Rimini Street, Inc. v. Oracle International Corporation

Opinion

3 UNITED STATES DISTRICT COURT

4 DISTRICT OF NEVADA

5 * * *

6 ORACLE INTERNATIONAL Case No. 2:14-cv-01699-MMD-DJA CORPORATION, et al., 7 ORDER Plaintiffs, 8 v.

9 RIMINI STREET, INC., et al.,

10 Defendants.

11 12 I. SUMMARY 13 This was a software copyright and unfair competition dispute between Plaintiffs 14 and Counter Defendants Oracle America, Inc., and Oracle International Corporation 15 (collectively, “Oracle”) and Defendants and Counter Claimants Rimini Street, Inc., and 16 Seth Ravin (collectively, “Rimini”) generally regarding Rimini’s unauthorized copying of 17 Oracle’s enterprise software into and from development environments created by Rimini 18 for its clients, along with disputes regarding allegedly false statements in marketing and 19 advertising and unfair competition. (ECF Nos. 1253 at 2, 1305 at 12-13.) Following a 20 bench trial, the Court mostly—but not entirely—found in Oracle’s favor and entered a 21 permanent injunction against Rimini. (ECF Nos. 1536 (“Bench Order”), 1537 (the 22 “Injunction”), 1538 (“Judgment”).) Before the Court is Oracle’s motion for attorneys’ fees 23 and costs seeking $70,602,364.20. (ECF No. 1560 (“Motion”); see also id. at 30.)1 24

25 1This docket entry contains redactions. The sealed, unredacted version of this motion was docketed at ECF No. 1568. The redacted version of Rimini’s response was 26 docketed at ECF No. 1588, and the sealed, unreacted version was docketed at ECF No. 1595. The redacted version of Oracle’s reply was docketed at ECF No. 1604 and the 27 sealed, unredacted version was docketed at ECF No. 1607. Some of the exhibits 1 Because Oracle has shown it is entitled to fees and costs, but the portion of its fees 2 request seeking recovery of fees incurred to prepare for a jury trial Oracle decided to 3 forego is slightly excessive, the e-discovery costs it seeks are not recoverable, and as 4 further explained below, the Court will grant in part, and deny in part, the Motion. 5 II. DISCUSSION 6 The Court first addresses whether Oracle is entitled to recover its attorneys’ fees, 7 then whether its requested fees are reasonable, and then Oracle’s request for costs. 8 A. Entitlement to Fees 9 The Court found in the Bench Order that Oracle prevailed on some of its copyright 10 claims, including its claims under the Digital Millenium Copyright Act (“DMCA”), and its 11 Lanham Act claims. The Court rejects Rimini’s contrary argument that this is a mixed 12 results case and thus no party is really the prevailing party. (ECF No. 1595 (sealed) at 13 15-17.) To start, the Court found that Oracle mostly prevailed in the Bench Order (ECF 14 No. 1536 at 1, 5 n.2, 193, 196) and then reiterated that finding in its order denying Rimini’s 15 emergency motion to stay the Injunction (ECF No. 1553 at 1, 5-6). The Court also stated 16 Oracle mostly prevailed in the Injunction. (ECF No. 1537 at 1 (“As explained in the Court’s 17 concurrently issued bench order, Oracle mostly prevailed both pretrial and at trial.”).) And 18 the Judgment also reflects that Oracle was the prevailing party. (ECF No. 1538 at 1.) But 19 perhaps most significantly, the Court decided to issue the Injunction,2 which reflects that 20 Oracle established Rimini’s liability as to some of its claims and some of Rimini’s support 21 processes. See Glacier Films (USA), Inc. v. Turchin, 896 F.3d 1033, 1038 (9th Cir. 2018) 22 (stating that actual success involves establishing liability even if the damages awarded 23 are nominal or nothing). The Injunction also reflects that Oracle obtained a material 24

25 associated with these motions were also filed under seal. United States Magistrate Judge Daniel J. Albregts granted the corresponding motions to seal. (ECF Nos. 1576, 1578, 26 1602, 1609.) The Court generally cites to the sealed versions of the documents in this order but indicates that they are sealed and does not disclose sensitive information. 27 2It remains stayed as of the date of entry of this order. (ECF No. 1553.) 1 alteration of the relationship between Oracle and Rimini—because it constrains what 2 Rimini can do, say about Oracle, and requires Rimini to issue a corrective press release. 3 See Cadkin v. Loose, 569 F.3d 1142, 1149 (9th Cir. 2009) (holding that “the material 4 alteration test the Supreme Court articulated in Buckhannon governs the prevailing party 5 inquiry under § 505 of the Copyright Act”); Fifty-Six Hope Rd. Music, Ltd. v. A.V.E.L.A., 6 Inc., 778 F.3d 1059, 1078 (9th Cir. 2015) (stating that the “material alteration” test also 7 applies to the determination of who is the prevailing party as to Lanham Act claims). In 8 contrast, Rimini obtained no injunction against Oracle. Oracle won the material alteration 9 in the parties’ relationship, not Rimini. This means that Oracle is the prevailing party even 10 though Oracle stipulated to the dismissal with prejudice of its damages claims shortly 11 before trial. 12 The Court further finds that Rimini’s approach of tallying up claims on which Oracle 13 prevailed (ECF No. 1595 (sealed) at 10-14, 16) does not reflect the reality that much of 14 the trial—and the Court’s corresponding findings in the Bench Order—was focused on 15 Oracle’s copyright infringement claims regarding the PeopleSoft product. (Compare ECF 16 No. 1536 at 10-62, 65-69, 134-57, 164-67 (findings as to PeopleSoft) with id. at 62-64, 17 73-76, 157-64 (findings as to all other products).) And Oracle largely prevailed on its 18 copyright infringement claims as to Rimini’s support processes for Oracle’s PeopleSoft 19 product. (See generally id.) Thus, even though Oracle did not prevail on all its copyright 20 claims, it substantially prevailed. Moreover, the Court’s Lanham Act findings are not 21 exclusively tied to any one Oracle product, so Rimini’s tallying approach does not quite 22 make sense when applied to them. (Id. at 76-95, 169-73.) In addition, Rimini did not 23 prevail on the only claim it went to trial on. (Id. at 95-115, 173-82.) In sum, Rimini’s tallying 24 approach to contest whether Oracle is the prevailing party paints an inaccurate picture of 25 the trial proceedings and the Court’s findings in the Bench Order. 