Chapter 1: Crush the Competition
Marketing is civilized warfare. If you find that metaphor too brutal, or if you are not prepared to fight, you should not enlist. As long as aggressive competitors exist -- and in this rich and dynamic world they always will -- you will be under attack. Your competitors' job is to capture business and then defend that new perimeter. So is yours.
Now, a lot of marketing is creative. It's strategic. Cerebral. But eventually you must make a move -- and then the fighting begins. Even the most brilliant campaigns suffer occasional setbacks, and it is during those moments of crisis that the true mettle of the marketing team is tested.
Every company faces marketing crises at intervals throughout its history. A company that fails to surmount one can slow to a halt, even atrophy, for many years. Just surviving such a test usually means only a return to the status quo ante.
But to triumph over such a crisis, to turn possible disaster into a resounding victory, can accelerate a company's growth in a burst of sustained business momentum. Meanwhile, such an unexpected turnabout can demoralize the competition or -- at the very least -- cause considerable discomfort.
Winning, beating the odds, converting defeat to victory -- that's the point of marketing. The stories of such marketing coups are our business legends -- what Iacocca did at Chrysler and what Townsend did at Avis. It is what Apple is trying to do right now in office automation.
And it is what Intel had to do in 1980. I know, because I was there. My career depended on a single victory.
Intel Corporation was founded in 1968 by Robert Noyce, the inventor of the integrated circuit, Gordon Moore, a legendary high-technology scientist and business strategist, and Andrew Grove, a now famous manager and executive.
Intel owed its success (Ben Rosen once called it the most important firm in America) to inventive genius, an ability to convert ideas into products (such as the famous microprocessor), Grove's dynamic management, and, not least, a talent for developing new markets for its new products. All those factors combined to give Intel one of the most remarkable starts in American business history.
But not all of Intel's success derived from intrinsic strengths. For a long time the company had also benefited from the benign neglect of more powerful firms in the same industry. Like many hot young electronics firms, Intel had focused on new markets, pursuing a path the industry giants had no interest in following. But the day of reckoning had come. By the mid-1970s Intel's achievements had become an embarrassment to its competitors and the target for most of the largest semiconductor manufacturers in the United States, Europe, and Japan. The list of competitors poised for attack was more than a little daunting: Texas Instruments, Motorola, National Semiconductor, Philips, Siemens, Nippon Electric Corporation (NEC), Hitachi, and Fujitsu, among others -- the Billion Dollar Club of the semiconductor industry.
Intel still prospered but was losing ground in some important markets and was threatened in others. Intel once had been the leading supplier of 1,024-bit "dynamic" RAMs (random access memory) chips, but had lost that leadership to a start-up company. We had been unable to regain that momentum. A number of companies also had jumped into the EPROM (erasable programmable read only memory) chip market and were applying pressure. Finally, by 1979, Intel's strong position in the microprocessor market, though relatively intact, had suffered inroads from a start-up company named Zilog, and from Motorola, the latter a number of times Intel's size.
By late 1979 Intel was under full siege. Such attacks were nothing new to Intel, and the company had won more than its share of battles. But this threat was different in one very important way: The product line in dispute, the model 8086 16-bit microprocessor family, was the linchpin of the entire corporation. A number of multimillion-dollar Intel businesses depended on its success.
In particular, the sale of every Intel 8086 and its companion chip, the 8-bit 8088, pulled along large numbers of peripheral, memory, and controller chips worth in total ten times as much as the 8086. Whenever an 8086 sale was lost, the departing customer would frequently turn to the new supplier for those ancillary products. On top of that, Intel had two very profitable systems businesses dependent upon the success of the 8086.
Les Vadasz and I had been co-general managers of the microprocessor division in 1976 when the 8086 was being planned. At the time we decided to make the product an extension of the then-successful 8080 family. That created some design problems, but they were more than counterbalanced, in our opinion, by the resulting access to a large existing software library.
The 8086 was introduced to the market in 1978. As the first high-performance, fully supported 16-bit microprocessor, it had quickly gained the top position in the market, capturing the lead from older and less capable products supplied by Texas Instruments and National Semiconductor. In response, Zilog and Motorola prematurely announced their own "paper tigers" (products that existed only on paper). Customers loved the features of the proposed products and were not too happy about some of the compromises Intel had made, so it was obvious that when and if those microprocessors ever emerged from the drawing boards, they would be a serious threat.
Meanwhile, as Intel had grown, the management had reorganized, and I left the microprocessor business to become the general manager of one of Intel's microprocessor-based systems businesses. Needless to say, any success I would have in my new role would be vitally dependent upon the survival of the 8086. So I remained in close touch with the 8086's marketing effort.
The 8086 marketing and sales group was suffering from apathy brought on by shattered morale. It was demoralizing to have one customer after the next lecture you about your employer's failures and your competitors' strengths. Many customers actually relished the opportunity to stick it to the famous Intel.
