Myles Harrington, president and co-founder of Grant Street Group, is either a brilliant businessman or an extortionist, depending upon whether you stand to benefit or suffer from the patent he holds.
Last December, Harrington’s Pittsburgh-based company won a patent on a system to conduct online auctions of new issues of fixed-income instruments. Having spent nearly four years waiting for the patent from the United States Patent and Trademark Office (USPTO), Harrington intends to enforce it. Of the several companies he has approached about licensing the technology, he claims that one, financial information giant Thomson Financial, has instead copied it. Thomson’s competing system, Parity, which processed more than 350 municipal bond sales last year, directly infringes on his patent, says Harrington, who has sued as a result.
“We’ve been in business since 1997,” he says, “and Thomson employees hit our site over 1,000 times prior to launching Parity. I can tell you that’s extraordinary behavior. When small companies like us innovate, and the barriers to entry are low, we don’t have a chance unless we can protect our ideas.”
Says a Thomson spokeswoman: “We believe this suit has no merit, and we are confident that the outcome of these proceedings will find in our favor.”
The case might seem cut and dried if Harrington’s patent covered the manufacturing process of a widget or some other end product. But because it applies to a method not of making something but of doing business, as well as to the computer technology behind it, Grant Street’s protection falls into the nebulous world of business method patents, the fastest-growing type — and the most controversial.
The holders of these patents say they are simply defending their hard-earned intellectual property and deserve no less protection than any other inventor. But critics assert that many business method patents are granted for inventions that simply computerize obvious techniques or prior inventions (such as municipal bonds and auctions), and therefore shouldn’t qualify.
Business method patents still make up less than 4 percent of the total number of patent applications the USPTO receives annually. But the number of business method patent applications received and granted by the office grew from 2,700 and 583, respectively, in 1999 to 7,900 and 899 in 2000, according to John Love, the group director overseeing business method patent examinations at the USPTO. He anticipates 12,000 such applications in 2001.
Despite this growth, the USPTO doesn’t strictly define business method patents. According to Love, his division examines Class 705 applications, which have to do with the computer-based manipulation of figures and information in the management or conduct of a business or enterprise, specifically dealing with the Internet, points of sale, or exchange of goods. To be considered for any patent, he says, an application must meet three criteria: It must be nonobvious, innovative, and useful. But critics claim that an overworked, understaffed, underfunded USPTO is unable to distinguish between what is innovative and what is common knowledge. Not so, responds Love, who explains that when a business method patent is approved, the technology itself may not be new, but its effect is.
In any case, these patents are allowing more companies to collect licensing fees and file infringement lawsuits. The USPTO estimates that the number of patent lawsuits filed in U.S. courts topped 2,000 last year, up roughly 25 percent from 1997. And patent holders are being helped by a new industry devoted to the patents’ purchase and enforcement. Companies like TechSearch LLC, based in Northbrook, Illinois, buy patents from holders that can’t afford to enforce the patents themselves, and then go after infringers, giving inventors a cut of the proceeds.
Infringement lawsuits aren’t cheap. According to a survey by the American Intellectual Property Law Association, the median cost of litigating a patent infringement case in which $1 million to $25 million is at stake is $1.5 million, with the top 25 percent of cases costing an average $2.5 million. And the stakes can be huge. Awards or settlements in business method patent infringement cases now reach into the hundreds of millions of dollars.
The Internet, not surprisingly, deserves part of the credit (or blame) for the rise of business method patents. Much of the software source code now widely used on the Web was anonymously created by developers under the “open source” ethic, in which software code was exchanged freely for the good of the Internet’s advancement. This wasn’t an issue until 1998, when the first business method patent was upheld in court. In a ruling in State Street Bank v. Signature Financial Group, the court found that Signature’s computer program to track mutual funds had a valid patent.
“Prior to the State Street ruling,” says Love, “conventional wisdom was that mathematical ideas were nonpatentable, because the laws of nature, abstract ideas, and anything that’s naturally occurring aren’t eligible for patent protection.”
Because very few of the previous innovations in this area were ever documented, there is little proof that an idea or program existed prior to a patent application. Finding such so-called prior art is the only way to invalidate a patent. The USPTO — which must search prior-art databases, including professional journals, before granting a patent — is actively seeking public input on new prior-art database sources that deal with business methods so that it can do a more thorough job of evaluating the validity of the patents. But unless the prior art is publicly available, which excludes most information related to the genesis of the Internet, finding it is a challenge.
