Companies are preparing for the intellectual-property battle
AS A 23-year-old engineer working for Hewlett-Packard in 1975, Joe Beyers had a clever idea. The new desktop computers, he thought, should be able to work on more than one task a time (something that at that point only much bigger computers could do), rather than forcing users to twiddle their thumbs as they waited for new applications to launch. After many late nights in the lab, he devised a way to make it happen, and duly had a patent granted on his invention—his first of many. Yet within a few years, all computer-makers were using the same technique.
Mr Beyers told Bill Hewlett and Dave Packard, the company's founders, that the idea was being copied, and tried to persuade them to enforce the patent. But both said no; the firm was benefiting from the invention in its own products, and that was enough. “I was disappointed,” says Mr Beyers. “I thought it was a very valuable patent, and HP deserved some value from it. I put it aside and moved on. But in the back of my mind, I felt that we had not done enough to get the full value of it.”
Times change, and so did Mr Beyers's role in the company. Thirty years on, he now manages a team of 50 people in HP's intellectual-property division. When he meets the boss, it is to report on the group's patenting and licensing business, an area where engineers and lawyers work hand in glove. Since the unit was created in 2002, HP's licensing revenue has increased from $50m to over $200m. “It represents a fairly dramatic shift in how HP treats IP,” Mr Beyers concedes.
The changes at HP are mirrored throughout the technology sector. Companies are now treating intellectual property as a business asset not very different from a product on a shelf. They are spending more on R&D, filing for more patents and licensing out their ideas, as well as licensing in other people's. They are also increasingly asserting their patents against others, demanding royalties and going to court to get them. “We are in the middle of an explosion in the use of IP to try to protect market positions,” says Intel's Mr Sewell.
The biggest reason for relying more heavily on intellectual property is that other firms are doing the same. Most companies readily admit to the drawbacks. It can make it tougher to build new products without accidentally infringing someone else's patents, as well as increasing costs and sometimes leading to spurious lawsuits. Yet technology executives insist they must harness the system lest they be crushed by it. Pressed on why his company is pushing to obtain more and wider patents, Henning Kagermann, the boss of SAP, a large German software firm, exclaims in exasperation: “These are the rules of the game!”
“There is certainly a level of mutually assured destruction among the big companies. If you build up your patent portfolio, I build up mine—nukes pointing at each other,” says Mr Papadopoulos at Sun Microsystems. But he sees it as an advantage rather than a problem. “That has exactly the right outcome. We sit here and exchange patents with each other. Ultimately, that's great: you have a set of companies doing more innovation than they would have otherwise,” he says.
Those inventions increasingly get turned into property rights. For example, Nokia has over 12,000 existing patents globally, and 10,000 innovations in the process of being patented. It files around 1,500 applications a year. IBM has around 40,000 patents and is granted 3,000 more every year, which has made it the number one recipient at America's patent office for the past 12 years. HP last year ranked fourth in America, with 1,783 patents; worldwide, it holds around 25,000. Microsoft has recently sprinted into the market, with around 10,000 applications pending.
Patents have grown in parallel with an increase in spending on research and development. One rule of thumb is that tech companies file almost two patents for every $1m they spend on R&D. The likelihood that research will yield a patent has increased hugely, thanks in good part to the growth in software patents. According to a 2004 study by James Bessen, a researcher at Boston University's School of Law, and Robert Hunt, an economist at the Federal Reserve Bank of Philadelphia, by the end of the 1990s firms were able to obtain more than twice as many software patents for every R&D dollar they spent than at the start of the decade.
Messrs Bessen and Hunt find it worrying that the growth in the number of patents exceeds the increase in R&D expenditure. They believe it indicates that these are “cheap” patents being used as a substitute for more R&D. In other words, the cake is not getting much bigger, it is merely being cut into more and ever thinner slices. Is the industry doing too much patenting merely for the sake of it?
“We have 10,000 patents—it's an awful lot of patents,” says Mr Sewell at Intel. “Would I be happy with 1,000 patents rather then 10,000? Yes, provided the rest of the world did the same thing.” John Kelly, who directs IBM's intellectual-property strategy, explains that: “Even though we have 3,000 patents [awarded annually in America], if we had to, I could make that number 10,000.”
Despite this proliferation, numerous economic studies show that only about 5% of patents end up having any value, and that a small handful of those account for most of the income received from patents. Moreover, according to a study by America's National Bureau for Economic Research in 2000, patents were less effective in protecting innovation than things like secrecy, speed to market and complementary manufacturing, sales or service.
With firms acquiring more patent ammunition each year, what if someone were to pull the trigger? “Should one company use its portfolio against another, it faces potential retaliation. The best that can happen is nothing happens,” says Joe LaSala, general counsel of Novell, a software company that sells proprietary applications along with the open-source Linux operating system.