26 And the pertinent caselaw Rimini cites does not compel a different conclusion. In 27 Kaloud, Inc. v. Shisha Land Wholesale, Inc., 741 F. App’x 393, 397 (9th Cir. 2018), the 1 United States Court of Appeals for the Ninth Circuit affirmed on abuse of discretion review 2 the district court’s decision not to award attorneys’ fees because the district court found 3 the case unexceptional, and indeed affirmed the district court’s denial of attorneys’ fees 4 to both sides. See 741 F. App’x at 396-97. The Court will analyze below whether this is 5 an exceptional Lanham Act case, and to preview, finds it is because the Court found 6 Rimini willfully made false statements. This case is accordingly not quite like Kaloud. 7 Similarly, the Ninth Circuit affirmed the district court’s decision not to award attorneys’ 8 fees in Guthy-Renker Corp. v. Bernstein, 39 F. App’x 584, 587-88 (9th Cir. 2002), again 9 on abuse of discretion review, because the district court found the moving party had 10 obtained only limited success on his pertinent claims, see id. The Court does not find 11 Oracle’s success so limited. And even that citation recognizes the Court’s discretion to 12 award fees in copyright cases. 13 The Court thus finds that Oracle is the prevailing party for purposes of its Motion. 14 The Court accordingly addresses Oracle’s entitlement to fees as to first Oracle’s copyright 15 claims and then Oracle’s Lanham Act claims. 16 1. Copyright Claims 17 The Court may award attorneys’ fees and costs to the party that prevails on 18 copyright claims. See 17 U.S.C. § 505. As to the DMCA in particular, “[t]he DMCA 19 authorizes a court, ‘in its discretion,’ to allow recovery of costs and to award ‘reasonable 20 attorney’s fees’ to the prevailing party.” Sony Computer Ent. Am., Inc. v. Divineo, Inc., 21 457 F. Supp. 2d 957, 967 (N.D. Cal. 2006) (quoting 17 U.S.C. § 1203(b)(4), (5)); see also 22 id. (finding award of attorneys’ fees and costs appropriate under the DMCA where the 23 DMCA infringement was willful). In determining whether to award fees and costs, the 24 Court must consider the factors described in Fogerty v. Fantasy, Inc., 510 U.S. 517, 534 25 n.19 (1994), which include: “frivolousness, motivation, objective unreasonableness (both 26 in the factual and in the legal components of the case), and the need in particular 27 circumstances to advance considerations of compensation and deterrence.” Perfect 10, 1 Inc. v. Giganews, Inc., 847 F.3d 657, 675 (9th Cir. 2017) (citation omitted). The Ninth 2 Circuit also instructs district courts to consider, “the degree of success obtained, the 3 purposes of the Copyright Act, and whether the chilling effect of attorney’s fees may be 4 too great or impose an inequitable burden on an impecunious plaintiff.” Id. (citation 5 omitted). But these factors are not exclusive, and not all of them must be met for the Court 6 to award fees. See id. 7 The parties proffer competing arguments on many of these nonexclusive factors, 8 so the Court addresses them in turn. Oracle first argues it achieved a high degree of 9 success on its Lanham Act claims because it got Rimini’s claim dismissed pretrial and 10 won the Injunction against Rimini, which provides in pertinent part that Rimini made false 11 statements and requires Rimini to issue a corrective press release about them. (ECF No. 12 1568 (sealed) at 18-19.) And Oracle argues it achieved a high degree of success on its 13 copyright claims because it proved many acts of infringement, the Court found Mr. Ravin 14 vicariously liable, and the Court’s Injunction requires both deletion of various infringing 15 files and requires Rimini to stop using its infringing automated tools. (Id.) Rimini counters 16 that Oracle only obtained a limited degree of success because Oracle dropped its 17 damages claims shortly before trial, and Oracle lost its copyright claims as to Oracle 18 products EBS, JDE, and Siebel—along with failing to obtain total victory as to PeopleSoft 19 and Database. (ECF No. 1595 (sealed) at 18.) The Court agrees with Oracle. 20 Indeed, the Court agrees with Oracle’s argument offered in its reply (ECF No. 1607 21 (sealed) at 7-8) that Rimini’s argument does not adequately address the Injunction. The 22 Injunction reflects the Court’s copyright findings in the Bench Order that Process 2.0 23 generally infringes, Rimini did not change its conduct following Judge Hicks’ findings in 24 Oracle I, Rimini violated the DMCA many times, and that Rimini must discontinue its use 25 of automated tools to perform PeopleSoft support. The Court also ordered Rimini in the 26 Injunction to delete many infringing files. The Injunction reflects a significant victory on 27 Oracle’s part even though Oracle dropped its damages claim because, if it goes into 1 effect,3 it will force Rimini to significantly change its PeopleSoft software support 2 business. On that note, even Rimini argued to the Court that compliance with the 3 Injunction would bring an end to an entire business line. (ECF No. 1542 at 9.) So while it 4 is true that Oracle dropped its damages claims, Oracle nonetheless achieved a significant 5 degree of success on its claims. 