ORDER GRANTING MOTION FOR PRELIMINARY APPROVAL OF CLASS ACTION SETTLEMENT
Re: Dkt. No. 206
VINCE CHHABRIA, United States District JudgePreviously the Court denied preliminary approval of a proposed class action settlement agreement in this case that included $12.25 million in monetary relief, as well as modest nonmonetary relief. The parties have returned with a new agreement that includes $27 million in monetary relief and slightly enhanced nonmonetary relief. Based on the current record, the Court concludes that the new settlement agreement is fair, reasonable, and adequate within the meaning of Rule 23(e)(2) of the Federal Rules of Civil Procedure. Accordingly, the motion for preliminary approval is granted.
*1033This ruling presumes that the reader is familiar with the background of the. case, which is discussed more fully in the Court’s prior rulings. See generally Cotter v. Lyft, Inc., 60 F.Supp.3d 1067 (N.D.Cal.2015); Cotter v. Lyft, Inc., No. 13-cv-04065-VC, 176 F.Supp.3d 930, 2016 WL 1394236 (N.D.Cal. April 7, 2016).
I.
The plaintiffs in this case are current or former drivers for Lyft in California. They filed a proposed class action in which they alleged that Lyft violated California law by classifying them as “independent contractors” rather than as “employees.” This misclassification, according to the plaintiffs, deprived Lyft drivers of significant compensation and benefits.
Before litigating the issue of class certification, the plaintiffs and Lyft reached a settlement that proposed to bind the entire class of California-based Lyft drivers. The plaintiffs filed a motion for preliminary approval of the settlement. Along with the motion for preliminary approval, the plaintiffs filed a motion to certify the class for settlement purposes only. They also sought permission to file an amended complaint, which included several claims they hadn’t originally' brought. The idea behind the proposed amended complaint was that it would, in combination with the waiver in the settlement agreement, give Lyft greater protection against future lawsuits. Specifically, it would ensure that all claims based on the assertion that California Lyft drivers are employees rather than independent contractors (and not just the particular misclassification claims that happened to be included in the original lawsuit) would be covered by the settlement agreement.1
The settlement did not reclassify the drivers as employees, but it limited the circumstances in which Lyft could terminate drivers, and it created a process by which drivers could challenge certain termination decisions. The settlement also contemplated a monetary payment by Lyft of $12.25 million, most of which would go to the drivers, but 30 percent of which would go to the plaintiffs’ lawyers.
The Court rejected the proposed agreement, primarily because plaintiffs’ counsel, during settlement negotiations, grossly underestimated the value of the drivers’ claim for reimbursement of expenses. This error resulted in a settlement amount that was unreasonably low considering the substantive strength of the plaintiffs’ claims and the value those claims would have if the plaintiffs ultimately got a judgment in their favor. The Court also expressed concern that the portion of the settlement total set aside to resolve the plaintiffs’ claims under California’s Private Attorneys General Act (“PAGA”), $122,250, was calculated in an arbitrary fashion.
The parties have returned with' a new settlement agreement, and once again seek preliminary approval. The total settlement amount has increased from $12.25 million to $27 million. The agreement contemplates that the plaintiffs’ lawyers will not seek any more in attorneys’ fees than they pledged to seek in the first settlement agreement, meaning that only 14 percent of the new settlement total could potentially go to the attorneys. And' of the $27 million, $1 million is earmarked for the PAGA claims, with the State of California to receive 75 percent of that amount.
But a new issue has come up since the prior motion for preliminary approval. A new lawsuit, Zamora v. Lyft, Inc., No. 16-*1034cv-02558-VC, has been filed by another group of drivers. The Zamora lawsuit alleges that, from August 2014 to the present, Lyft has deprived drivers of certain gratuities or other payments meant for them. Specifically, the new lawsuit alleges that when Lyft imposes a “Prime Time” surcharge for rides given during peak hours, it falsely tells riders that the surcharge goes solely to the driver, when in fact Lyft takes a 20 percent cut of the surcharge (just as it takes a 20 percent cut of the regular fare).
