Appeals by the Numbers

Where there is any doubt, then, applicants should be more likely to overcome Section 2(e) descriptiveness refusals than Section 2(d) refusals. This paper tests whether the dramatic difference in standards of proof for 'close' cases impacts the success rate of appeals by reviewing the ex parte decision data from the Trademark Trial and Appeal Board's e-FOIA Final Decisions database. The underlying facts of the individual appeals probably make more of a difference than the standards, which apply only where there is 'doubt,' but it seems reasonable to assume that there was at least enough doubt in each case for the decision to spend the time and money to file and brief an appeal.

While there is less data for appeals of Section 2(a), 2(b), 2(c), or 2(f) data, appeals of final refusals under these sections have been reviewed as 'controls,' since there is no standard of proof that requires weighting of the evidence one way or the other in 'close' cases.

Description of the Data

This article relies on summaries of ex-parte appeal decisions reported on the USPTO's 'Final Decisions of the Trademark Trial and Appeal Board' e-FOIA page. TTAB decisions extracted here were issued between Sept. 4, 1996 and Oct. 26, 2012.

Each application is a single data point. A company could apply for a mark in three classes as a single application or as three applications. Assume the company gets the same refusals for each application. On appeal, it prevails in two classes but not the third. If it had filed multiple applications, the decisions would be coded as one 'affirmed' appeal and two 'reversed' appeals. If it filed one multi-class application, the decision would be coded as a single 'split' decision appeal.

The analysis also excludes genericness refusals from consideration, since the standard applied to Section 2(e) descriptiveness refusals does not apply.

The Data

Comparing All Section 2 Refusal Appeal Success Rates

[IMGCAP(1)]

[IMGCAP(2)]'

Somewhat surprisingly, there is very little difference between the success in appealing Section 2(d) and Section 2(e) refusals. Section 2(e) refusals were overturned in full in 18.5% of cases and in part in a further 1.4%, while Section 2(d) refusals were overturned in full in 16.1% of appeals and in part in a further 1.2%.

Appeals of Section 2(a), 2(b), 2(c), or 2(f) refusals, the 'control' group, were successful over 78% of the time ' insignificantly lower than the Section 2(e) appeal success rate and only a bit lower than the Section 2(d) appeal success rate. The appeal success rate in the 'control' group is mainly driven by Section 2(a) appeals, where almost 36% of appeals were successful.

Comparing Section 2(d) and Section 2(e) Appeal Success'Rates over Time

[IMGCAP(3)]

[IMGCAP(4)]'

The success rate of appeals of Section 2(e) refusals has remained remarkably consistent, hovering between 19% and 21%. There have been no definite or significant trends in the volume of appeals of Section 2(e) refusals.

[IMGCAP(5)]

[IMGCAP(6)]

The success rate of appeals of Section 2(d) refusals has dropped from around 19% ' 20% (1995 to 2004) to around 15% ' 16% (2005 to the present). The number of appeals per year has generally risen, too. The period between 1996 and 1999 averaged 131 appeals per year; the figure jumped to 228 appeals per year between 2000 and 2004, and again to 265 appeals per year from 2005 ' 2009, before dropping to around 215 appeals per year from 2010 to the present. There has been no change in the law that would account for either the change in the number or success rates of appeals.

The success rate may be dropping for three possible reasons. First, the wider adopting of e-filing through ESTTA has resulted in more appeals being filed or appeals being filed less selectively. This seems unlikely to be true because the success rate of Section 2(e) appeals over the same time has not changed. The next two possible reasons are related to the explosion in active registrations: according to the USPTO Data Visualization Center, from 2008 to Q4 2012, the number of active registrations increased from just over 1.4 million to over 1.8 million. More registrations is likely to mean that more marks are facing Section 2(d) refusals than ever before. A second reason for the rise in appeals is that applicants may be increasingly exasperated over the number of Section 2(d) refusals they are getting, prompting more applicants to appeal instead of (or in addition to) simply re-filing for a new mark. Third, the increasing number of registrations make it more likely that applicants' proposed marks are attempting to join more than one prior mark that shares a similar term, thereby increasing the number of times applicants feel that their mark should also be allowed to co-exist in an increasingly crowded field ' a complicated and subjective issue that many applicants may wish to have re-aired before the Board.