Some of the younger marketing people couldn't take the humiliation. It was easier to work on other projects. Being abused by customers -- and even Intel's own sales force -- wasn't fun.
Management encouragement had been ineffective at correcting what was becoming a destructive situation. In late November Don Buckout, an Intel field engineer on Long Island, sent management an incisive and desperate eight-page telex. The discussion of Buckout's telex at the executive staff meeting the following Tuesday couldn't have been more unpleasant. By the end of it I had either volunteered or been asked by Grove to run a marketing task force charged with solving the 8086 problem.
That was the beginning of Operation Crush.
A blue-ribbon group of the best sales and marketing people in the company was quickly assembled on December 4. We met continuously for three days. Among the "volunteers" were Jim Lally, the general manager of board products; Rich Bader, one of Jim's product managers; Dave House, the general manager of the microprocessor division; Jeff Katz, the marketing manager for microprocessors; Casey Powell, the regional manager to whom Buckout reported; and Regis McKenna, Silicon Valley's lop marketing consultant. That was the first thing we did right. We did not delegate the job.
I appreciate that this runs counter to the principles in most textbooks on management and that many managers become trapped following such a path, but in the current crisis delegating responsibility had already failed. And, I would argue, the great marketing crusades of the past were led by the top people in the company: Lee Iacocca and Avis's Robert Townsend, to name two.
The first thing the group did was agree on the problem. That wasn't hard. There were three of us in the race: Motorola was going to be first, Zilog second, and Intel was headed for obscurity. All of us agreed that if we whipped Motorola, we would win. For that reason we made our goal not simply regaining market share but restoring Intel's preeminence in the market.
In the semiconductor business, the only market share you really care about is the one you maintain when the market is mature. To accomplish that, a firm must convince sufficient numbers of customers to "design in" (that is, integrate) your chip into their products. So the task force established a goal of achieving two thousand "design wins" by the end of 1980.
That was the second thing we did right. We had set a shockingly high goal. Knowledgeable observers thought a few hundred wins more reasonable. We decided that every salesman could get one win a month. By simple arithmetic, the number two thousand fell out. We trusted our people to come through.
As the discussion developed, we increasingly talked about what our real objective was. It was Jim Lally who articulated the need to "crush the competition." The word was wonderful. It captured the essence of our attitude. It also left no doubt about the single-mindedness of our purpose.
The code name Crush was never supposed to be made public. Roger Borovoy, the corporate counsel, was concerned about the implications of such a loaded word. But the name already was spreading like wildfire throughout the company. Everyone loved it. We had been kicked around enough; Crush signaled that we now meant to stand our ground and fight aggressively. And it meant we were going to win.
We decided to kick off the campaign before Christmas, not waiting until the first of the year. Now that we had a concept, there was no reason to defer action because of the holidays.
CRUSH INVENTS A PRODUCT
Our first task was to define the market and its competitive environment precisely. Hours were spend discussing customers and why we had won or lost various accounts. By the end of the discussion we had concluded that the customers could be divided into three general groups: hardware-oriented companies; software-oriented firms wanting to use Intel software; and software-oriented companies wanting to write their own software. We were doing well with the first two groups but nearly always lost out with the third.
That exercise all but confirmed what we already knew: Software-oriented customers, many of whom had migrated from the minicomputer field, wanted a microprocessor "architecture" (design) with precisely the features we lacked and Motorola and Zilog had. Moreover, those computer people did not really understand the advantages of the Intel products and were not crediting us with our strengths.
Thus, we decided, what we needed was a new product that better fitted the needs of our customer base. We would have to invent one.
Everyone on the task force accepted the harsh truth that Motorola and Zilog had better devices. If Intel tried to fight the battle only by claiming our microprocessor was better than theirs, we were going to lose. But we also knew that a microprocessor designer needed more than just the processor, and we had our competitors beaten hands down when it came to the extras. We had been playing to competitors' strengths, and it was time to start selling our own.
What were those strengths? We concluded that Intel's competitive advantages were these:
1. A fine image as a technology leader: Customers were concerned if they left Intel they would lose out on future developments.
2. A more complete product family and a plan to enhance it: Motorola was weak in this area. If we could make customers aware of that fact, it would be a great advantage to us.
3. A well-focused and superbly trained technical sales force: The Motorola sales force was a group of generalists. They lacked technical support in the field as well. Many were afraid of the microprocessor. We knew that if we could just get the customer to ask Intel before making a design decision, we usually could beat the competition.
4. Better performance at the system level: If the customer evaluated total capability -- a system with the 8086, math coprocessors and peripheral circuits -- we came out ahead. We also had a well-thought-out interconnection scheme. Here, too, Motorola was weak.