Included in the Zamora lawsuit, are six claims, all based on the factual allegations just described. One is a purely statutory claim. Section 351 of the California Labor Code provides that “gratuities” are “the sole property” of “employees” and may not be taken by employers in whole or in part. But there is no private cause of action under section 351, so the Zamora plaintiffs have asserted this first claim in the name of the state, through PAGA, and seek penalties on behalf of the state and themselves for Lyft’s alleged failure to comply with the Labor Code by taking a portion of the Prime Time surcharges, which the Zamora plaintiffs argue are “gratuities” within the meaning of section 351. And because section 351 applies only to “employees,” the Zamora plaintiffs’ section 351/PAGA claim depends on the contention that Lyft drivers have been misclassified as independent contractors and are in fact employees under California law.
The Zamora plaintiffs’ five other claims—brought under California’s Unfair Competition Law (“UCL”), Cal. Bus. & Prof. Code § 17200 et seq., and under the common law—seek various forms of restitution. In other words, these claims are based on the idea that Lyft must return all or a portion of the commissions it has taken from the drivers in connection with the “Prime Time” surcharges. These claims don’t necessarily depend on whether the drivers should be classified as “employees” or as “independent contractors.” Although two of these five claims do argue in the alternative that the drivers are entitled to relief because Lyft has violated section 351—the Zamora plaintiffs argue that Lyft’s taking a portion of the Prime Time surcharge is “unlawful” under the UCL because it violates section 361, and that Lyft has converted the drivers’ property because the Prime Time surcharges are gratuities and therefore the drivers’ property by virtue of section 351—neither claim rests solely on section 351 or requires a finding that the drivers are employees.
In this case (that is, the Cotter case), the plaintiffs originally included a UCL claim for restitution of gratuities Lyft allegedly took in violation of section 351, and a PAGA claim for penalties for those same alleged section 351 violations. Counsel for the Cotter plaintiffs considered the value of those gratuity claims when negotiating the settlement. But the Cotter gratuity claims related to an earlier time period—specifically, a period in 2013, when Lyft did not charge riders a set fare but instead accepted voluntary donations, which the Cotter plaintiffs argued were gratuities under California law. Counsel for the Cotter plaintiffs admit they never considered the Prime Time gratuity claims that the Zamora plaintiffs now assert. And counsel for the Cotter plaintiffs admit that, when they negotiated this settlement and analyzed its fairness for the proposed class, they never took the Prime Time gratuity claims, or their value, into account. Nonetheless, the settlement agreement for which the Cotter plaintiffs seek preliminary approval would release all gratuity claims for the class that depend oh the contention that the drivers are “employees” rather than “independent contractors.” Similarly, the proposed amended complaint includes any claim for gratuities that depends on the *1035contention that the drivers are employees (as opposed to the original complaint, which only included the gratuity claim from 2013). This means that any driver who is part of the Cotter class would fore-go the PAGA gratuity claim asserted in the Zamora case. The same would be true of the conversion and UCL claims asserted in the Zamora case to the limited extent those claims depend on a contention that Lyft’s conduct violated section 351, since those theories of relief depend on the mis-classification claims that are released by the settlement agreement.
The plaintiffs in the Zamora case have filed a motion to intervene in the Cotter case. The Zamora plaintiffs contend their interests have not been adequately protected by the Cotter plaintiffs and their lawyers, because the Cotter plaintiffs and their lawyers are proposing a settlement agreement that would partially waive their new Prime Time gratuity claims for. all drivers, without ever having considered the strength or value of the claims. The Court declined the Zamora plaintiffs’ request to shorten time on the hearing on their motion to intervene. But it allowed counsel for the Zamora plaintiffs to appear at the hearing on the motion for preliminary approval of the Cotter settlement, to ensure a thorough assessment of the agreement’s fairness and adequacy. Following the hearing,' the Court requested supplemental briefs from the parties, to further explore the strength and value of the new gratuity claims.
II.
District courts review class action settlements in two stages. First, the plaintiffs file a motion for something that -is typically described as “preliminary approval.” Along with that motion, the plaintiffs typically fíle a motion for class certification, requesting that the class be certified for purposes of settlement only. If the district court grants preliminary approval and certifies the class, typically the class members are notified and given a chance to object to the settlement, or to opt out of the settlement. Then the plaintiffs file a motion for final approval. After hearing objections (if any) to the settlement, the district court decides whether to grant final approval.