Comparing Different Section 2(e) Refusal Appeal Success Rates

[IMGCAP(7)]

[IMGCAP(8)]

Appeals under Section 2(e)(1) were by far the most common, accounting for over 83% of all Section 2(e) appeals. Appeals of Section 2(e)(4) refusals that a mark is primarily a surname and of Section 2(e)(5) refusals that the applied-for matter is functional, at 22.4% and 23.4% success rates, respectively, were slightly more likely to succeed than were appeals of all Section 2(e) refusals, which had an overall success rate of 18%. Appeals of Section 2(e)(3) refusals that the applied-for mark was primarily geographically deceptively misdescriptive were the least successful, succeeding even in part under 14% of the time. The sample sizes for appeals of Section 2(e)(2)-(5) refusals are tiny in comparison to Section 2(e)(1) refusals, and it is possible that that appeal success rates would converge further if there were a larger sample size.

Conclusions

While the Trademark Trial and Appeal Board must resolve any reasonable doubt about a 2(e) refusal in favor of the applicant (by publishing the application for opposition), and it must resolve any reasonable doubt about a 2(d) refusal in in favor of the prior registrant (by refusing to register the application), the standards of proof do not appear to have much impact on appeal success rates. What might this mean?

It is quite possible that the Board views most cases as clear cut, even if the applicants who chose to appeal thought the matters were in enough doubt to justify the time and cost of an appeal. If we assume that the Board's decisions are impacted by the different legal standards in all 'close' cases where the Board has real doubt (and that the different standards only impact 'close' cases), then the approximately 2.4% greater success rate in Section 2(e) appeals than Section 2(d) appeals could have two meanings. First, not many appeals present 'close' cases. Second, the different standards are having a significant impact in 'close' cases. Assuming that only a small subset of appeals are 'close' significantly exacerbates the small-sample-size problem, though; if we assume that only 'close' cases (but every 'close' case) were impacted by the different legal standards, as assessed by the percentage variance in appeal success rate of appeals of Section 2(e) and 2(d) refusals, the data set shrinks from around 5,700 appeals to only around 138 appeals.

If we instead assume that all applicants are assessing their appeals rationally and that only 'close' cases are appealed, the difference in appeal success rates appears to be insignificant. The appellant is quite likely to lose, regardless of what Section 2 refusal grounds are under appeal.

We could further test the hypothesis by reviewing the success rate in overcoming refusals during prosecution before Examining Attorneys. This would provide a much larger data set, and perhaps provide some insight into whether a refusal is more likely to be overcome in prosecution than on appeal. Unfortunately, examination data is not available in a systematic database format that allows for an easy comparison of success rates.

Beyond measuring the impact of the different standards of proof for appeals of Section 2(d) and Section 2(e) refusals, it's worth emphasizing that four in five appeals of all Section 2 refusals fail completely. This data should lead careful practitioners to counsel clients that their appeal is unlikely to succeed, even if there are some apparently compelling arguments in their client's favor. As a result, having discussions about the appeal and how to carefully prepare the record for appeal are not enough. The client also needs to be given concrete advice about what additional actions it should be taking immediately, operating under the assumption that the appeal will probably fail. These suggestions could include considering alternative filing strategies, surveys (for Section 2(e) refusals), investigations or cancellation actions (for Section 2(d) refusals), or clearing and filing for an alternative mark as a backup.

'

Matt Schneller is an Associate and Erin Hennessy is a Partner at Bracewell & Giuliani LLP in Seattle, WA.

'

Two of the main substantive causes for refusals of U.S. federal trademark applications are descriptiveness refusals (under Section 2(e) of the Lanham Act, 15 U.S.C. '1052(e)(1)) and likelihood of confusion refusals (under Section 2(d) of the Lanham Act, 15 U.S.C. '1052(d)). The U.S. Patent & Trademark Office's examiners and the Trademark Trial and Appeal Board's judges, who review ex parte appeals from examiners' final refusals to register pending applications, apply very different standards as 'tiebreakers' when the evidence and arguments submitted by the applicant against the two types of refusals create doubt about the appropriate outcome.

