Patent Searching - Better Safe than Broke
Kurt O. Baumgartner - President,
The Devine Corporation

Do a search, don't do a search or maybe have a search done. It costs too much to have a search done and the economics of a search don't make having it done worthwhile. But, the economics of not doing a search could be more severe. Since most inventor's deal with a patentability search, finding out if it is already patented, remarks made by Curtis Harrington (Inventor's Digest, Jan-Feb, 1998) are very important. He states, on patentability searches, the cost will be at least $1,000 and it is not profitable to incur those costs considering the additional $2,900 in filing costs. I feel that those comments are misleading and his search cost figures are highly inflated. What is the inventor to do? Who should they believe? These are good questions with no real hard and fast answers. Hopefully this article will aid inventors in making the appropriate choices.

Most inventors do not have just a single idea. While each idea is unique, it is only one of many that the inventor produces. In fact, many inventors carry notebooks with them to write down their ideas as they occur. The truth is, most of these ideas never get past the notebook stage and most of the ones that do never get to the patent stage. This is why it is important for inventors to know just what they are getting into when they start the patent process.

Economics is the key. It costs money to patent an invention. As stated in Harrington's article, the filing cost of $2,900 is just the beginning of the process and the costs. There are patent office and attorney fees all along the process with a substantial fee required when the patent office says it is going to issue your patent. Failing to pay that fee means no issuance and no patent. Then, at 32, 72 and 112 years, a maintenance fee is due or the patent expires. A cost of $5000 for a utility patent from start to finish, including attorney fees, is not unusual and most often a fair approximation of the average cost. With a first office rejection almost automatic for every patent filed, to expend another $1,000 for a patent search would seem to be a little excessive. If a rejection is going to occur anyway, why would anyone waste time with a patent search?

The obvious answer is that it is not a waste of time or money. In order to write an application a patent attorney must write acceptable claims concerning the invention. Poorly written claims are one reason for a first office rejection. Another is that the claims are already covered by previous works of art (patents-U.S. and foreign). If the attorney has already read the previous works, claims can be written that will not infringe on existing work and reduce the chances of a first office rejection. Writing claims without reviewing the prior patent literature is almost guaranteeing a first office prior art rejection. For the inventor, a prior art search can mean saving time and effort. As mentioned, inventors have many ideas, the one under patent consideration may seem most promising. However, if it looks like the patent literature, either US or foreign, seems to cover the idea well, it may not be worth the additional time and expense to even try to file an application. A patent may get issued eventually, but the claims will be so restricted as to have little value. There is no reason to go through the attorney fee and filing expense to have the patent office indicate that there is nothing to patent when that could have been determined prior to filing. It makes more sense to concentrate on those ideas, which have greater potential.

Determining the prior art is not that difficult and does not have to cost anywhere near the $1000 indicated by Mr. Harrington. But, determining what needs to be searched can be as important as how much or where to look. The patent office is going to conduct its own search of an invention. The office will search, at a minimum, the US patent database that they maintain and the foreign application/patent database compiled by Derwent, Inc. In the case of some scientific areas, they may search appropriate literature databases to find previously published articles. The most critical areas of prior art are US and foreign patents, not the worldwide accumulation of knowledge mentioned by Mr. Harrington. A good faith effort to search this area of prior art then dealing with any items found, in the application or in the information disclosure statement, often reduces or eliminates an examiner's prior art opposition to the application. Examiners tend to look more favorably at applications that have done an adequate prior art search.

The how and where to do an adequate prior art search is a problem for many inventors. Obviously, the best place to perform a prior art search is at the patent office. Not only the US patent office, but at the patent office of every country in the world. This, of course, is not practical. That is why online databases and the Internet have become so popular. But where and how to look can be time consuming, confusing and costly. The following is a brief discussion concerning where inventors can go to search and why in some cases it is better to let an information provider do the search for you.

Free Searching - US Patents

If you live close to a patent repository library, you can go to the government documents section of the library and do a manual search using CASSIS. This allows you to search the information provided in the official patent gazette. CASSIS is limited in its scope and ability to find relevant patents. However, if you find relevant patents, copies should be available right there on microfilm. The USPTO can provide you with a list of repositories in the US. This is a beginning.

You may not find anything here. The next step is to look online at either the USPTO Website (http://www.uspto.gov) or the IBM Patent Website (http://www.ibm.patent.com). Both of these sites allow you to search US patents back to 1976. However, you can search only limited parts of the patent: title, abstracts, inventors, assignees and citations. Knowledge of Boolean logic (And, Or, Not) and nesting is beneficial. The search engines are not sophisticated and the tendency is to get a large number of irrelevant records. Relevancy is based upon how often a search word(s) appears in one record in relation to the other records not whether the patent is pertinent to your topic. A great deal of time is spent looking at many patents to find the few that deal with the actual topic. If relevant records are found, the IBM site can provide copies of the images (figures and drawings) from the patents. For mechanical and design patents, this can be very helpful in determining whether the patent covers your invention. Beginning in 1999, the USPTO has plans to make the fulltext of patents searchable and include access to images. This is as far as you can go searching US patents without incurring some searching costs.