6 Oracle next argues that Rimini’s litigation position was objectively unreasonable 7 because the Court found some of Rimini’s statements about the non-infringing nature of 8 its support services were false—and that Rimini knew they were false when it made them. 9 (ECF No. 1568 (sealed) at 19-21.) Rimini counters that Oracle also took some 10 unreasonable positions, most notably not dropping its damages claim until right before 11 trial, Rimini prevailed on some of its arguments, and it made other arguments in good 12 faith even though the Court ultimately disagreed with them. (ECF No. 1595 at 18-20.) The 13 Court agrees with Rimini’s arguments for the most part, but finds they overlook the Court’s 14 findings about the willfulness of Rimini’s copyright infringement and Rimini’s related false 15 statements. 16 “[W]illful infringement is an important factor favoring an award of fees[,]” Hist. Rsch. 17 v. Cabral, 80 F.3d 377, 379 (9th Cir. 1996), and the Court’s findings regarding Rimini’s 18 willful copyright infringement highlight an objectively unreasonable position at the core of 19 this case. Indeed, it was Rimini who filed this case in an alleged attempt to establish that 20 its ‘Process 2.0’ did not infringe Oracle’s copyrights. (ECF No. 1.) But Rimini “knew when 21 it made them” that statements about its purportedly revised support processes were false. 22 (ECF No. 1536 at 77; see also id. at 12-18 (finding that Rimini did things it knew infringed 23 and lied about the timing and purpose of the migration), 76-79 (finding other statements 24 about Rimini’s revised support processes were false).) Rimini also decided not to comply 25 with the orders Judge Hicks issued in Oracle I. (Id. at 69-73.) In other words, this case 26 27 3Again, it is stayed pending a ruling from the Ninth Circuit. (ECF No. 1553.) 1 began with some false representations to the effect that Rimini had changed its practices 2 already found infringing in Oracle I that Rimini knew were false when it filed this case. So 3 what the Court views as the core position Rimini took in this case was objectively baseless 4 regardless of the fact that Rimini took some other, reasonable positions—some of which 5 the Court ultimately agreed with. 6 Oracle next argues that it needs the fee award it requests to be made whole (ECF 7 No. 1568 (sealed) at 21), and Rimini counters that Oracle had every opportunity to be 8 made whole by seeking a damages award, but forwent that opportunity when it dropped 9 its damages claims right before trial (ECF No. 1595 (sealed) at 20). The Court agrees 10 with Rimini on this argument. Damages are traditionally the vehicle for making parties 11 whole, and the Supreme Court overruled the precise portion of McCulloch v. Albert E. 12 Price, Inc., 823 F.2d 316, 323 (9th Cir. 1987) that Oracle cites to support this argument 13 in Fogerty, 510 U.S. at 114 in any event. This ‘make whole’ factor accordingly does not 14 favor awarding Oracle fees for the copyright claims it prevailed on. 15 Oracle then argues a fees award is necessary to deter Rimini from future 16 infringement, highlighting that Judge Hicks previously found this factor supported 17 awarding fees in Oracle I and that the deterrence rationale is only stronger now that the 18 Court has found Rimini a ‘recidivist infringer.’ (ECF No 1568 (sealed) at 21-22.) Rimini 19 counters in pertinent part that it is already deterred and has no interest in infringing 20 Oracle’s copyrights. (ECF No. 1595 (sealed) at 20.) But this very case illustrates that 21 Rimini’s response is not credible. 22 Deterrence is a valid reason to award fees. See Perfect 10, 847 F.3d at 675. And 23 deterrence is also a valid reason to award fees here. Again, Rimini filed this case based 24 on its false representations that it had changed its processes to avoid infringing Oracle’s 25 copyrights, see supra, but it had not stopped infringing and the Court found Rimini a 26 recidivist infringer (ECF No. 1536 at 190). In addition, Rimini litigated this case for some 27 nine years before the Court found it a recidivist infringer who lied to conceal its 1 infringement, enabling it to successfully compete against Oracle for the support of 2 Oracle’s products during those nine years. Rimini’s false statements and history of 3 recidivism suggest that more deterrence is required to prevent Rimini from infringing on 4 Oracle’s copyrights going forward. The deterrence factor strongly favors awarding Oracle 5 fees. 6 Oracle finally argues that awarding it fees would be in line with the purposes of the 7 Copyright Act because a fee award would reward Oracle for pursuing its rights under the 8 Copyright Act. (ECF No 1568 (sealed) at 22.) Rimini counters in pertinent part that this 9 factor does not favor awarding Oracle fees because some of Oracle’s positions—which 10 the Court rejected—were inconsistent with the purposes of the Copyright Act and would 11 have stifled competition if Rimini did not defend against them. (ECF No. 1595 (sealed) at 12 20-21.) Rimini further notes that meritorious defenses can also further the purposes of 13 the Copyright Act. (Id.) The Court finds both parties’ arguments as to this factor somewhat 14 persuasive and accordingly finds it neutral. 