5. Ultimately, perhaps the most important advantage Intel had was that Motorola's customers were experiencing great difficulty making that chip work in their products. Intel had great customer service and support. We could assure a customer's success with our device. By comparison, choosing the Motorola path clearly presented a risk to the customer.
By the end of the three-day meeting, we had a "product" -- at the least the idea of one. We also had a preliminary schedule for delivering that product to market. Now, we needed to organize the company to deliver our message.
MOUNTING THE CRUSADE
The task force finished its preliminary work on a Friday. By the following Tuesday the multimillion-dollar program had been approved. Within a week the new strategy had been presented to the sales force and had earned its support.
I cannot stress the importance of that last step. Too often marketing programs are designed in an ivory tower. The sales force can instantly recognize a plan that will not work, so feedback from the field is critical. If the salespeople don't buy in at the outset, you should probably start over.
Fortunately, our sales force liked what it saw. The salespeople wanted a good fight as much as anyone in the firm, if not more.
Ultimately, Crush encompassed top management, the sales force, four marketing departments at three geographic locations, and a corporate communications group. All had to work together to pull off the internal portion of the operation. In all, Crush employed the talents of more than a thousand employees. The next big step would be to organize this army to march single-mindedly in one direction. The only common authority over the diverse organization was the president himself, Andy Grove.
Years later I learned that Dave Packard, one of the founders of Hewlett-Packard, used to say that marketing is too important to be left to the marketing department. If any event proved his point, it was the Crush kickoff meeting. It was held at the San Jose Hyatt House, with more than a hundred Intel managers in attendance.
As people walked in the door, they received a brown button with "Crush" spelled out in large orange letters (we used the orange color of the Denver Broncos, whose defensive team was referred to as the "Orange Crush" that year). The key speakers were Bob Noyce and Andy Grove. Bob let people know how important winning was to the company. Andy explained that Crush would remain a corporate focus until the job was done. As subtlety is not one of Andy's strengths, the managers had no doubt about what that statement meant.
There was a lot of work to be done. The key to accomplishing it all was getting everyone to do his or her share. The task force toured the company, explaining to groups the Crush plan and what we wanted the employees to do. Intel is a great place for teamwork, and people were quick to sign up.
The Crush crusade had begun.
The task force next chartered a number of interdepartmental committees to work out the details of implementation. That meant converting what until then had been mere ideas into actual plans of action. New sales aids were needed. System-level benchmarks had to be developed. Numerous articles had to be written for the trade magazines. An effort was even launched to get our customers to write about their experiences using the 8086. In all, more than fifty articles were published in the trade press.
We committed ourselves to preparing, within ninety days, a catalog of Intel's future products. That meant a massive effort writing preliminary data sheets on a large number of parts. In the end the new catalog ran more than a hundred information-packed pages. The "Futures Catalog," as it came to be called, served as a tangible demonstration of Intel's resurgent position in the market, a cornerstone for the seminar blitz that followed.
The seminar series was a tremendous enterprise. Our strategy was to focus first on large customers, as winning them was crucial. We had to make sure those customers appreciated the benefits of our products and our future plans in components, development systems, and software. For the first quarter we targeted twenty-five major customer seminars around the world. They were successful and were followed soon after by nearly fifty full-day seminars for the general public.
The seminars were attended by thousands of potential customers. One reason they came was to get a copy of the Futures Catalog. But to do that, each attendee had to register and fill out a qualification form. Intel then hired college students and put them in sales offices with the assignment to follow up on those leads. In most cases they were quite successful.
That burst of activity was merely a prelude to a climax: a users' forum at which we would discuss with our most important customers the in-depth details of our new products. To guarantee attendance at that event, we promised the top customers an opportunity to get together with Intel managers and engineers -- not only to learn about our future plans but actually to influence them.
The seminar program turned out to be a tremendous drain on our corporate resources, but we didn't dare stop, because it was working. The morale in the field was picking up, the factory staff was feeling better, and most important, we had lured our principal competitor into fighting us on our own turf. Motorola even published its own "futures" catalog. As ours had been the result of dedicated efforts of entire marketing groups, Motorola's catalog seemed second-rate by comparison, adding to our credibility and undermining theirs. Motorola's response to our announcement of a co-processor chip was a device that not only didn't solve the customers' problem but exposed the inadequacy of that firm's product line. Ultimately the Motorola catalog became an Intel sales tool, possibly the best one we had.
The one large client we had to win over was IBM. And we did -- though why is still not clear to me. Dave House says IBM believed Intel had the only product that could be supplied in volume to support its needs. I myself suspect that availability of software for the Intel product line played the decisive role. The software existed in part because we had chosen to make the 8086 an extension of the 8080 and also because of the momentum built by Crush.