Any reasonable doubt as to whether a mark is merely descriptive or contains any suggestive element must be resolved in favor of the applicant. The rule grew out of Section 2(e)(1) refusals of merely descriptive marks, see, e.g., In re Aid Laboratories, Inc., 221 USPQ 1215 (TTAB 1983), In re MVP Group Int'l, Ser. No. 85074276 (TTAB May 16, 2012), and also applies to applies to 2(e)(2) refusals that the mark is primarily merely geographically descriptive,'see, e.g., In re International Taste Inc., 53 USPQ2d 1604, 1605-06 (TTAB 2000), and to 2(e)(4) that the mark is primarily merely a surname, see, e.g., In re Yeley, 85 USPQ2d 1150, 1151 (TTAB 2007); In re Benthin Mgmt. GmbH, 37 USPQ2d 1332, 1333-1334 (TTAB 1995). The standard of proof does not apply to Section 2(e)(3) refusals that a mark is primarily geographically deceptively misdescriptive or to 2(e)(5) refusals that the matter applied for is functional.

In contrast, if there is any reasonable doubt as to whether there is a likelihood of confusion with a prior registration, that doubt must be resolved in favor of the prior registrant. See, e.g., In re Hyper Shoppes (Ohio), Inc., 837 F.2d 463, 464-65, 6 USPQ2d 1025, 1026 (Fed. Cir. 1988); In re G.B.I. Tile & Stone, Inc., 92 USPQ2d 1366, 1372 (TTAB 2009).

Where there is any doubt, then, applicants should be more likely to overcome Section 2(e) descriptiveness refusals than Section 2(d) refusals. This paper tests whether the dramatic difference in standards of proof for 'close' cases impacts the success rate of appeals by reviewing the ex parte decision data from the Trademark Trial and Appeal Board's e-FOIA Final Decisions database. The underlying facts of the individual appeals probably make more of a difference than the standards, which apply only where there is 'doubt,' but it seems reasonable to assume that there was at least enough doubt in each case for the decision to spend the time and money to file and brief an appeal.

While there is less data for appeals of Section 2(a), 2(b), 2(c), or 2(f) data, appeals of final refusals under these sections have been reviewed as 'controls,' since there is no standard of proof that requires weighting of the evidence one way or the other in 'close' cases.

Description of the Data

This article relies on summaries of ex-parte appeal decisions reported on the USPTO's 'Final Decisions of the Trademark Trial and Appeal Board' e-FOIA page. TTAB decisions extracted here were issued between Sept. 4, 1996 and Oct. 26, 2012.

Each application is a single data point. A company could apply for a mark in three classes as a single application or as three applications. Assume the company gets the same refusals for each application. On appeal, it prevails in two classes but not the third. If it had filed multiple applications, the decisions would be coded as one 'affirmed' appeal and two 'reversed' appeals. If it filed one multi-class application, the decision would be coded as a single 'split' decision appeal.

The analysis also excludes genericness refusals from consideration, since the standard applied to Section 2(e) descriptiveness refusals does not apply.

The Data

Comparing All Section 2 Refusal Appeal Success Rates

[IMGCAP(1)]

[IMGCAP(2)]'

Somewhat surprisingly, there is very little difference between the success in appealing Section 2(d) and Section 2(e) refusals. Section 2(e) refusals were overturned in full in 18.5% of cases and in part in a further 1.4%, while Section 2(d) refusals were overturned in full in 16.1% of appeals and in part in a further 1.2%.

Appeals of Section 2(a), 2(b), 2(c), or 2(f) refusals, the 'control' group, were successful over 78% of the time ' insignificantly lower than the Section 2(e) appeal success rate and only a bit lower than the Section 2(d) appeal success rate. The appeal success rate in the 'control' group is mainly driven by Section 2(a) appeals, where almost 36% of appeals were successful.

Comparing Section 2(d) and Section 2(e) Appeal Success'Rates over Time

[IMGCAP(3)]

[IMGCAP(4)]'

The success rate of appeals of Section 2(e) refusals has remained remarkably consistent, hovering between 19% and 21%. There have been no definite or significant trends in the volume of appeals of Section 2(e) refusals.

[IMGCAP(5)]

[IMGCAP(6)]

The success rate of appeals of Section 2(d) refusals has dropped from around 19% ' 20% (1995 to 2004) to around 15% ' 16% (2005 to the present). The number of appeals per year has generally risen, too. The period between 1996 and 1999 averaged 131 appeals per year; the figure jumped to 228 appeals per year between 2000 and 2004, and again to 265 appeals per year from 2005 ' 2009, before dropping to around 215 appeals per year from 2010 to the present. There has been no change in the law that would account for either the change in the number or success rates of appeals.