With the plethora of patents swishing around the system, it is impossible to avoid potential infringement, so self-defence becomes imperative. One former executive at a search-engine company says that all the major search firms—Google, Yahoo! and MSN—actually infringe each other's patents in some way. Their intellectual-property strategies are designed to ensure a balance of power. The same is true in nearly all other areas of IT and telecoms. Everyone has an interest in preserving this precarious equilibrium lest the whole edifice come crashing down.
Pure intellect
Qualcomm is the very model of an intellectual-property company. Where many companies line their walls with pictures of their products, in the firm's San Diego headquarters the corridors are covered with plaques of its patents. Qualcomm created a technology called CDMA, which now forms the basis of third-generation wireless networks. Around one-third of the company's revenues (and 60% of its profits) come from royalties on all equipment that uses the technology; the remainder comes from selling the chips that rely on that intellectual property, where it has a market share of over 80%.
Because its technology underlies the third-generation mobile-phone standard, Qualcomm has become a toll bridge that all equipment-makers must cross. “Our licensees don't like to pay us royalties, but they forget the work we put in to get the business. We provide them with quite a bit more than a patent licence,” explains Steve Altman, Qualcomm's president. “If we were just an IP shop, we would not have been successful. What caused us to be a success was that very early on we didn't just license patents, we enabled the manufacturers to get to market quickly.”
The licensing practice began when Qualcomm was young and struggling in the early 1990s, helping its cashflow. At first, the company made the mobile phones as well as developing the underlying technology, but in 1999 it sold its handset division in order to focus on the less tangible—and more lucrative—part of the business. Today, it spends almost $1 billion a year, or 19% of revenue, on R&D. It has amassed 1,800 patents, and 2,200 applications pending.
“We could have very easily said, ‘Let's close up shop, sit back and wait for royalties to come’. But that would have been a short-lived business: the technology evolves very quickly,” says Mr Altman. In August, Qualcomm paid $600m for Flarion, a firm with little revenue but around 100 patents either issued or pending on a new generation of wireless technology. If all goes as planned, this will allow Qualcomm to dominate the next phase of high-speed mobile communications too.
“IP companies can be very profitable. That doesn't mean we are extortionists by any means. If you get the technology right, you get to license it many times,” explains Tudor Brown of ARM, a British firm that creates the intellectual property behind microchips used in nearly all mobile phones and other wireless devices. ARM is the most ubiquitous company no one has ever heard of, with its technology in use in over 70% of all mobile phones. Whereas Qualcomm still keeps a foothold in the physical world by supplying chips, ARM does nothing but R&D and licensing.
But Mr Tudor has a warning for firms that want to concentrate exclusively on the intellectual-property business: licensing usually works only alongside a basket of products or services. For IBM, for example, the majority of its intellectual-property revenue comes from the sale of know-how, not patent licences alone. In essence, the difference is that between the recipe for a dish and a list of ingredients.
ARM was fortunate to be in the right place at the right time, when numerous chip-design firms all wanted to outsource the basic technology for wireless chips so they could innovate on top of it. Also, ARM understood that it needed to offer a lot more than just patent licences, such as documentation and support for its licensees. “They are successful only if they get it right, and we are here to help them get it right,” Mr Tudor says.
Yet it is easy to get it wrong. Take BT, Britain's telecoms incumbent, which in 2000 announced that it had a patent on hyperlinks, a technology that allows people to click on a web-page link to go to another web address. BT claimed it had developed this innovation more than a decade earlier, before the web even existed. Its executives whispered that it was worth billions. The company sent out menacing letters to firms, seeking to enforce its “rights”. But an American judge laughed its claims out of court.
Mr Beyers at HP says he also gets many demands to pay up for infringing someone else's patents. But he is often able to show that the accusers are themselves infringing HP patents, and occasionally ends up getting them to pay him instead.
The Microsoft factor
One of the newest entrants to the intellectual-property game is also the most feared: Microsoft. In the past two years, the company has reshaped its entire strategy around innovation and patents. This will have serious consequences for the rest of the industry.
In 2003, Bill Gates, Microsoft's founder and chairman, faced a number of problems that centred around intellectual property. First, the company found it was being sued for patent infringement more often and had to pay hundreds of millions of dollars in damages. Second, antitrust regulators were forcing Microsoft to open its technology to rivals to allow different systems to work together. Third, the company recognised that its monopoly on its operating system and desktop software would be eroded over time, in part by open-source alternatives, and wanted to delay that process. Lastly, Microsoft was spending around $5 billion a year on R&D and wanted some revenue to help offset that outlay.
