KELLY, Circuit Judge.

Michael L. Baum, Pedram Esfandiary, R. Brent Wisner, Baum & Hedlund, Los Angeles, CA, Bryce B. Bell, Mark W. Schmitz, Bell Law, LLC, Kansas City, MO, Clark Andrew Binkley, Kim E. Richman, Petra Renee Wicklund, Richman Law & Policy, Irvington, NY, for Plaintiffs-Appellees.

Amy M. Crouch, Attorney, Douglas B. Maddock, Jr., James Patrick Muehlberger, Shook & Hardy, Kansas City, MO, Patrick E. Hogan, John J. Rosenthal, Winston & Strawn, Washington, DC, Jeffrey Scott Wilkerson, Winston & Strawn, Charlotte, NC, for Defendant-Appellee.

Theodore H. Frank, Adam E. Schulman, Hamilton Lincoln Law Institute, Washington D.C., Jonathan R. Whitehead, Whitehead Law Office, Lee's Summit, MO, for Objector-Appellant.

David M.S. Dewhirst, Kathleen L. Smithgall, Attorney General's Office, Helena, MT, for Amici on Behalf of Appellant(s).

Before SMITH, Chief Judge, BENTON and KELLY, Circuit Judges.

KELLY, Circuit Judge.

Anna St. John objected to a class action settlement between Defendant Monsanto and Plaintiffs Lisa Jones, Horacio Torres Bonilla, and Kristoffer Yee, on behalf of a class of consumers. The district court overruled St. John's objections, approved the settlement, and awarded Plaintiffs attorney's fees. St. John appeals, and we affirm.

I. Background

Plaintiffs filed suit in February 2019, pleading multiple claims arising out of the allegedly deceptive labelling of Roundup products manufactured by Monsanto. Specifically, Roundup products bore a label indicating that the active ingredient, glyphosate, "targets an enzyme found in plants but not in people or pets." Plaintiffs alleged, however, that Monsanto knew that glyphosate is in fact present in gut bacteria in both humans and animals, so the label was false.

In August, the parties attended a formal mediation. Throughout the fall and winter, they continued to exchange discovery and negotiate the details of a settlement. As part of this process, both parties commissioned experts to quantify the measure of damages. The experts surveyed consumers to determine how much less they might expect to pay for the Roundup product without the misleading label. Plaintiffs’ expert estimated that the misleading label constituted 7.9% to 15.9% value. Plaintiffs concluded, therefore, that 15.9% of the value of the products purchased was the best-case damages after victory at trial. Monsanto's expert found no significant difference in the value of a product with and without the challenged label and estimated no more than 2.5% of the value as damages.

An initial proposed settlement agreement was presented to the district court for preliminary approval in March 2020. The parties agreed to a total Common Fund of $39.55 million. They agreed that Monsanto would not object to Plaintiffs’ counsel seeking 25% of that amount as an attorney's fee. Class members who filed claims were to receive 10% of the average retail price for the product(s) they bought, and any remaining funds after the costs of administration would be distributed cy pres .

Before the district court ruled on that motion, the parties executed a Second Corrected Class Action Settlement Agreement that made four changes to the initial agreement: (1) narrowed the scope of the class members’ release of claims; (2) added Plaintiffs’ intent, unopposed by Monsanto, to seek an incentive payment of $2,500 for each named plaintiff; (3) proposed two cy pres recipients—the National Consumer Law Center and the National Advertising Division of the Better Business Bureau—and clarified the cy pres selection process; and (4) extended the notice period and opt-out deadline. The notice documents were updated to reflect these changes, though they did not identify the cy pres organizations specifically. The district court granted preliminary approval, certified a national settlement class, and approved notice to putative class members.

The 90-day notice period began on May 28 and ended on August 28, 2020. The initial forms of notice included: publication in an issue of Better Homes & Gardens; banner notices on Google, Yahoo!, Facebook, Instagram, and YouTube targeting individuals with an interest in lawn and garden maintenance; radio and banner notices on Pandora streaming radio targeted to lawn and garden enthusiasts; sponsored search advertising on Google Ads for key words related to the litigation; nationwide news release; and creation of a settlement website and hotline. In July, midway through the notice period, the parties directed the claims administrator to initiate a supplemental notice program to augment the notice obtained by the methods described above. This supplemental notice included: more targeted digital banners; email distribution to a purchased, curated list of individuals; advertisements in four digital newsletters on relevant topics; and notices on two class action aggregation websites. The claims administrator calculated that these combined notice efforts reached 82% of class members with an average frequency of 2.51 contacts.