The success rate may be dropping for three possible reasons. First, the wider adopting of e-filing through ESTTA has resulted in more appeals being filed or appeals being filed less selectively. This seems unlikely to be true because the success rate of Section 2(e) appeals over the same time has not changed. The next two possible reasons are related to the explosion in active registrations: according to the USPTO Data Visualization Center, from 2008 to Q4 2012, the number of active registrations increased from just over 1.4 million to over 1.8 million. More registrations is likely to mean that more marks are facing Section 2(d) refusals than ever before. A second reason for the rise in appeals is that applicants may be increasingly exasperated over the number of Section 2(d) refusals they are getting, prompting more applicants to appeal instead of (or in addition to) simply re-filing for a new mark. Third, the increasing number of registrations make it more likely that applicants' proposed marks are attempting to join more than one prior mark that shares a similar term, thereby increasing the number of times applicants feel that their mark should also be allowed to co-exist in an increasingly crowded field ' a complicated and subjective issue that many applicants may wish to have re-aired before the Board.

Comparing Different Section 2(e) Refusal Appeal Success Rates

[IMGCAP(7)]

[IMGCAP(8)]

Appeals under Section 2(e)(1) were by far the most common, accounting for over 83% of all Section 2(e) appeals. Appeals of Section 2(e)(4) refusals that a mark is primarily a surname and of Section 2(e)(5) refusals that the applied-for matter is functional, at 22.4% and 23.4% success rates, respectively, were slightly more likely to succeed than were appeals of all Section 2(e) refusals, which had an overall success rate of 18%. Appeals of Section 2(e)(3) refusals that the applied-for mark was primarily geographically deceptively misdescriptive were the least successful, succeeding even in part under 14% of the time. The sample sizes for appeals of Section 2(e)(2)-(5) refusals are tiny in comparison to Section 2(e)(1) refusals, and it is possible that that appeal success rates would converge further if there were a larger sample size.

Conclusions

While the Trademark Trial and Appeal Board must resolve any reasonable doubt about a 2(e) refusal in favor of the applicant (by publishing the application for opposition), and it must resolve any reasonable doubt about a 2(d) refusal in in favor of the prior registrant (by refusing to register the application), the standards of proof do not appear to have much impact on appeal success rates. What might this mean?

It is quite possible that the Board views most cases as clear cut, even if the applicants who chose to appeal thought the matters were in enough doubt to justify the time and cost of an appeal. If we assume that the Board's decisions are impacted by the different legal standards in all 'close' cases where the Board has real doubt (and that the different standards only impact 'close' cases), then the approximately 2.4% greater success rate in Section 2(e) appeals than Section 2(d) appeals could have two meanings. First, not many appeals present 'close' cases. Second, the different standards are having a significant impact in 'close' cases. Assuming that only a small subset of appeals are 'close' significantly exacerbates the small-sample-size problem, though; if we assume that only 'close' cases (but every 'close' case) were impacted by the different legal standards, as assessed by the percentage variance in appeal success rate of appeals of Section 2(e) and 2(d) refusals, the data set shrinks from around 5,700 appeals to only around 138 appeals.

If we instead assume that all applicants are assessing their appeals rationally and that only 'close' cases are appealed, the difference in appeal success rates appears to be insignificant. The appellant is quite likely to lose, regardless of what Section 2 refusal grounds are under appeal.

We could further test the hypothesis by reviewing the success rate in overcoming refusals during prosecution before Examining Attorneys. This would provide a much larger data set, and perhaps provide some insight into whether a refusal is more likely to be overcome in prosecution than on appeal. Unfortunately, examination data is not available in a systematic database format that allows for an easy comparison of success rates.

Beyond measuring the impact of the different standards of proof for appeals of Section 2(d) and Section 2(e) refusals, it's worth emphasizing that four in five appeals of all Section 2 refusals fail completely. This data should lead careful practitioners to counsel clients that their appeal is unlikely to succeed, even if there are some apparently compelling arguments in their client's favor. As a result, having discussions about the appeal and how to carefully prepare the record for appeal are not enough. The client also needs to be given concrete advice about what additional actions it should be taking immediately, operating under the assumption that the appeal will probably fail. These suggestions could include considering alternative filing strategies, surveys (for Section 2(e) refusals), investigations or cancellation actions (for Section 2(d) refusals), or clearing and filing for an alternative mark as a backup.

'

Matt Schneller is an Associate and Erin Hennessy is a Partner at Bracewell & Giuliani LLP in Seattle, WA.

'