Some of the most important technologies of the past 50 years--the transistor, the relational database and the microprocessor--weren't the slam dunks for their creators that you might expect.
Some inventors lost their lead through a lack of insight. Corporate politics sometimes plays a role. More often, the delay of payback is simply the result of poor timing--a reasonable strategy at the wrong time. Take the microdrive at the heart of many of today's MP3 players, for instance. It was invented long before the world was ready for something like the iPod.
Still, those billions of dollars in research and development eventually paid off for at least some technology makers.
Here's some notable examples of inventions gone wrong and opportunities missed:
1. The transistor
In 1947, scientists at AT&T's Bell Labs created the world's first silicon transistor. Three of its scientists would later win the Nobel Prize in physics for the invention. Bell Labs obtained a patent for the device, but the invention was licensed to, among others, IBM, Texas Instruments and the forerunner of Sony. The goal was to avoid antitrust problems with the U.S. government. (In a 1956 consent decree, AT&T agreed to license the transistor freely.)
But relatively easy licensing terms cost AT&T millions in royalties.
"There are trillions of transistors in use," said Richard Belgard, a patent consultant.
On the bright side, the foundational patent would have expired in the mid-1960s, years before the computer revolution. By contrast, AT&T got to keep its phone monopoly until the mid-1980s.
AT&T had subsequent brushes with near-greatness, but these seem tougher to explain. It invented, but didn't become the dominant name in Unix. It passed on an opportunity to own cellular licenses in the '80s (although it got into cellular later). It also tried its hand at PCs.
2. Owning a bit of the Internet
Back in the early '90s, Robert Cailliau of the European Organization for Nuclear Research, or CERN, contacted venture capitalist Sven Lingjaerde to see whether the lab could get funding for its World Wide Web project.
"When the project grew in size, more money was needed, and the top management of CERN then decided to cut the budget, claiming it was not directly linked to fundamental research and it was starting to cost too much," Lingjaerde said in an e-mail. "We were considering putting money behind the project, but only if a strong U.S. VC would join. We knew that our small means (would) not be enough. The business model was also not clear."
He tried to contact two well-known U.S. venture capitalists. The first never responded, despite several attempts. The second, whom Lingjaerde sent a five- page fax to, was interested but said, "I don't see how you can make money with (the) Internet." A few years later, both became huge promoters of the Internet.
Still, as Linjaerde points out, the market did look iffy back then, and everyone since has made quite a bit off of it. And it's tough to say how big the phenomenon would have become, had it been commercial at first.
Jerry Kaplan had burned through $75 million while running GO Computing, a foiled attempt at pen-based computing chronicled in his book "Startup: A Silicon Valley Adventure." But in 1994, he co-founded Onsale, one of the first online-auction companies. Backed by Kleiner Perkins Caufield & Byers, it became an early leader.
"It's like money from heaven," he described Onsale to BusinessWeek.
Heaven was short-lived. In 1995, eBay was born. A few years later, Onsale went on to merge with Egghead and got auctioned off in pieces.
"If you don't get the model exactly right, capitalism can be unforgiving," Kaplan said in an interview. eBay created a forum for people to sell things to each other. Onsale specialized in auctioning off remainders. "That was a fundamental difference," he said.
By the time the merger with Egghead came along, so much venture money was flowing that online retailers "were buying products for $1 and selling them for 95 cents, and trying to make up the difference in volume," he joked.
Still, Kaplan asserts that Onsale can be looked at as a success. The company's quarterly revenue nearly reached $100 million during some periods. eBay also eventually bought some of the company's patents. In the meantime, Kaplan has written another book, a fictionalized account of the boom years called "Rocket Ride," and is looking for a publisher. He also kicked off a new game start-up called Winster.
4. Silicon nanowires
Silicon nanowires, tiny filaments of silicon, could well become the foundational technology for the chip industry in the coming decades. The foundational patents--filed by, yes, AT&T's Bell Labs--were first issued in 1964 and expired years before a practical market could evolve.
AT&T, however, isn't the only nano pioneer that may not profit much from its inventions. NEC invented single-walled nanotubes in 1991. Although the company is actively licensing its patents, the initial ones will expire circa 2008. The nanotube industry is only just getting fired up, and the technology may not affect the electronics market until after 2010. One major semiconductor maker, speaking on condition of anonymity, said it will let the clock run out.
IBM also secured fundamental nanotube patents at about the same time.
5. Busicom's processor
The Intel 4004, the world's first microprocessor, debuted in 1971. The rights to the chip, however, initially belonged to a Japanese calculator maker called Busicom, which commissioned Intel to build it in 1969.
By the time the chip came out, calculator prices had dropped, and Busicom wanted a discount. Intel agreed on the condition that it could sell the 4004 (technically a bundle of three chips) outside the calculator market. Busicom agreed.
Still, it took a while for the invention to catch on.