In October 2020, the parties sought approval from the district court for another updated settlement and notice. First, the parties proposed amending the settlement to allow for a possible upward adjustment of payments to claimants of up to 50% of product value rather than the 10% figure previously agreed to. They also added a third proposed cy pres recipient, the Berkeley Center for Consumer Law & Economic Justice. The parties proposed an additional notice period of 90 days for the updated notice, which would include the original forms of notice and the supplemental forms of notice initiated in July, plus new television and radio advertising. The revised notice would inform class members of the possible pro rata increase in payments to claimants. The district court approved this proposal.

The supplemental notice and claim period ended on February 16, 2021. The following week, the claims administrator reported that it had received 285,399 total claims accounting for slightly more than 1 million products, though it anticipated rejecting approximately 43,000 of those as duplicative or deficient. This represented a 2–3% estimated claims rate based on total sales of almost 89 million units during the relevant period. The validity of some claims had not been verified at the time of briefing, but the parties indicate that the value of the valid claims will range between $11.72 million and $13.34 million. The 25% award to the attorneys is $9.89 million, and the administrator's fees amounted to $1.8 million. This leaves approximately $14 to $16 million to be distributed cy pres , depending on the final value of the valid claims.

St. John made three objections to the settlement, all of which she renews on appeal. First, St. John argues that there are further steps the parties could take to identify and encourage the participation of more class members. At the very least, St. John argues, the payment to class members who have made claims should be increased to 100% of the price of the products purchased before donating proceeds cy pres . Second, St. John argues that the district court's order allowing funds to be donated to the cy pres organizations constitutes compelled speech in violation of her First Amendment rights. Finally, St. John argues that the cy pres should be excluded from the total value of the Common Fund for purposes of calculating the attorney's fee and that time spent on related litigation in another district court should be excluded from the compensable time considered in the lodestar analysis.

II. Legal Standard

"We review a district court's order approving a class action settlement for an abuse of discretion." Rawa v. Monsanto Co., 934 F.3d 862, 868 (8th Cir. 2019). "In doing so, ‘we ask whether the district court considered all relevant factors, whether it was significantly influenced by an irrelevant factor, and whether in weighing the factors it committed a clear error of judgment.’ " Id. (quoting Marshall v. Nat'l Football League, 787 F.3d 502, 508 (8th Cir. 2015) ).

III. Discussion

A. Size of the Cy Pres

St. John's first objection is to the size of the cy pres distribution. St. John argues that the district court should have (1) required the parties to take additional steps to identify additional class members and (2) increased the pro rata portion of the Common Fund up to 100% of the weighted average retail price.

The district court did not abuse its discretion in concluding that notice to the class was sufficient in light of the comprehensive notice plan and the estimated results from the claims administrator. This court has noted that "a claim rate as low as 3 percent is hardly unusual in consumer class actions and does not suggest unfairness." Keil v. Lopez, 862 F.3d 685, 697 (8th Cir. 2017) (affirming the district court's conclusion that a settlement was fair, reasonable, and adequate where the potential class covered 3.5 million households, an estimated 87% of those received notice, and 105,173 claims were submitted against a settlement fund of $32 million). St. John points to cases in which the parties subpoenaed consumer data from retailers and were able to make direct payments to consumers based on those records, including to consumers who did not opt in to the class. The district court engaged the parties about that possibility during a hearing, asking, "So in light of the objector's objection, have you done any additional investigation as to whether – whether additional notice is possible, the cost of additional notice, the reference that the objector makes to subpoenaing records from big-box retail locations, actions or steps of that sort?" Plaintiffs’ counsel responded that after conferring with the claims administrator, they concluded that the notice plan already in place

was actually more effective than seeking subpoenas from retailers who have increasingly imperfect data. Especially with ongoing privacy concerns, retailers – major retailers are now getting rid of a lot of that data, they're not holding on to it in the way that they used to. They, of course, aren't tracking people who make purchases with cash and, of course, it would not include people who purchased from smaller retail outlets. So it was our conclusion that that would not have been a most effective form of updating notice and that the steps that we already took were, in fact, more effective.