The USPTO’s database efforts are just part of a reform program, begun in March 2000, that also set up a review of every new business method patent application and created an advisory committee. Andy Gibbs, CEO of PatentCafe.com Inc., in Sacramento, and a charter member of the committee, concedes that the USPTO’s reform efforts have made it more difficult to receive a business method patent today than it was three years ago. But, he adds, “that doesn’t reverse the fact that many questionable patents have been issued. Many of them still force companies to defend against infringements.”
In particular, detractors cite patents held by Amazon.com on a “one-click” Web shopping technique, Sightsound .com on a process for selling music online, and Intouch on a method for sampling music, as examples of patent grants on overly broad and obvious concepts. (Amazon’s injunction against Barnes&Noble.com’s one-click shopping function was recently lifted by the U.S. Court of Appeals, which said that B&N’s case “raised substantial questions” about the patent’s validity.)
Consultants contend that many companies nonetheless don’t realize that they are sitting on a gold mine. “Many have a lot of patents that they haven’t looked at in a while,” says Kevin G. Rivette, CEO of Cupertino, California-based patent management company Aurigin Systems Inc. and the co-author (with David Kline) of Rembrandts in the Attic: Unlocking the Hidden Value of Patents (Harvard Business School Press, 1999).
Rivette credits savvy patent management for the turnarounds at Apple Computer, IBM, and Texas Instruments. He says that Wall Street analysts are paying closer attention as a result, citing a recent buy recommendation on Universal Electronics Inc., makers of programmable remote controls, based on the strength of its patent portfolio.
The challenge for companies out to mine their intellectual property for patent potential is to “protect those methods that are differentiators, that make your business better than your competitors’,” says Rivette. For example, IBM, whose patent portfolio generated a whopping $1.7 billion in licensing fees in 2000 — or 2 percent of Big Blue’s total revenue — uses what are called “picket fence” patents. It files for one patent, but then publicly discloses all the obvious improvements that someone else might try to patent around the original filing. By publicly disclosing those improvements in separate documents, it has “created” prior art without going through the laborious process of additional patent applications.
Love says it takes longer for the USPTO to grant a business method patent — about 30 months — than it does for some other types, because of the new review process. When requests for licensing fees from alleged infringers don’t work after a patent has been granted, more and more companies are calling in the big guns, like Ray Niro, a patent attorney with Niro, Scavone, Haller & Niro, an intellectual property law firm that represents Grant Street Group’s Harrington as well as TechSearch.
Averse to such conflict? It’s common whenever new technology comes along, says Rivette. “Patents have always been called extortion,” he notes. “And they’ve always been called that by people who don’t have them. Steamboats, the telegraph, lightbulbs…with every major technology shift, there’s always a period of time when things are really mushy. We’ll see a lot of chaff come out of the system in the next couple of years.”
That’s cold comfort to companies that are currently facing lawsuits over dicey patents, or that have already paid licensing fees under them. Even if the patents are overturned, those fees are unrecoverable unless a license agreement stipulates that a refund is due.
Kris Frieswick is a staff writer at CFO.
With the advent of business method patents, thousands of companies that thought they were using technology or concepts that were in the public domain are now getting slapped with infringement notices. And patent infringement is difficult to defend against, says Charles Cella, CEO and founder of Boston-based BountyQuest Corp., which acts as a middleman for awards to anyone who finds publicly available documentation related to a disputed patent. Such documentation is known as prior art.
“Aside from proving you’re not using a similar technology or process,” says Cella, “the only way to [invalidate a patent] is by finding prior art.” Cella notes that while most people search databases of prior art to find out whether a business method has been patented before, they aren’t finding out if it has been used before. “We saw an enormous gap between these two questions,” he says.
So for $2,500, a company that is being sued for infringement can now list the patent on the BountyQuest Web site and post a reward for valid prior art. If prior art is found, which has happened for 15 percent to 20 percent of the postings, the finder gets $10,000 and BountyQuest gets a success fee. (Since BountyQuest’s inception in October 2000, at least $1 million worth of awards have been posted on the site.) Ironically, in light of Amazon’s infringement suits over the business method patent it holds for its “one-click” online shopping technology, company founder Jeff Bezos is a partner in BountyQuest and recently wrote an open letter stating his concerns with the process for granting such patents.
All of which raises a question: When it reviews a patent application, why doesn’t the United States Patent and Trademark Office do what BountyQuest does? John Love, the group director overseeing such patent examinations at the USPTO, blames confidentiality concerns. “It’s dangerous to just call up a company and find out if they’re doing anything like it,” he says. “It’s a tip to what application we’re working on.” — K.F.