During all of this selling activity, Intel kicked off a big PR and advertising campaign. The old ad program was scrapped, and Regis McKenna created a new one around the theme, "There is only one high-performance VLSI computer solution -- Intel delivers it." In support, Intel executives visited the business and trade press around the world.
At that point Operation Crush seemed to be working pretty well. The rate of design wins had picked up, as had the sale of development systems. We were monitoring our progress every two weeks, and by June things were looking good. In marketing, Dave House's groups delivered the sales support material on time, including the "Klingon Neutralization Kit," a 4-foot wooden box containing sales aids. "If the arguments did not work," he joked, "you could always drop it on the competition." We never had to; as far as we could tell, Motorola had already been stunned by the intensity of our effort.
Still, not everything had gone as planned. We appeared to be falling short of our goal of two thousand design wins.
When we kicked off Crush, we had promised the field salespeople a contest. Jim Lally was responsible for designing the program, and he had originally thought of sending the winners to Hawaii. As the program gained momentum, Tahiti was chosen as even more appealing. After all, these people were the key to the program, and they were killing themselves.
In June, as we looked at the numbers, it seemed we would fall far short of the goal. We became concerned that if there were only a very few winners, the contest would backfire. We would end up demotivating the sales force instead of motivating it. So we relaxed some of the criteria for validating a design win.
As it turned out, we didn't need to. The way Jim had designed the program, poor performance by a few could jeopardize the opportunity for others to win. By the third and fourth quarters, therefore, the peer pressure in the field on laggards was enormous. As important as all the other Crush activities were, the competition was probably the most important reason for our ultimate victory. The field was absolutely brutal in its pursuit of design wins and in self-enforcement.
In the end we did reach the two thousand design wins target. As a reward, almost the entire field sales force went on a trip to Paradise. They deserved it.
By the time Crush was over, our victory was almost complete. Intel all but owned the business application segment of the 16-bit microprocessor market. Today the Intel-type microprocessor architecture has about an 85 percent market share.
That result was far better than any of us would have dreamed possible. Even if we had lost IBM, the company would have been better off because of Crush.
Still, we had failed to utterly Crush Motorola. Intel had beaten it in 8-bit and 16-bit microprocessors and had won the battle in the general-purpose microcontroller market, but that did not stop Motorola from entering the 32-bit battle. As I write, the two companies are again locked in a struggle for market share. Motorola is a much tougher company today. Its executives apparently learned as much from Crush as we did at Intel.
CRUSH IN HINDSIGHT
The process we at Intel went through with Operation Crush began as an intellectual exercise. We first had to understand the market segments and why were losing or winning in each of them. Once that was done, marketing could devise a product to meet the needs of the customer. We did not ask engineering to do anything different; that would have taken too much time. Time was the one thing we didn't have. So instead, we simply took the devices we had, adding Intel's credibility and a future direction, and then "dynamically repositioned" the product line (as Regis McKenna would say) as a complete solution. Marketing took what it had and created a "new" product line that the customers believed they needed. In the process we produced a strategy the field sales force could believe in.
Motorola also helped us. It had the opportunity to consolidate its victory yet instead fell into the trap of confronting our strengths head on. It could have been different but chose to be the same. Motorola had the chance to debunk our "futures" strategy as an act of desperation -- which it was. Instead, our competitor legitimized our program by putting out an inferior imitation. Had Motorola chosen to remain aloof from our challenge, I think Intel would have been in deep trouble.
Motorola also had an incomplete product. It lacked many of the required peripherals and did not have the support to meet customer needs. Motorola couldn't assure its customers' success. Intel, on the other hand, could. Intel gave good service; Motorola (because it had failed to invest in the support infrastructure) could not. On top of that, Motorola had failed to realize it needed a different type of sales force to sell microprocessors. Intel had in place a group of specialists. We had been hiring people with computer backgrounds for a number of years, people who could effectively deal with the engineers who were our customers.
Probably the most important lesson that came out of Crush was a realization that a big crisis is best answered by a "crusade." Our greatest promotion was more an act of leadership than a flash of creative brilliance. Intel was loaded with product champions and marketing intellectuals, but in the final analysis what made Crush work was conviction and grit.
At Intel, people assumed that any problem could be solved. That made the job a lot easier. The team had no doubt that Regis McKenna would figure out how to position the product. It never entered Jim Lally's or Casey Powell's mind that we could fail. Their confidence was infectious.
Behind everything was Andy Grove, Intel's president, who supported the crusade with his time, energy, and conviction.
All the key ingredients -- the organizations, the products, the people -- had been there before Crush. The difference was that with Crush we stopped cowering at the competition and started believing in ourselves. As we regained our confidence, Intel exhibited hope rather than despair. The market sensed that change, and soon our customers were cheering on Intel's counterattack.
Yes, marketing is civilized warfare. In the pages that follow, I hope to teach you how to fight it.
Copyright © 1986 by William H. Davidow