There is no further discussion in the record of the feasibility of St. John's proposed approach. We do not doubt that there are circumstances in which pursuing records from retailers is a reasonable and effective way to get relief to class members, especially because it might allow for direct payments to affected consumers without a cumbersome claims process. Based on this record, however, the district court did not abuse its discretion by not requiring the parties to pursue this approach in addition to the notice plan that had already been implemented, which advertised the settlement in a targeted way across numerous platforms and was revised twice in an effort to reach more consumers.

The second issue St. John raises is whether the class members who have been identified are entitled to a larger proportion of the price of the product, up to 100%, before the residual funds are allocated cy pres . Relying on In re BankAmerica Corp. Securities Litigation, 775 F.3d 1060 (8th Cir. 2015), St. John argues that because class members’ damages are unliquidated, they should be able to recover up to the full purchase price before the district court may order cy pres distribution. Concerns about a windfall to class members, St. John asserts, are not relevant in the context of unliquidated damages.

This argument overstates BankAmerica’s holding. In BankAmerica, we held that unclaimed funds may only be distributed cy pres where existing class-member claimants have been fully compensated and further distribution to remaining class members is not feasible. Id. at 1064. Where class members have claims for liquidated damages, they are fully compensated when those claims are "100 percent satisfied by the initial distribution." Id. (quoting Klier v. Elf Atochem N. Am., Inc., 658 F.3d 468, 475 (5th Cir. 2011) ). When damages are unliquidated, class members are not necessarily " ‘fully compensated’ by payment of the amounts allocated to their claims in the settlement." Id. at 1065. If the settlement provides "only a percentage of the damages" sought by the plaintiffs and "the settling parties disagree as to both liability and damages, and do not agree on the average amount of damages per share that would be recoverable by any of the Classes," then "the notion that class members were fully compensated by the settlement is speculative, at best." Id. at 1066. Contrary to St. John's assertion, however, this does not require that class-member claimants receive the full amount of unliquidated damages claimed in the complaint before cy pres distribution. Rather, it requires the district court to make its own assessment of the damages "that would be recoverable" by class members before approving distribution of the residual funds cy pres . The reversible error in BankAmerica was not that plaintiffs had not received the full change in stock value but that the district court had not determined the measure of class members’ damages and whether they had been fully compensated before granting a cy pres distribution.

In this case, the district court conducted such an analysis, and we find no abuse of discretion in its conclusion that a payment to class members of 50% of the average weighted retail price for the items they purchased "fully compensated" the class members and that they had no equitable claim to the remaining funds, which were appropriately distributed cy pres . The district court reasoned that even if class members claimed they would not have purchased Roundup if it had not borne the allegedly misleading label, their damages would still have to be reduced from 100% to account for the value they received from using Roundup. The conclusions of both parties’ experts also support this finding—Monsanto's expert's survey found a 2.5% differential and plaintiffs’ expert's survey found a differential of 7.9% to 15.9%. We see no clear error of judgment in the district court's conclusion.