15 The Copyright Act’s “ultimate aim is ... to stimulate artistic creativity for the general 16 public good.” Mattel, Inc v. MGA Ent., Inc., 705 F.3d 1108, 1111 (9th Cir. 2013) (citation 17 omitted). But Oracle is not an artist. Pertinent to the Court’s copyright infringement 18 findings, Oracle is an enterprise software company who acquired PeopleSoft. (ECF No. 19 1536 at 10.) And Oracle’s argument as to this factor treats as equivalent the creator of a 20 work and someone who later acquired it. But no reasonable person would consider an art 21 collector similarly situated to the person who painted the painting. So the Court does not 22 find Oracle’s argument on this factor entirely persuasive. That said, the Court agrees with 23 Oracle that it found Rimini widely and willfully infringed Oracle’s copyrights despite what 24 should have been clear indications from Judge Hicks’ prior orders that the way Rimini ran 25 its PeopleSoft support business was infringing. Oracle’s assertion of its copyrights to 26 deter such willful infringement accordingly does seem somewhat consistent with the 27 purposes of the Copyright Act, as embedded within it is the presumption that artists 1 unable to deter infringement may be less inclined to create—because they would less 2 reasonably expect compensation for their original works. The Court also agrees with 3 Rimini that meritorious defenses can serve the purposes of the Copyright Act. But only 4 some of Rimini’s defenses were meritorious. As explained above, some of them were 5 objectively baseless. In sum, the ‘purpose of the Copyright Act’ factor is a mixed bag that 6 nets out to neutral. 7 Weighing all these factors together, the Court finds that Oracle is entitled to fees 8 for substantially prevailing on its copyright claims as to PeopleSoft and on its DMCA 9 claims. 10 2. Lanham Act Claims 11 The Court may also award fees to the prevailing party in exceptional Lanham Act 12 cases. See 15 U.S.C. § 1117(a). ‘“[A]n ‘exceptional’ case is simply one that stands out 13 from others with respect to the substantive strength of a party’s litigating position 14 (considering both the governing law and the facts of the case) or the unreasonable 15 manner in which the case was litigated.” SunEarth, Inc. v. Sun Earth Solar Power Co., 16 839 F.3d 1179, 1180 (9th Cir. 2016) (quoting Octane Fitness, LLC v. ICON Health & 17 Fitness, Inc., 572 U.S. 545, 554 (2014)). In deciding whether to award fees to the 18 prevailing party on a Lanham Act claim, district courts should consider the totality of the 19 circumstances, exercising their equitable discretion considering the nonexclusive factors 20 outlined in Octane Fitness and Fogerty, and using a preponderance of the evidence 21 standard. See id. But because the standard imposed by SunEarth “is less stringent than 22 the previous ‘willful infringement’ standard, it stands to reason that [a] case of willful 23 infringement would satisfy the SunEarth test.” Jason Scott Collection, Inc. v. Trendily 24 Furniture, LLC, 68 F.4th 1203, 1223 n.13 (9th Cir. 2023), cert. denied, 144 S. Ct. 550, 25 (2024). 26 This is an exceptional case for the reasons described above as to Oracle’s 27 copyright and DMCA claims, and because the Court found Rimini willfully made the false 1 statements included in the Injunction. See id. (stating willful infringement would satisfy the 2 SunEarth test). The Court included 15 false statements in the Injunction; three of which 3 relate to whether Rimini’s support processes infringe Oracle’s copyrights (the Court 4 includes the TomorrowNow statement in this category), and 12 which relate to the relative 5 security offerings of the two companies. (ECF No. 1537 at 4-7.) As to the statements 6 about Rimini’s support processes, the Court found they were false and misleading, Rimini 7 knew they were false and misleading, and Mr. Ravin was involved in making them. (ECF 8 No. 1536 at 72, 77, 95, 115.) Similarly, the Court found Rimini’s statements regarding 9 security false and misleading as well. (Id. at 169-73, 195-96.) The Court also found the 10 false and misleading nature of these statements irreparably harmed Oracle. (Id. at 188.) 11 In sum, the Lanham Act portion of this case was exceptional because of Rimini’s willful 12 violations of the Lanham Act. Oracle is accordingly entitled to its fees as to its Lanham 13 Act claims. See Jason Scott Collection, 68 F.4th at 1223 n.13. 14 B. Reasonableness of Fees 15 Now that the Court has determined Oracle is entitled to fees, the question 16 becomes, how much? Oracle argues the Court should award it full fees of $68,913,126.65 17 (ECF No. 1561-3) because its successful and unsuccessful claims are intertwined, Oracle 18 was charged reasonable hourly rates, and the amount billed on the case was reasonable. 19 (ECF No. 1568 (sealed) at 23-29.) 20 Rimini counters that Oracle’s requested fees are unreasonable and proposes 21 basically four different approaches to reduce the amount of fees the Court awards 22 supported by declarations from three experts:4 (1) only award fees for the claims Oracle 23 prevailed on, not the claims Oracle did not prevail on; (2) apply across-the-board 24 reductions of certain percentages; (3) deduct fees incurred in certain bills because the 25 bills are unreasonable or violate principles of honest billing; or (4) deduct fees incurred in 26