Foreign Patents

This area of searching is much more difficult. The only free site is the World Intellectual Property Office or PCT office site (http://pctgazette.wipo.int/eng). Most foreign applications are first filed with this office. The inventor can indicate which foreign countries, which belong to the Patent Cooperation Treaty, they intend to file. While these are just applications, the invention and claims are outlined. This database only covers 1998 and again has limited searchable fields. This does not cover foreign patent applications not filed in the PCT office.

Fee Based: US Patents Only

Patent Miner: An Internet based (http://www.patentminer.com) service that allows you to search twenty-five years of US patents full text. It utilizes a basic search engine typical of the Internet. While Boolean logic is supportable, it is still a basic search format. You can get patent copies from them, see the full text of the document and hypertext over to the IBM site to see the images. As with any Internet based services, some days it is fast, some days slow and some days you can't get to it at all. Fee: There is a cost for this that ranges from $12.95 (single concept search) to $49.95 (Unlimited searching for a 12-hour period)

Corporate Intelligence: A non-Internet service that requires special software, which allows you to connect directly into their serve. They have US patents fulltext back to 1945. They have a more sophisticated search engine than the Internet providers. However, if you wish to do a sophisticated search with nesting and proximity, the search can get to be very complicated. You can receive patent copies via e-mail or regular mail at $3 each. However, retrieving a large number of patents via e-mail can cause problems if your system has limited space. A short patent can run 1 megabyte while a very long biotech patent can be over 50 megabytes. Fee: The software costs and you can pay as you search or sign up for a variety of monthly pay plans.

US and Foreign Patents

Dialog & STN: these are two different for fee services. Both have US patents full text back to 1971 and claims back to 1950. They also have a variety of foreign patent databases such as the Derwent World Patent Index and JAPIO (Japanese patents) among others. They have very sophisticated search engines which allow for great flexibility in Boolean logic and proximity searching. While these are 'pay-as-you-use' services, there is a nominal initial set-up or account fee. However, there is no minimum usage and you pay only for what you use. All information providers discussed in the next and the USPTO will use one or both of these services.

Information Providers/Brokers

There are a number of these services available. Some can be found in the classified section of this magazine. You can also contact your local inventor's group or chamber of commerce as they may have names of appropriate companies or individuals. Also, fellow inventors may be able to recommend a service they have used. The purpose of the information providers is simple. For a fee, they will search the online patent literature to see if the inventor's idea has already been patented or for foreign countries, applied for. This is a subject or term search using the terms and concepts that best describe the invention. If an appropriate class has been determined for the invention, a search of that class can be made. Some of these services will also perform a physical search of the patent office files for patents similar to the invention that date prior to the online databases. Depending upon how detailed an inventor wants the searcher to go, a search can run from $100 for US only to $1000+ if a manual search is completed at the patent office. Normally, a search of the online US and foreign patent databases should run $200-$300. This price varies due to how much is being searched, what kind of results the client will receive, and how large the search firm is (overhead costs). Costs need to be discussed and fully understood with the provider before a search is performed. Be prepared to pay part or the entire search cost up front.

Conclusion

Much has been said that about spending $500 or $1000 to do a patent search when all that is going to be covered is a small portion of the worldwide literature. And any publication from that multitude of literature can be declared prior art and prevent your patent from being issued. That is true and because it is true, an inventor could forego a patent search. The inventor will still spend the $2,000-$3,000 for the attorney to prepare and submit a patent application. Then, when the patent office rejects the patent application, and it will, because it found prior art, the attorney must argue that what was found is not relevant and/or change the claims. The result will be thousands more in attorney and re-filing fees. If the application is dropped, the inventor will loose most or all the their money. The patent office keeps some or all of the application fee. And unless there is an agreement with the attorney to return his fees, none of that is returned. In addition, it has taken the USPTO about a year to provide that negative opinion. One year and a couple thousand dollars down the drain. To spend $200-300 to find the obvious prior art patents and maybe prevent spending the $2900, plus waste a year, seems to be a very profitable trade-off. If nothing is found or similar patents are found, the attorney can 1) write claims to avoid obvious prior art problems and 2) explain those away in the application so that the examiner sees why they are not relevant. A much better initial application should result. If patents are found that are right on target of the invention, the invention is dropped there and then. The result is minimal costs and time. While a patent search will not guarantee clear sailing and no rejection, it will increase chances of getting through a first office action without a prior art rejection. I believe that is money and time well spent.