B. First Amendment

St. John's next argument is that the district court ordering a cy pres distribution to particular charitable organizations is a form of compelled speech of the class members in violation of the First Amendment. We disagree. The First Amendment prevents the government from "compel[ling] the endorsement of ideas that it approves." Knox v. Serv. Emps. Int'l Union, Local 1000, 567 U.S. 298, 309, 132 S.Ct. 2277, 183 L.Ed.2d 281 (2012). "Compelling a person to subsidize the speech of other private speakers raises similar First Amendment concerns." Janus v. AFSCME, Council 31, ––– U.S. ––––, 138 S. Ct. 2448, 2464, 201 L.Ed.2d 924 (2018). But class members have not been compelled to subsidize speech when residual funds are distributed cy pres . As discussed above, residual funds may only be distributed cy pres after class members who have filed claims are "fully compensated" and no further allocation of funds to other, remaining class members is feasible or appropriate. See BankAmerica, 775 F.3d at 1064–66. So while settlement funds "are the property of the class," id. at 1064 (quoting Klier, 658 F.3d at 475 ), residual funds do not belong to any individual class member who has received his or her portion of the settlement fund. Nor are the class members who fail to claim their portion of the settlement fund compelled to subsidize speech; they could have filed a claim to collect the funds themselves or opted out of the settlement and preserved their right to pursue their claims individually, but they have no claim to residual funds. And neither of these situations is analogous to the facts considered by the Supreme Court in Janus. The compelled speech in Janus involved automatic deductions taken from employees’ paychecks. A cy pres distribution, in contrast, "involves funds that, regardless of the cy pres provisions, could not feasibly be paid to class members," In re Google Inc. St. View Elec. Commc'ns Litig., 21 F.4th 1102, 1118–19 (9th Cir. 2021), and so cannot be money "taken" from any member of the class, cf. Janus, 138 S. Ct. at 2486 (First Amendment requires that "employees clearly and affirmatively consent before any money is taken from them" (emphasis added)). Cy pres distribution of residual funds pursuant to the settlement agreement neither constitutes speech by any individual class member nor infringes on their First Amendment rights.

C. Attorney's Fee

Finally, St. John challenges the attorney's fee of 25% of the Common Fund to be paid to class counsel. St. John urges the court to exclude the cy pres from the value of the lawsuit in calculating the attorney's fee because the cy pres is not a benefit to the class. But the funds that are ultimately allocated cy pres were available for class members to claim. If the court affirms the adequacy of the notice to the class, then the court cannot fault plaintiffs’ counsel for the fact that class members, for myriad possible reasons, did not submit enough claims to exhaust the Common Fund. Furthermore, by its very name, a cy pres distribution "must be for the next best use," that is, "for indirect class benefit," and "for uses consistent with the nature of the underlying action." BankAmerica, 775 F.3d at 1067 (quotation omitted). Because the cy pres is "distributed for a purpose as near as possible to the legitimate objectives underlying the lawsuit [and] the interests of class members," id. (quotation omitted), the district court did not abuse its discretion in including the amount allocated cy pres in calculating the attorney's fee.

St. John also argues that the district court erred in assigning any value to the parties’ agreement that Monsanto will change the Roundup label since the settlement does not give Plaintiffs any say in the wording of the new label, and in fact, Monsanto had already begun the regulatory process to change the label before the settlement agreement was reached. Again, we disagree. Because the prior label had been approved by the EPA, it was presumptively legal, and Plaintiffs could not have obtained an injunction against the label from the court. The district court did not abuse its discretion in determining that Monsanto agreeing to change its label was an element of the class's overall success. The fact that the settlement does not control the text of any new labeling Monsanto may adopt does not persuade us otherwise.

St. John's final argument is that the district court erred in including work that was done in prior litigation, Blitz v. Monsanto Co., No. 17-473 (W.D. Wis.), in the fee award. In Miller v. Dugan, this court acknowledged the general principle that a fee award could include time spent on separate litigation "if the effort resulted in work product that was actually used in the instant case, the time spent was inextricably linked to issues raised in the instant case, and the plaintiff was not otherwise compensated for counsel's work in the ancillary proceeding." 764 F.3d 826, 832 (8th Cir. 2014). We are satisfied that the district court did not abuse its discretion in concluding that the close relationship between Blitz and this case permitted time spent on Blitz to be included in the lodestar analysis. Blitz also raised state-specific and nationwide class claims based on the allegedly false Roundup label. Monsanto and the Plaintiffs here stipulated to the use of discovery from Blitz, including depositions, and avoided duplicating in this litigation work that had already been done. And the settlement agreement that resolves this case also resolves Blitz, so the attorneys will not be compensated separately for their related work on that case. It was therefore not a clear error in judgment for the district court to include Plaintiffs’ counsel's work on Blitz in its assessment of a reasonable attorney's fee.

For these reasons, we affirm the order of the district court approving the class action settlement in this matter.