27 4The Court has not certified them as experts, but the Court nonetheless refers to them as experts in this order. 1 preparation for the jury trial that Oracle decided to forgo. (ECF No. 1595 (sealed) at 22- 2 30; see also ECF Nos. 1597, 1598, 1599 (sealed expert declarations as to fees).) 3 The Court agrees with Rimini that it should deduct categories of fees closely 4 associated with preparation for the jury trial that Oracle decided to forgo and deduct an 5 additional approximately $30,000 for billing entries documenting administrative tasks 6 billed hourly but will otherwise award Oracle the rest of the fees it requests. “The district 7 court may attempt to identify specific hours that should be eliminated, or it may simply 8 reduce the award to account for the limited success.” Hensley v. Eckerhart, 461 U.S. 424, 9 436-37 (1983). The Court exercises its discretion and finds the approach of identifying 10 specific hours that should be eliminated the better one here. See id. (giving courts this 11 discretion). The Court further describes below why it finds some of Rimini’s proposed 12 approaches for fee reductions unpersuasive, and others based on the elimination of 13 specific hours for jury trial preparation and administrative work persuasive. 14 To start, Rimini argues that some fees should be deducted because they were 15 incurred while Oracle was attempting to establish causation and damages, and Oracle 16 dropped its damages claims shortly before trial. (ECF No. 1595 at 24-25.) But Rimini 17 overlooks that causation was still an important issue in the bench trial, and indeed the 18 Court found in the Bench Order that Rimini’s actions caused Oracle harm in deciding to 19 issue the Injunction. (ECF No. 1536 at 183-91 (finding that Rimini caused Oracle harm in 20 finding that the irreparable harm and inadequacy of monetary damages prongs of the 21 permanent injunction analysis favored issuing the Injunction).) Said otherwise, the Bench 22 order contained significant causation findings. (Id.) So a reduction in fees for fees incurred 23 establishing causation and damages without separating one from the other would be 24 unreasonable because causation was still an important issue during the bench trial. 25 Rimini next argues that the Court should apportion fees and decline to award 26 Oracle fees for claims on which it did not prevail, which Rimini contends were not 27 intertwined with the claims Oracle prevailed on. (ECF No. 1595 (sealed) at 25-27.) This 1 approach is valid in the abstract, but the Court does not find it the best approach here— 2 for two reasons. First, for the reasons Oracle provides in its Motion (ECF No. 1568 3 (sealed) at 23-26), Oracle’s successful and unsuccessful claims were intertwined. And 4 allocation is not required where there is a common core of facts between successful and 5 unsuccessful claims. See Oracle USA, Inc. v. Rimini St., Inc., 783 F. App’x 707, 711 (9th 6 Cir. 2019) (affirming Judge Hicks’ decision not to apportion fees in Oracle I). Second, and 7 perhaps more importantly, Rimini’s request to apportion fees is based on Rimini’s expert’s 8 John L. Trunko’s analysis where he “identified at most only 9,172.39 hours totaling 9 $3,902,190.68 in fees associated with specific claims and issues on which Oracle 10 prevailed, such as PeopleSoft, the DMCA, and the Lanham Act.” (ECF No. 1598 (sealed) 11 at 13.) That estimate is unrealistically low considering that Oracle seeks around $70 12 million in fees and the Court has determined that Oracle mostly prevailed. A reasonable 13 estimate would thus put fees around at least more than half of that amount. Moreover, 14 Mr. Trunko supports his estimate with a citation to his Exhibit 6, which does not show the 15 actual billing entries he has determined are related to the claims he understands Oracle 16 prevailed on, and instead shows a series of totals of hours and fees by timekeeper. (ECF 17 No. 1598-6 (sealed).) This does not give the Court enough information to assess the 18 validity of Mr. Trunko’s analysis. The Court accordingly declines to apportion fees based 19 the claims Oracle prevailed on. 20 However, the Court agrees with Rimini that it would be unfair and unreasonable to 21 award Oracle fees incurred in preparation for a jury trial that Oracle chose to forego. (ECF 22 No. 1595 (sealed) at 25-26.) Indeed, it was Oracle who initiated the stipulation to convert 23 the trial from a jury to a bench trial. (ECF Nos. 1409, 1411, 1416, 1417.) It would thus be 24 unreasonable to award fees to Oracle for work that it decided to render unnecessary, 25 particularly after years of litigation where Rimini understood it was defending against 26 some $1.4 billion in alleged damages. (ECF No. 1595 (sealed) at 9-10, 24-25.) Rimini’s 27 expert Mr. Trunko specifically identified three categories of bills that the Court finds 1 reasonably relate to the jury trial: (1) the mock jury exercise ($1,042,558.10); (2) third 2 party subpoenas because those were intended to prove damages and causation on a 3 customer-by-customer basis ($6,589,460.61); and (3) jury trial tasks ($3,058,185.98). 4 (ECF No. 1598 (sealed) at 17-18.) These deductions total $10,690,204.69. As to the 5 subpoenas specifically, the Court agrees with Rimini that, had Oracle never sought 6 damages, “[t]he case could have been litigated based on Rimini’s processes and a 7 handful of clients without the need for 800 client subpoenas.” (ECF No. 1595 (sealed) at 8 25.) Thus, the deduction for time spent on third-party subpoenas is reasonable. And 9 overall, this $10,690,204.69 deduction for tasks related to a jury trial that never happened 10 is reasonable. 