them design a very techy look and feel ad where we flew inside a computer - flew into the disc drive so that you're inside the computer, and the flew around inside the computer showing all the circuitry, and it ended up on the processor. And made a series of very powerful ads that did this showing the inside of the computer, and positioning "Intel. The computer inside." This is the microprocessor. It's in the center. It's the heart of your computer. It's the brain of your computer. And - and very dramatic fly-through technique, which at the time had never been done. So it was very visible and got the marketplace's attention. That went a long way to really establishing the Intel image to the broader marketplace, to the non-engineer if you will, to the PC buyer, and helped really establish the cache of the Intel Inside logo. And by that point we had also moved on and were actually getting the logo on the bezel of the PC, the little logo. And so that was the beginning of the Intel Inside Program. And that was about - that was early '92 by the time we were running those ads.

RW: Well, Dave House is - is normally credited with the Intel Inside. Any idea why that is?

DC: Why - I - I have never heard that.

RW: Yeah. It's - and is - on his resume or on his biography.

DC: Really?

RW: Yeah.

DC: Well, no, I have no idea why that is.

RW: So was he in charge of microprocessors at the time?

DC: He was - yeah, he was. He was in charge of microprocessors, and during the - during the time that the 386 SX campaign ran that I talked about, I reported to him during that period. As - as we move on, there were various reorganizations and he moved running other organizations, and I - I don't remember the timing of what happened where. So I - I - I reported to…

RW: You know it's very successful if - if everybody's claiming that they were involved.

DC: Well, that's right. The - the - the great thing about - about - the good feeling about doing anything that's successful, success has many fathers.

RW: Yes.

DC: And failure, failure stands alone.

RW: But you still had the problem of the number that…

DC: We did.

RW: That indeed AMD was able to put out a 486. So what was the answer to the - the number problem?

DC: Well, the answer to the number problem then, we had already launched the 486, and so that was there in the marketplace. And by then there were many competitors who were - who were selling new processors, many start-up companies who were designing processors and coming to market with them. And the Cyrix, I believe, introduced a product that they called the 486. And I forget - I think this was 1993. And they called it a 486. And looking at it architecturally, our architects felt, you know, gee, this is what they've done is gussied up a 386. This is not a 486 architecture. And yet they're calling it a 486. And so the feeling started to build inside that we had to figure a way to address this. And so I decided we're going to name what internally everybody's thinking of as the 586. We're not going to call it a 586; we're going to name it something. And again, Andy was very supportive of this, and supportive of the - of the processing. We couldn't place a very complex process. Karen Alter drove this process. And - the process was very complicated and we went out and polled lots of people for names - potential names and so on, and ended up with a list of names including Intellect was - which you had mentioned earlier which is the name of one of our development systems. We still owned the trademark on that. It had long since not been used, and so it had no association. Came up with a list of names that we could potentially call this new 586 generation of microprocessor. And Andy convened a meeting where we got all of the key executives in a room who felt that they had ownership for the microprocessor. And we - we got them in a room and went around and said what is your favorite name from this list, because we've cleared all these names. We could use any of them. What's your favorite name? And we got all the way around the room and I - there were twenty-odd people in the room, and I think every one of the ten possibilities had two votes. There was no clear winner. And so Andy said - he closed the meeting and he said, okay, well, you know, Dennis and I are going to go away and we're going to name this. And - and I'm - I, Andy, am going to announce it Monday on national television. And so he - he was very in favor of the - of Pentium processors, the name, and so that was what we chose and - but kept it quiet. Nobody except a couple of PR people knew that. And - and - and - and he and I. And Monday he went on national television and announced the name. It was quite exciting. And so that was the way the Pentium processor was launched. And at that point also, to address the issue of this constant, you know, Intel Inside means any processor from Intel, but one of our strategic problems is selling the latest. And so we want to talk about the Pentium processor, and yet can the consumer tell that they're getting a Pentium processor now when they get to the store? All they see is Intel Inside. And so we have affixed Pentium to the Intel Inside logo. And that, too, is very controversial. Nobody liked that. But I insisted on that. And - and - that was the - the launch - that was the change of the - of the logo in the launch of - of - of the naming of processors.

RW: Again, hugely successful.

DC: It - it did turn out to be hugely successful. You know, the - the thing about that whole era was, again being engineers, very data driven. And I went and did market research on everything. We built an incredibly strong market research department, hired Ellen Konar who had been market research at IBM, and she built just a crack-team of market researchers. And we - we had all kinds of data. So we would watch Pentium processor as it got accepted in the market by people as it started to mean something. And for a long time, even though we didn't introduce - no one introduced the 586. It was just such a logical progression. 586 had more equity than Pentium as a name for a processor for a long time. But eventually it changed and Pentium became - Pentium processor became more widely know, and then displaced it. So that was really an interesting - most of these were nail biting exercises because you're - you know, you're really taking profound risks, and - fundamental risks so they're - to they're company strategy. But, yeah, the - that one worked.

