Look like a duck: How Intel came from behind

Summary : Look like a duck: cool and collected on the surface, while paddling hard underneath. But never stop paddling, or you will be a sitting duck.

Look like a duck: cool and collected on the surface, while paddling hard underneath. But never stop paddling, or you will be a sitting duck.


The best companies in the valley emulate this principle.  Intel immediately comes to mind, with Andy Grove’s oft repeated mantra, “only the paranoid survive”.  On the surface, Intel never looked like its feathers were ruffled by a problem, while underneath working feverishly to correct it.  During the mid-80s RISC/CISC battle: RISC (Reduced Instruction Set Computer) was new and fast; CISC (Complex Instruction Set Computer) was old and slow.  All the experts were certain RISC was the superior technology and that Intel was in trouble.  Motorola, Sun Microsystems, MIPS, HP, DEC, IBM, and many others planned to unseat them by redefining the market as a RISC versus CISC battle.  Intel, in a seemingly arrogant move, chose to ignore RISC: they avoided publicly acknowledging the trend.  They did come out with a RISC processor.  But more importantly, they focused on the real need of the customer: SPEED and COST.  Behind the scenes they began to put RISC technology in their processors.  The 486 was the first processor to have some RISC elements, the Pentium had more, and the Pentium Pro was mostly a RISC processor.  They improved their manufacturing prowess so they could be the cost/performance leaders.  Starting with the Pentium, they openly compared their chips with other RISC processors, placing it in the middle of the pack.  This could have been suicide.  But Intel deftly positioned Pentium as the PC processor.  The message was, ‘All those other processors are for high end machines, you will never be able to afford, we will give you affordable performance.’   With Pentium Pro, Intel now had performance that could match the best processors at very low cost.  Today, Intel is confronted with the sub-$1000 PC and again they look cool, calm, and collected.  But look closer, under the surface, and you will see them paddling like crazy.

Had Intel recognized RISC, they would have instantly legitimized this way of looking at the world of processors, because they were the leader.  By not recognizing RISC as having a technology leadership position, Intel – as the market leader – was able to leave the legitimacy of the RISC/CISC argument open to question.  This bought them the precious time needed to catch up.

Be careful though, it’s not market share leadership that counts, it is mind share leadership.  Exactly what constitutes leadership varies in time.  Customers may see market share as the important leadership quality one time, technology another, and something else another.  The constant is that customers always choose one leadership quality as important.  If your marcom initiatives do not align to the customer’s viewpoint, they will fail. 

A classic failure mode example of this is Apple’s ad from the eighties when IBM entered with its Personal Computer, or PC.  Apple’s ad welcomed IBM into the market, essentially legitimizing IBM as a leader in market.  It was a rare, but costly error from Steve Jobs.  Immediately IBM was seen as a real computer maker instead of as a stodgy crasher late to the party.  Apple was no longer thought of as the innovative leader.  No, they were just a toy maker for geeks.  They made ‘home computers,’ while IBM’s was a ‘Personal Computer.’  It wasn’t long before IBM passed Apple.

In another Intel example, the shoe was on the other foot.  It was trying to get its 32 bit 386 to market and Motorola had beat it to the punch.  Motorola had quickly gained mind share as the 32 bit leader – in part because at that time, no microprocessor supplier had ever kept its leadership through a bit-scale jump.  The Wintel network concept had yet to emerge.  Customers were starting to design-in Motorola.  Intel chose to fight back in this case.  It started promoting the 286 heavily, with open comparisons to how it was better than Motorola’s processor.  Motorola lost their cool and came out with its own comparisons to Intel’s 286.  Motorola couldn’t paddle any harder, as their chip was in the market.  But they should have waited.  By comparing their 32 bit processor to the 16 bit 286, Motorola locked the customer’s mind into the idea that their part was really a match for the 286.  By acknowledging the 286, they instantly ceded their leadership position back to the market share leader.  Had they waited, the 386 would have been seen rightfully as late to market.  Instead it was perceived as the new technology leader with no peer.  Because of Motorola’s marcom error, when the 386 arrived, Intel was able to capture the crown of technology leader with ease.  It was a gutsy move by Intel’s Howard High and his team in marcom, who knowingly took the risk that Motorola might not respond.  But Andy Grove always appreciated gutsy moves, which made it possible organizationally.

Applied Materials is another company who consistently looks like a duck.  Their first cluster tool, the Precision 5000, was the most complex deposition tool ever fielded when it was introduced, and it had major reliability problems as a result.  Many, myself included, questioned if cluster tools could ever be built reliably.  Applied quietly moved ahead, installing the most massive quality improvement program in the history of the industry.  The lessons learned from that program were transferred to their Endura platform, which became a roaring success.

An important aspect in these cases is that Intel and Applied did not deny their problems, but they did not waste time acknowledging them either.  They simply fixed them quietly, displaying confidence in their solution, while never allowing themselves to be seen sweating the issue.  This approach, if done correctly, wins points with customers and infuriates competitors.  It also provides another proof of how important the M&M’s of business are – Management and Marketing – and why they belong in the same candy dish.

By G Dan Hutcheson

                                      Copyright © VLSI Research Inc.  All rights reserved.

You may like this also:

Access to and use of this Website is subject to TechInsights' Terms of Use (including Copyright Policy & Claims) and Privacy Policy. By accessing or using this Website you agree to TechInsights' Terms of Use (including Copyright Policy & Claims) and Privacy Policy.

Copyright © 2024 TechInsights Inc. All rights reserved.