11 Rimini also argues for across-the-board percentage reductions in Oracle’s fees 12 based on declarations from their experts Dennis L. Kennedy and William G. Ross (ECF 13 No. 1598 (sealed) at 27-30), but the Court finds the deductions adopted above primarily 14 to account for Oracle’s decision to dismiss its damages claims shortly before trial and 15 forego a jury trial provide a more specific and reasoned basis for deductions than an 16 across-the-board percentage would. Moreover, it is possible that the amount deducted 17 for third party subpoenas is overinclusive because, as the Court noted above, causation 18 remained relevant to the bench trial, even though damages were no longer. It is also 19 possible that all these deductions are somewhat overstated because Rimini’s expert is 20 the one providing the numbers. See Schwarz v. Sec’y of Health & Hum. Servs., 73 F.3d 21 895, 905 (9th Cir. 1995) (“[A] district court does not abuse its discretion when it resorts to 22 a mathematical formula, even a crude one, to reduce the fee award to account for limited 23 success.”). And while across-the-board percentage reductions have been found 24 acceptable, see id., reductions grounded in categories of unreasonable billing 25 nonetheless seems more precise. See Fox v. Vice, 563 U.S. 826, 838 (2011) (“The 26 essential goal in shifting fees (to either party) is to do rough justice, not to achieve auditing 27 perfection.”). In sum, the Court rejects Rimini’s proposal to apply across-the-board 1 percentage reductions to Oracle’s fees and further notes that the reductions for Mr. 2 Trunko’s third party subpoenas category should account for some or all the purported 3 overbilling Rimini complains about. 4 Rimini’s expert Mr. Trunko also proposes additional categories of unreasonable 5 bills that he argues warrant further reductions to Oracle’s fees award. (ECF No. 1598 6 (sealed) at 10-18.) The Court addresses the remaining categories in the order Mr. Trunko 7 presents them. Mr. Trunko first argues for a reduction for block billing, but the exhibit he 8 cites does not provide examples of the bills he contends are block billed, much less the 9 actual entries. (Id. at 10-11; see also ECF No. 1598-3 (sealed) (containing a list of 10 timekeepers and totals of their hours and fees).) The Court also finds Oracle’s 11 counterargument persuasive that a small amount of block billing is acceptable, and even 12 taking Mr. Trunko at his word, the alleged amount of block billing is 4.7% of the total hours 13 Oracle seeks its fees for. (ECF No. 1607 (sealed) at 14.) The Court therefore rejects an 14 additional reduction based on block-billing. 15 Mr. Trunko next argues for reductions based on ‘vague’ time entries, but the Court 16 does not find the examples he provides too vague. (ECF No. 1598 (sealed) at 11-13.) Mr. 17 Trunko further explains that many of the time entries he finds objectionable under this 18 category ‘contain’ words that indicate vagueness but does not list out the rest of the bill 19 entries or point the Court to an exhibit that shows all the entries he objects to. So, the 20 Court cannot really say whether these entries are vague, and it is unreasonable to infer 21 that an entire bill entry is vague because it contains a single word suggesting vagueness 22 without providing the rest of the entry. In addition, the entries that seem the vaguest in 23 Mr. Trunko’s report, such as ‘attention to customer subpoenas’ and ‘work on third party 24 productions’ should already be captured in the deduction for hours spent on the third party 25 subpoenas that the Court has already decided to apply to Oracle’s fees award. The Court 26 accordingly does not find a further deduction for vague entries warranted, understanding 27 1 that some of these entries are presumably included in the deduction category for third 2 party subpoenas. 3 Mr. Trunko then argues a further deduction for long billing days is warranted, but 4 many of the examples he points to appear to be either right before or during trial, and long 5 billable days are unfortunately to be expected preparing for and during trial, particularly a 6 trial as complex as this one. (ECF No. 1598 (sealed) at 13-15.) Moreover, the exhibits (11 7 and 12) that Mr. Trunko uses to support this portion of his declaration do not give the 8 Court any insight as to why various timekeepers may have billed long days on the 9 occasions Mr. Trunko identified, so the Court cannot find unreasonable the fees billed for 10 work performed on those long days. (Id.; see also ECF Nos. 1598-11 (sealed), 1598-12 11 (sealed).) The Court declines to further reduce Oracle’s fees for purportedly long days. 12 Mr. Trunko next argues a further deduction for excessive conferencing and team 13 meetings between attorneys is warranted. (ECF No. 1598 (sealed) at 15.) But the Court 14 would both hope and expect that attorneys are meeting in a matter as complicated as this 15 one, involving this many attorneys, as hours spent meeting could eliminate many hours 16 of duplicative or unnecessary work. The Court is indeed unsure how a case like this one 17 could efficiently progress without team meetings. Regardless, the Court rejects Rimini’s 18 invitation to further reduce Oracle’s fees for time spent on meetings between attorneys. 19 Mr. Trunko then argues for a further reduction for time spent performing 20 administrative tasks. (Id. at 15-16.) To support this argument, Mr. Trunko cites two 21 exhibits: Exhibit 10 and Exhibit 14. (Id.) The Court agrees with Mr. Trunko that the 22 examples provided in Exhibit 10 are clerical tasks billed at hourly rates, and accordingly 23 agrees a reduction is warranted. See Gilead Scis., Inc. v. Merck & Co., No. 13-CV-04057- 24 BLF, 2017 WL 3007071, at *8 (N.D. Cal. July 14, 2017) (citing Nadarajah v. Holder, 569 25 F.3d 906, 921 (9th Cir. 2009), and agreeing that a fees reduction is warranted for clerical 26 tasks such as maintaining files billed at hourly rates). The fees for the examples detailed 27 in Mr. Trunko’s Exhibit 10 total $30,700.85. (ECF No. 1598-10 (sealed) at 13-15.) The 1 Court will accordingly apply an additional reduction of $30,700.85 to Oracle’s fees award. 