"I think it gave Intel its future, and for the first 15 years, we didn't realize it," Intel Chairman Andy Grove said in a 2001 interview. "It has become Intel's defining business area. But for...maybe the first 10 years, we looked at it as a sideshow. It kind of makes you wonder how many sideshows there are that never become anything more."
6. Relational database software
IBM's engineers can take credit for inventing the hard disk drive, the RISC (reduced instruction set computing) chip and speech recognition software, among other technology. The company has been granted in excess of 22,000 patents in the last decade, more than its top 10 competitors combined. But the company doesn't take all of its inventions to market successfully. Take the relational database, for instance.
A young IBM engineer named Edgar Codd defined the concept and structure of the relational database back in the '60s and '70s. Codd's revolutionary idea was to organize data into tables of rows and columns, and to relate that data to other tables. His work produced the blueprint for how to build a relational database, as well as the foundation for what would become SQL, or Structured Query Language, a standard way to access data.
At the time, however, the technology didn't mesh with IBM's corporate strategy. The company was heavily invested in an older database model. The result: IBM didn't market a product based on Codd's ideas until 1978--one year after a young entrepreneur named Larry Ellison founded Oracle. Ellison's company went on to become the leader in relational database software, a $13.5 billion market that Oracle leads to this day.
Microsoft got into operating systems by chance, but this story starts with IBM.
For its first PC, IBM initially considered using the C/PM system from Digital Research. Because Digital Research wouldn't sign a nondisclosure agreeent, IBM asked Microsoft, then developing applications for IBM, for MS-DOS.
MS-DOS, however, was actually based on QDOS, an operating system created by Tim Paterson of Seattle Computer Products. Microsoft bought QDOS from SCP (which didn't know about the IBM deal) for $50,000.
The sale became the basis of an empire. Subsequently, Paterson worked temporarily at Microsoft.
8. SGI buys almost all of Cray
When Silicon Graphics Inc. bought Cray in 1996, the company was one of the giants in Silicon Valley. SGI execs mingled with Bill and Hillary Clinton, and its technology was behind blockbusters like "Jurassic Park."
But SGI failed at the time to buy one part of the business--the UE10000 server, constructed out of 64 UltraSparc processors. Instead, Sun Microsystems bought the UE10000 and transformed it into the E10000 line, a flagship family of Unix servers that let Sun compete directly against IBM at the high end of the market.
Sun vaulted to prominence in the dot-com years, partly on the strength of the systems, which often are priced at more than $1 million.
Meanwhile, SGI, for a variety of reasons, sank to the sidelines and became a source of vacant office space for, among others, Netscape and Google.
9. Yahoo passes on Google
Most technology mergers don't work, but there are cases in which an established company could have avoided big headaches later. Google was a project at Stanford University's engineering labs in 1998, when the founders showed it off to Yahoo co-founder David Filo. According to Google, Filo said he wanted to talk with them when the technology was fully developed and scalable. Sources said Yahoo even had a chance to buy Google.
Since then, Google, of course, has become Yahoo's biggest competitor. But in retrospect, it seems somewhat reasonable that Yahoo wouldn't have been jumping to buy the company. Search was a flooded field at the time. Stanford had little luck finding early investors.
Yahoo, however, isn't alone. It offered itself--in vain--to Netscape back in the mid-1990s.
10. The microdrive and hard-drive-based MP3 players
IBM started shipping an invention called the microdrive--a mini hard drive with a 1-inch diameter platter, in 1999 and waited for business customers to snap it up. And waited...and waited.
Sales never materialized, and IBM, which invented the hard drive back in the '50s, continued to lose money on drives. (HP also came up with a small drive in the '90s but snuffed it.)
"IBM didn't see the consumer," said Bill Healy, senior vice president of product strategy and marketing at Hitachi Global Storage Technologies and a former IBMer. "Hitachi is the GE of Japan. They make rice cookers, refrigerators, nuclear-power plants."
Mistake? Hitachi has had more luck selling drives, but the business is still notoriously competitive and profits are often elusive. And, unlike IBM, Hitachi faces a slew of competitors in this market.
On a somewhat related note, Compaq Computer, Dell and others marketed MP3 players with hard drives before Apple did. However, they were home systems with standard PC hard drives. In January 2001, it seemed like a promising market. In October 2001, Apple came out with the first iPod based on a novel 1.8-inch drive that had interested few manufacturers. Portability won out.
11. Xerox PARC
Move along, folks, nothing to see here. Xerox has been flayed mercilessly for allowing concepts such as the desktop PC, Ethernet networking and the laser printer--all invented at its famed Palo Alto Research Center, or PARC--to get exploited by others.
The photocopier giant is now trying to stay afloat in a world going paperless. Still, PARC did help launch the careers of a number of people: James Clark, Alan Kay, Robert Metcalf and Lawrence Tesler, among others.
CNET News.com's Mike Ricciuti contributed to this report.