Moore’s Law arguably has underpinned business models across the computing, software, consumer electronics and communications industries for decades. Despite concerns that Moore’s Law–the observation that transistor density doubles about every 18 months–would cease to operate at some point, so far that concern has been misplaced.
Reed Hastings is among CEOs who recently have relied on a continuation of Moore’s Law as the underpinning of a radical change in strategy, shifting from shipping DVDs using the postal service to streaming.
As logical as that decision might seem in retrospect, it was a gamble. Hastings had gambled before. Not many recall that the original Netflix business entailed selling DVDs.
When Netflix was launched in 1998, “I would guess that 95 perent of our revenues were coming from the sales of DVDs ,” said Marc Randolph, former Netflix CEO.
But Netflix already could see a major change was needed. “it was obvious to us that this was not a sustainable business,” said Randolph. “It was inevitable that at some point in the near future we would have Amazon entering the DVD business.” And Walmart and other mass market retailers.
“All of which would have crushed our margins and slowly but surely driven us out of business,” said Randolph. And remember, in switching to DVD rentals, Netflix sacrificed 95 percent of its revenue.
The decision to focus on streaming also was a gamble. In 1998, some argue, it would have cost Netflix as much as $270 to stream a single movie. By 2010 it might have cost about five cents for Netflix to stream a movie of equivalent image quality.
That is why DVDs by mail was necessary. Netflix could not build a business on streaming. But Hastings considered Moore’s Law.
“We took out our spreadsheets and we figured we’d get 14 Mbps to the home by 2012, which turns out is about what we will get,” said Hastings. “If you drag it out to 2021, we will all have a gigabit to the home.” That meant streaming delivery was indeed possible.
Nor is that the first time a major video innovation was built on assumptions about Moore’s Law.
Back in the mid-1980s, when standards for high-definition TV were being debated, the lead proposals were for hybrid analog-digital systems, using as much as 45 MHz of spectrum, where standard television used 6-MHz channels.
Tom Elliot, former Tele-Communications Inc. SVP, recalls the potential disruption of the cable TV business model. Elliot estimated at the time that it would take 15 years to recover the investment in early proposed hybrid versions of high definition TV.
The problem, as Elliot saw it, was that the hybrid system would in turn soon be eclipsed by all-digital TV, requiring two major generations of investments inside of 15 years.
Elliot recalls TCI SVP John Sie coming into his office and asking how the company could avoid being “run over” by broadcasters and consumer electronics firms. Elliot recalls arguing that Moore’s Law would hold, though many at the time argued progress was about to flatten.
So the big gamble was on Moore’s Law. If the law held, then it would be only a matter of time before decoders roughly corresponding to a “mainframe in the living room” could be built for consumers.
That amount of processing power would allow an all-digital signal representing perhaps 100 MHz of raw, uncompressed video to be encoded, and then decoded on the fly, by a consumer terminal the industry could afford. Virtually nobody thought that was feasible.
The big gamble was Moore’s Law holding, allowing both low-cost decoding and use of standard 6-NHz TV channels.
Almost nobody but Elliot, and later TCI, actually thought it could be done. He recalls that position as being “really lonely.”
But Elliot knew Moore’s Law. He had talked to clean room techs. Yield might be an issue initially, but if Moore’s Law held, an affordable set-too was economically possible, even if–at the time–that required the equivalent of a mainframe in the livingroom.
Elliot was proven right, as controversial as the call was at the time.
So Moore’s Law has been the foundation for transformative business decisions more than once.
TCI gambled its future on digital TV built on Moore’s Law. Netflix likewise gambled its business on access bandwidth. Both were bold, dangerous business decisions, built on assumptions about Moore’s Law continuing to operate.