2 However, Mr. Trunko’s Exhibit 14 suffers from the same deficiency the Court has 3 identified above as to several of his other exhibits—it lists timekeepers, totaling some of 4 their hours and their fees billed, but does not show the Court the billing entries Rimini is 5 objecting to. (ECF No. 1598-14 (sealed).) Mr. Trunko further explains in his declaration 6 that the entries reflected in Exhibit 14 contain words like ‘upload,’ ‘download,’ ‘finalize,’ 7 and ‘forward.’ (ECF No. 1598 (sealed) at 16.) But in addition to not showing the Court 8 what the rest of the entries say, the Court notes that ‘finalizing’ a document such as a 9 motion for summary judgment can be a substantive task that Oracle should be able to 10 recover its fees on. The Court accordingly declines to further reduce Oracle’s fees to 11 reflect the entries covered by Mr. Trunko’s Exhibit 14. 12 Mr. Trunko next argues for a reduction in Oracle’s fees for multiple entries on the 13 same day with the same description for each. (Id. at 16-17.) But this proposed reduction 14 is illogical. It may take several days or even weeks to draft a motion. An attorney may 15 work on it all morning, take a break for lunch, and then work on it again in the afternoon. 16 So multiple entries on the same day with the same description does not necessarily 17 indicate duplication—it may just indicate a longer project. The Court declines Rimini’s 18 invitation to further reduce Oracle’s fees based on the ‘identical entries’ category. 19 Mr. Trunko finally argues for a reduction in fees for excessive staffing, particularly 20 calling out time spent by attorneys added to the case after Oracle’s lead attorneys 21 switched firms, suggesting that the newly added attorneys were billing time for 22 familiarizing themselves with things that already happened in the case. (Id. at 17.) This is 23 a fair critique, but the Court declines to further reduce Oracle’s fees based on this 24 category because of the unique circumstances of this case. First, this case was filed in 25 2014 and went to trial in late 2022. It is reasonable to expect that some attorneys would 26 change over time given the length of this litigation. Second, the Court agrees with Oracle 27 (ECF No. 1607 (sealed) at 16) that Rimini’s allegations of overstaffing are hypocritical 1 given that Rimini too employed several law firms and many lawyers over the life of this 2 case. The Court declines to further reduce Oracle’s fees based on Mr. Trunko’s 3 ‘excessive staffing’ category. 4 Zooming back out, Rimini does not challenge the reasonableness of the hourly 5 rates Oracle was charged, and the Court finds the hourly rates reasonable. (ECF No. 6 1595 (sealed).) Indeed, the fact that Rimini does not challenge the reasonableness of the 7 hourly rates Oracle’s attorneys charged contributes to the Court’s finding that the hourly 8 rates Oracle’s attorneys charged were reasonable. But other factors contribute to this 9 finding as well. First, Judge Hicks found that similarly high hourly rates were reasonable 10 in Oracle I and awarded Oracle some $28 million in fees. See Oracle USA, Inc. v. Rimini 11 St., Inc., 324 F. Supp. 3d 1157, 1171 (D. Nev. 2018). Second, the Ninth Circuit affirmed 12 Judge Hicks’ attorneys’ fees ruling and did not specifically address the hourly rates, but 13 did not find them unreasonably high, either. See Oracle, 783 F. App’x 707, 711. Third, 14 Oracle’s in-house counsel James Maroulis explains that a big part of the reason he hired 15 these attorneys for this case is that they also represented Oracle in Oracle I. (ECF No. 16 1574 (sealed) at 2-3.) So Oracle’s attorneys in this case are some of the same attorneys 17 from Oracle I, and both Judge Hicks and the Ninth Circuit found their hourly rates 18 reasonable in that very similar case. Finally, Mr. Maroulis also explains that Oracle is a 19 sophisticated consumer of legal services, it will only pay hourly rates it finds reasonable, 20 Oracle received discounted rates from its attorneys in this case, and he reviewed the bills 21 and found them reasonable. (See generally id.) All these factors support finding that the 22 hourly rates charged by Oracle’s counsel were reasonable. 23 In sum, the Court finds that Oracle is entitled to $10,720,905.50 24 ($10,690,204.69+$30,700.85) less in fees than the $68,913,126.65 it requests in its 25 Motion (ECF No. 1561-3 (listing this as the full amount of fees sought)). That means 26 Oracle is entitled to $58,192,221.15 in attorneys’ fees. 27 /// 1 C. Costs 2 This brings the Court to Oracle’s request for its full taxable costs. (ECF No. 1568 3 (sealed) at 29-30.) Rimini only objects to Oracle’s request for $1,368,960 in certain e- 4 discovery costs. (ECF No. 1595 (sealed) at 30.) Rimini does not object to Oracle’s other 5 requested costs, which total $320,277.55 ($185,303.81 + $130,601.22 + $4,372.52) for 6 deposition costs, costs related to service of process, and costs related to paper copies 7 and trial exhibits for court. (ECF No. 1568 (sealed) at 29-30.) And the Court agrees with 8 Oracle that award of these costs is proper for the reasons Oracle provides. (Id.) The Court 9 accordingly finds Oracle is entitled to $320,277.55 in unobjected-to taxable costs. 