RW: Well, okay. So what else did you do at Intel, then.

DC: Well, that was pretty much it I guess. The Intel Inside program continued and it went through a lot of changes. Every - every year was a new challenge and a new change. We - we expanded in the early days as I explained, the - the - I only allowed print ads, the logo to be used in the co-op on print ads, and I did that for a lot of reasons. One of which was to keep it focused because if you - the - the - the net result of doing only print ads was that when you flip through a PC magazine you saw that Intel Inside logo frequently enough that it - it became em - emblazoned. People recognized it. And if you - and another reason was because it was accountable. We could - in terms of payout, we could - we knew what we were paying for. You can - you can get rate sheets from magazines and know what an ad costs and so on. So it was accountability, and there was the visibility, and there was just general focus. But as time moved on then we were so successful with TV, we wanted to expand to the co-op portion of the program so that it would encompass television as well. And we did that. And - but television provided a new challenge for the logo which was, in a print ad, even though the logo appears only in the lower right hand corner of the ad say, is visible, and yet not very intrusive, although the art director would think it's intrusive. In truth, it's not very intrusive on the ad. So you - you only get a little bit of the space. You do it consistently. It's not invasive to the - to the - to your partner - to the OEM, and yet you get something out of it. With TV it's a very different challenge, because if you flash a little logo for a few seconds in a TV ad, it's not seen. If you put it there large enough so it is seen, it becomes very intrusive and it - and works against you and the, and the person making the ad's just not going to stand for it. So we needed a different way that was - that would communicate it. And what we came up with was the bong - the ba ba ba ba! And - and so the TV signature, the logo flashes, you probably don't even really see it, but you hear that ba ba ba ba! bong, and the Intel Inside words. And so that became very powerful as a - as a - as a ingredient branding device for the TV ad. So that was a change. Later the Internet came, and that was a change. And the program grew, and there were new challenges. And so for the next half dozen years until I left Intel in - in the year 2000 - at the end of 2000, basically that's what I spent the rest of my time on. Evolving that program, and evolving marketing at Intel. The marketing organization obviously grew. By the - by - through the '90s I - I ran the - the corporate marketing for Intel which came to be a very large budget and - and a lot of people, and trying to use that very efficiently. So, yeah, that was pretty much what I did at Intel.

RW: Well, why'd you leave?

DC: Well, by the year 2000 I had - I had done this for a long time and felt like - I mean there were - there still challenges and - and things to do, and Intel's done a lot of great changes to the program since I left, and - and things to do. So there were still things to do but I had been doing that for a long time. And frankly, was just very tired. I needed some time off. It was a very grueling period. Very challenging period. Takes a lot out of you.

RW: Well, Intel has really been successful for many, many years. And why - why is that? I mean, what would say was the - the secrets of Intel's success?

DC: I - I think there's - it boils down to the people first of all. Incredible - you know, I - it was - it was wonderful working there. I - I was there for almost twenty years. I was there for nineteen-plus years, and never a dull day because of the people. That - just brilliant people that you had the - the - the pleasure to work with. They're - everywhere from the very visible top people who - who were brilliant and, you know, changed the world. Robert Noyce, who co-invented the - the integrated circuit and was just a brilliant charismatic man, and Gordon Moore, who's just a - a wonderful human being and a brilliant manager and a brilliant scientist, and Andy Grove who we talked about. All of those men were brilliant and dynamic and charismatic and great leaders. And so that - that's part of - of the - the success of Intel, is that it had this legacy of great leaders from Noyce to Gordon to Andy to Craig Barrett, and just brilliant, very strong technical people. Great leaders. But the people that they hired and cultivated were great as well. Just the - the quality of people that I worked with over the years at Intel was amazing. And it - it made coming to work, you know, challenging and - and intellectually - it drove you to do your best, and you - you always knew there were people there you could rely on, that - that would come up with solutions for whatever problems you were faced with. And that was good. So - I - in - in the end I think whatever makes any company successful is the people. And I - I would have to say that that was Intel, is the people. And one way of - of continuing that culture and that - of attracting that kind of people was Intel in the early days formed a - a very strong culture, and actually codified the culture. And we all carried around - on our badges we have little attachments that explained the corporate culture, and we had six, you know, things that were sort of pillars of the corporate culture. Risk taking, and quality, and great place to work, and customer orientation, and discipline, and results orientation, things like that. And so they were - they were codified for you so that you always kind of had this philosophy to fall back on and on driving you day to day. But in the end, what really Intel focused on in my opinion is results. We really drove for results. You didn't sit around being esoteric a lot. You - you looked for - what - what's the problem? What do I need to solve the problem? Is it on track to being solved? A lot of accountability that way. So meetings were very efficient. Plans were very well drawn. Lots of - lots of creative energy but very focused on results. I think those made Intel successful. Have made Intel successful and will continue to make it successful, because it's still got a lot of, you know, terrific people there.

RW: Absolutely. Well, thank you, Dennis. It's been great.

DC: You bet. Thank you, Rob. This was enjoyable.