One of the advantages of IFI CLAIMS Patent Services is the quality of our data – the result of years of editorial work that evolved to support advanced patent analytics. We work closely with our partners and clients to enable access to patent data that analysts transform into valuable insights into patent portfolios, technology trends, and IP market insights.
The yin yang between patent quantity and quality has long been an issue among innovative businesses. It has yet to be resolved. It is not a simple matter of those with a lot of patents must have more questionable ones, but high numbers do tend to beg the question: which patents are really important and to whom? In addition to claims, factors like industry, timing, need and perception play an even increasing role in patent quality and value. – Bruce Berman, a principal in Brody Berman Associates, an IP communications firm that advises patent holders and managers and helps to position IP assets.
A recent analysis of the legal representatives associated with US patents demonstrated the importance of data quality while raising some interesting questions about exactly how we evaluate our top US IP law firms. A few Google searches, and you will find a several ranked lists of the top patent firms. Averyindex.com recently published a list of the top 100 patent firms by the number of patent attorneys with Finnegan Henderson Farabow Garrett & Dunner (Finnegan) taking the top spot with 276 Patent Attorneys or Agents. http://www.averyindex.com/top_patent_firms.php
Similar to how IFI ranked the Top US Assignees, we ranked the top US firms based on the number of US utility patents granted in 2011 for which they were the legal representative. Here are the 2011 Top 50 Patent Law Firmsthat account for over 23% of the 2011 US granted patents.
Comparing our results to other lists and counts at the USPTO surface inevitable anomalies. One real puzzle is why several firms like Nixon & Vanderhye, that ranked #11 on our list, didn't appear on other lists at all? Most of the data differences are likely the result of spelling variations. Like assignee names in patent data, legal representative data needs to be cleaned in order to accommodate spelling or name variations. For example in the case of Birch Stewart Kolasch and Birch, there are 29 name variations for 2011 alone.
In cases where the patent counts are higher than those reported by IFI CLAIMS, it is possible that source aggregates data from an acquisition or from some sort of subsidiary. Without clear explanations or reasonable access to the underlying it is nearly impossible to resolve these data discrepancies. However at IFI CLAIMS we'll dig as deep as needed to explain and expose the differences.
The top patent law firm for 2011 is Oblon Spivak McClelland Maier & Neustadt (Oblon) with 5613 US granted patents. Finnegan is ranked #7 with 2089 US granted patents.
Averyindex.com ranks Oblon #22 with only 82 registered patent attorneys and agents. How does one explain the dramatic discrepancy between the performance of Finnegan vs Oblon when you consider the number of patent grants per attorney?
For both firms their top 10 assignees are almost all non-US companies and these clients account for almost 60% of their 2011 grants. We could dig further into the priority information, but it is likely that many of the applications are written off-shore, translated, and reviewed and filed by the US firms.
Finnegan | Oblon | |||
Assignee | Grants | Assignee | Grants | |
Toshiba Corp JP | 338 | Toshiba Corp JP | 1141 | |
Sony Corp JP | 179 | Ricoh Co Ltd JP | 571 | |
Caterpillar Inc | 152 | Sony Corp JP | 502 | |
AOL Inc | 111 | Tokyo Electron Ltd JP | 264 | |
Toyota Jidosha K K JP | 109 | Toyota Jidosha K K JP | 217 | |
L'Oreal S A FR | 82 | Mitsubishi Denki K K JP | 215 | |
SAP AG DE | 66 | NTT DoCoMo Inc JP | 126 | |
Research In Motion Ltd CA | 53 | BASF SE DE | 115 | |
Huawei Technologies Co Ltd CN | 52 | SNECMA FR | 104 | |
Telecom Italia SpA IT | 51 | Asahi Glass Co Ltd JP | 94 |
Taking another perspective we can look at single assignee like Toshiba Corp and see that there 2011 patent filings were managed by 40 different legal representatives. Even this simple report starts off with almost 90 representatives more than half of which are name variations on the 40 firms. Oblon accounts for 45% of their 2011 grants and the following 10 firms account for over 90%:
Oblon, Spivak, McClelland, Maier & Neustadt, L.L.P. | 1141 |
Finnegan, Henderson, Farabow, Garrett & Dunner, L.L.P. | 338 |
Turocy & Watson, LLP | 318 |
Knobbe, Martens, Olson & Bear LLP | 127 |
Patterson & Sheridan, LLP | 119 |
Blakely, Sokoloff, Taylor & Zafman LLP | 90 |
Pillsbury Winthrop Shaw Pittman, LLP | 69 |
Nixon & Vanderhye, P.C. | 57 |
Foley & Lardner LLP | 56 |
Ohlandt, Greeley, Ruggiero & Perle, L.L.P. | 43 |
For both Oblon and Finnegan their top 20 US classes (to 3 digits) account for almost 50% of the granted patents. However the firms have only 10 common classes in their respective top 20 which of course suggests some level of specialization.
Finnegan | Oblon | |||
US CLASS | Grants | US CLASS | Grants | |
705 | 99 | 257 | 326 | |
438 | 91 | 399 | 244 | |
709 | 79 | 365 | 195 | |
257 | 76 | 386 | 191 | |
370 | 73 | 438 | 171 | |
707 | 71 | 428 | 168 | |
514 | 59 | 430 | 133 | |
382 | 59 | 370 | 130 | |
455 | 58 | 375 | 126 | |
008 | 51 | 348 | 121 | |
424 | 47 | 382 | 116 | |
701 | 46 | 455 | 112 | |
345 | 42 | 701 | 110 | |
060 | 40 | 514 | 110 | |
435 | 36 | 435 | 91 | |
600 | 33 | 429 | 87 | |
704 | 32 | 358 | 82 | |
348 | 31 | 369 | 79 | |
715 | 28 | 709 | 70 | |
430 | 28 | 359 | 69 |
(note: you can look up US patent classes at: www.uspto.gov/web/patents/classification/)
You can see certainly see further evidence of specialization such as class 705 (DATA PROCESSING: FINANCIAL, BUSINESS PRACTICE, MANAGEMENT, OR COST/PRICE DETERMINATION) where Finnegan was involved in more than twice the 2011 grants than Oberon.
And as you can continue to drill down into the data examining correlations between inventors, classification, claims counts, latency between filing and grant, citation counts, etc. Often claims counts are presented as a factor in patent valuation. Average claims counts vary a great deal depending on the domain, but they may also contribute to insights into the performance or efficiency of one firm vs another. Why do Oblon’s patents average 13.7 claims per patent while Finnegan’s average almost 17.4?Good insight into to questions like these depend on deeper analysis that requires high quality data. Modern data analytics requires rich interfaces like CLAIMS Direct and a team of experts that can help you define the dimensions of your problem getting you the data you need when you need it.
Check out our webinars or contact us to learn more about how IFI CLAIMS Patent Services can help you streamline your patent analysis with quality patent data.