10 Turning to Rimini’s objection to Oracle’s request for $1,368,960 in certain e- 11 discovery costs, Rimini argues Oracle’s request is frivolous because the Supreme Court 12 held e-discovery costs are unrecoverable in an appeal from Oracle I. (ECF No. 1595 13 (sealed) at 30 (citing Rimini Street, Inc. v. Oracle USA, Inc., 139 S. Ct. 873, 878 (2019)).) 14 Oracle counters in reply that these e-discovery costs are taxable costs, and the Supreme 15 Court ruling only applies to non-taxable costs—as Rimini acknowledged later in Oracle I. 16 (ECF No. 1607 (sealed) at 17.) Both parties support their arguments with declarations. 17 (ECF No. 1568 (sealed) (citing in pertinent part ECF No. 1573 (sealed) at 2-3); ECF No. 18 1595 (sealed) at 30 n.5 (citing ECF No. 1600 (sealed) at 7).) And while resolution of this 19 issue requires slightly more nuance than Rimini presents in its brief, the Court agrees with 20 Rimini. 21 It is true that the Supreme Court, in the opinion arising from Oracle I, held that full 22 costs under the Copyright Act means the costs specified in the general costs statutes, not 23 other types of costs. See Oracle I, 586 U.S. at 346. In addition, in that opinion, Justice 24 Kavanaugh (writing for the Court) wrote that the general costs statutes, “in turn do not 25 authorize an award for expenses such as expert witness fees, e-discovery expenses, and 26 jury consultant fees, which were expenses encompassed by the District Court’s $ 12.8 27 million award to Oracle here.” Id. at 341. The disputed costs that Oracle seeks to recover 1 here are ‘e-discovery expenses’ (ECF Nos. 1573 (sealed) at 2-3, 1573-2 (sealed)), so 2 Rimini is arguably correct that Oracle’s request for these costs is simply barred by the 3 Supreme Court’s opinion. 4 However, Oracle points to a stipulation with Rimini in Oracle I where Rimini agreed 5 that the Supreme Court’s opinion only related to non-taxable costs, and thus did not 6 disturb the award of taxable costs to Oracle. (ECF No. 1607 (sealed) at 17 (citing Oracle 7 I, No. 2:10-cv-00106-LRH-MDC, ECF No. 1233 at 1:17-19).) Oracle then argues that the 8 $1,368,960 in certain e-discovery costs it seeks are taxable costs permitted under 28 9 U.S.C. § 1920(4) (“Fees for exemplification and the costs of making copies of any 10 materials where the copies are necessarily obtained for use in the case”) and thus not 11 necessarily foreclosed by the Supreme Court’s opinion. (ECF Nos. 1568 (sealed) at 30, 12 1607 (sealed) at 17).) But Rimini has a response to this argument, too, in the form of a 13 declaration from its expert, Peter S. Vogel. (ECF No. 1595 (sealed) at 30 n.5 (citing ECF 14 No. 1600 (sealed) at 7.).) And the Court finds Mr. Vogel’s critique of the declaration Oracle 15 proffers from Oracle’s employee Laura Fichet persuasive. 16 To simplify the disagreement between Ms. Fichet and Mr. Vogel, Ms. Fichet says 17 that Oracle is entitled to some $1.3 in taxable costs for making copies necessary to 18 respond to Rimini’s discovery requests, but Mr. Vogel counters that the supporting 19 document Ms. Fichet attaches to her declaration does not show that Oracle paid for 20 copies. Mr. Vogel is right. Mr. Vogel first points out some discrepancies because Ms. 21 Fichet says her Exhibit 2 is documenting ProSearch charges, but the document itself does 22 not refer to ProSearch at all, instead apparently documenting fees paid to the firm Ryley 23 Carlock in an Oracle system. (ECF No. 1600 (sealed) at 5; ECF No. 1573-2 (sealed) at 24 2-3.) And setting that aside, Mr. Vogel explains that the document shows costs incurred 25 for processing and production of information, which, in his experience, is not the same as 26 ‘copying.’ (ECF No. 1600 (sealed) at 6-7.) The Court agrees with Mr. Vogel’s read of Ms. 27 Fichet’s Exhibit 2. For example, the $971,295 for ‘total production cost’ includes charges 1 || for 2,260 hours of project management, which is not copying. (ECF No. 1573-2 (sealed) 2 || at 2.) And even more generally, the charges shown for processing and production do not 3 || say copying, and both logic and Mr. Vogel’s experience suggest that they reflect charges 4 || for broader activities than simply copying. (/d. at 2-3; see also ECF No. 1600 (sealed) at 5 || 6-7.) 6 Thus, to the extent the Supreme Court opinion in Oracle | does not simply bar 7 || Oracle’s request for $1,368,960 in certain e-discovery costs, those costs are not 8 || recoverable because Oracle has not shown that those costs reflect “[flees for 9 || exemplification and the costs of making copies of any materials where the copies are 10 || necessarily obtained for use in the case[.]” 28 U.S.C. § 1920(4). See also In re Online 11 || DVD-Rental Antitrust Litig., 779 F.3d 914, 926 (9th Cir. 2015) (stating that, “[a]s a general 12 || rule, costs and fees should be awarded to the prevailing party[,]’ but “a district court’s 13 || discretion to award costs is limited to particular types of costs enumerated in 28 U.S.C. § 14 || 1920”) (citations omitted). 15 In sum, Oracle is only entitled to $320,277.55 in taxable costs. Adding that to the 16 || attorneys’ fees total mentioned above sums to $58,512,498.70 ($58,192,221.15 + 17 || $320,277.55). 18 || Ill ©CONCLUSION 19 It is therefore ordered that Oracle’s motion for attorneys’ fees and costs (ECF Nos. 20 || 1560, 1568) is granted in part, and denied in part, as specified herein. 21 It is further ordered that Oracle is entitled to $58,512,498.70 in attorneys’ fees and 22 || costs. 23 The Clerk of Court is directed to enter a separate judgment in Oracle’s favor 24 || reflecting this award of attorneys’ fees and costs. 25 DATED THIS 23% Day of September me 1 CL) 26 27 CHIEF UN TED STATES DISTRICT JUDGE 28 20