Mr Gates summoned the father of IT patent licensing, Marshall Phelps, a cheerful lawyer in his 60s who had recently retired after 28 years with IBM, where he set up and ran the company's highly regarded licensing programme. Mr Phelps told Mr Gates that he had three options: keep the firm's intellectual property within the company to use in its products but make no ancillary revenue from it; assert the company's patents against others (and face the wrath of regulators and rivals); or license it. In June 2003, Mr Gates plucked Mr Phelps out of retirement; by December, Microsoft's intellectual-property licensing division was born.
Since then, in addition to offering technology to other firms, Microsoft has struck a score of cross-licensing deals with big companies. Last May the company started a new business unit, Microsoft Intellectual Property Ventures, to license technology—in areas such as graphics, security and databases—to venture capitalists and start-up firms, sometimes in return for stakes in the companies. That is a new departure: in the past, Microsoft has tried to “cut off the air supply” of rivals, in the celebrated phrase of one Microsoft executive.
Microsoft is coming quite late to the patent game, for the obvious reason that it enjoyed a monopoly in its business area (albeit one judged illegal) for a long time and had little need of patents to protect it. Software copyrights and trade secrets sufficed.
The change in the company's strategy mirrors the wider shifts taking place in the industry as a whole. In 1976, Bill Gates wrote an “open letter to hobbyists”, asking people not to copy the firm's software illegally but to pay for it instead, because that would enable him to fund further improvements in the product. At the time, software was protected by copyright, which automatically comes with any creative work; software patents started to be allowed only in the 1980s, but were rarely sought. When they were first introduced, Mr Gates did not think much of them. In 1991, he wrote in a company-wide memo:
If people had understood how patents would be granted when most of today's ideas were invented and had taken out patents, the industry would be at a complete standstill today...A future start-up with no patents of its own will be forced to pay whatever price the giants choose to impose. What changed, explains Mr Kaefer of Microsoft, was the need to make the technology more open to others, to share it and to interconnect with other firms' technology. “Software patents take the place of trade secrets that we relied on before,” he says. “Software is built on the shoulders of giants—no one can build the whole thing. Patents are a property right that allow the innovation to be exchanged.”
However, there is a controversial side to Microsoft's intellectual-property ambitions. Since 2003, the company has claimed patents on some basic aspects of information technology that many computer scientists say are well established already and should be open to all, such as the “file allocation table” system that computers use to manage files. It has also asserted rights over file formats and so-called “application programming interfaces” that let different types of software interact. These things, critics argue, are not truly intellectual property but merely translations to allow interoperability. Moreover, Microsoft has used its recent multi-billion-dollar antitrust settlements with AOL, Sun, Novell and others to strike broad cross-licensing agreements, thus neutralising potential patent foes.
Taken together, Microsoft seems to be preparing to use its intellectual property as a way to demand a sort of “interconnection fee” from competitors so that their software can interact with Microsoft's (in the same way as a dominant telecoms operator uses interconnection to its network to thwart rivals). The company, suggest the critics, is simply trading its illegal monopoly for a legal one.
No wonder that Microsoft is strongly resisting a requirement by the European Union's antitrust regulators to disclose its networking protocols, which enable different software to interact, arguing that it should be compensated for its intellectual property. This, in particular, hurts open-source software developers, who lack the formal corporate structure (and funds) to license the code. This summer Microsoft took the EU to court over the matter.
In fact, the chief challenge to the company comes from open-source software, whose supporters insist that innovations should be shared, not kept proprietary. Mr Gates calls them “new, modern-day sort of communists” who “don't think that those [intellectual-property] incentives should exist.” And Microsoft has trumpeted that it will indemnify its users against potential patent-infringement liability—which is clearly meant to show up open-source software, where some infringement cases have been brought.
All in all, it looks as though Microsoft is preparing to use intellectual property as a new competitive weapon. Last year, Mr Gates told financial analysts that the firm would increase its patent filings to around 3,000 in 2005, up from 2,000 the year before and the low hundreds in the 1990s. The company currently holds over 6,000 patents, and has around 10,000 applications pending.
At the same time, Microsoft is among the companies most frequently sued for patent infringement: it is currently involved in 32 patent disputes, and spends close to $100m a year in legal costs. Conversely, in the area of copyright, Microsoft's software is the most pirated in the world, causing billions of dollars of potentially lost sales each year.
Industry-wide, there is a danger that big technology firms' infatuation with intellectual property may cause them to exploit their dominant role, forcing smaller firms to pay up or go under. If that happened, only the biggest companies with stacks of patents could hope to survive comfortably. Smaller companies would have to start taking out ever more patents to defend their interests.
To critics of the patent system, this is a sign that the pendulum has swung too far in favour of intellectual-property owners, resulting in an inefficient market for technology. There have been calls for legal reforms and changes in business culture in favour of a more balanced and open approach. But in fact these changes are already taking place—and the impetus is coming not from policymakers or lawyers, but from within the technology industry and from the marketplace itself.