The standard for reviewing class action settlements at the final approval stage is well-settled. Rule 23(e)(2) of the Federal Rules of Civil Procedure states that the district court may only approve the settlement if “it is fair, reasonable, and adequate.” See also Hanlon v. Chrysler Corp., 150 F.3d 1011, 1026 (9th Cir.1998). “It is the settlement taken as a whole, rather, than the individual component parts, that must be examined for overall fairness.” Id. The district court must balance “the strength of the plaintiffs’ case; the risk, expense, complexity, and likely duration of further litigation; the risk of maintaining class action status throughout the trial; the amount offered in settlement; the extent of discovery completed and the stage of the proceedings; the experience and views of counsel; the presence of a governmental participant; and the reaction of the class members to the proposed settlement.” Id. (citing Torrisi v. Tucson Elec. Power Co., 8 F.3d 1370, 1375 (9th Cir.1993)). Additionally, “settlement approval that takes place prior to formal class certification requires a higher standard of fairness. The dangers of collusion between class counsel and the defendant, as well as the need for additional protections when the settlement is not negotiated by a court[-]designated class representative, weigh in favor of a more probing inquiry than may normally be required under Rule 23(e).” Id.
The Ninth Circuit .has not specified whether a different .standard of review should apply at the preliminary approval stage. However, district courts often state *1036or imply that scrutiny should be more lax. For example, in one recent high-profile case, a district court noted: “At the preliminary approval stage, the bar to meet the ‘fair, reasonable and adequate’ standard is lowered.” In re Nat’l Football League Players’ Concussion Injury Litig., 961 F.Supp.2d 708, 714 (E.D.Pa.2014). Typically, at the preliminary stage, courts therefore ask merely whether the agreement “falls within the range of possible approval.” Id. (quoting In re Tableware Antitrust Litig., 484 F.Supp.2d 1078, 1080 (N.D.Cal.2007)). This approach seems to contemplate that a district court might preliminarily approve a class settlement even if the court has serious concerns about whether final approval would be appropriate, putting off a decision about whether those concerns should tank the settlement. See, e.g., Harris v. Vector Mktg. Corp., No. 08-cv-5198-EMC, 2011 WL 4831157, at *1 (N.D.Cal. Oct. 12, 2011) (preliminary approval granted despite concerns about adequacy; final approval denied). This approach also seems to contemplate that district courts should take nothing more than a quick look at the settlement to see if it has any “obvious deficiencies.” In re Prudential Sec. Inc. Ltd. P’ships Litig., 163 F.R.D. 200, 209 (S.D.N.Y.1995) (quoting Manual for Complex Litigation (Third) § 30.41 (Fed. Judicial Ctr. 1995)).
Nobody appears to have offered a rationale for this approach. Perhaps it’s just assumed from the word “preliminary,” which suggests that something fuller or more important will come later. But “preliminary” is just a label, not a rationale. Nor does the word “preliminar/’ appear in Rule 23(e)(2). Preliminary approval could just as easily have been labeled “initial” approval or “first round” approval.
In any event, the idea that district courts should conduct a more lax inquiry at the preliminary approval stage seems wrong. Certainly nothing in the text of Rule 23 suggests courts should be more forgiving of flaws in a settlement agreement at the preliminary stage than at the final stage, or that courts should merely give settlement agreements a “quick look” at the outset. And lax review makes little practical sense, from anyone’s standpoint. If the district court, by taking a quick look rather than a careful one, misses a serious flaw in the settlement, the parties and the court will waste a great deal of money and time notifying class members of the agreement, only to see it rejected in the end, requiring the parties to start over. The same is true if the district court does identify a potentially serious flaw at the preliminary stage but waits until final approval to conclude that it’s fatal. What’s worse, if a court waits until the final approval stage to thoroughly assess the fairness of the agreement, momentum could have a way of slanting the inquiry, in a manner that deprives the class members of the court protection that Rule 23 demands.
This approach may also inadvertently disadvantage class members. Class members will receive a notice saying that the settlement has received preliminary approval from a federal judge. A layperson may take the court’s preliminary approval to imply that she shouldn’t really worry about whether the settlement is in her best interest, because surely the court, which is more familiar with the law and the facts of the case, has already taken care of that. But that is a misimpression if the judge has merely glanced at the settlement or decided to hold off adjudicating a potential problem until final approval.
This is not to suggest that rigorous inquiry at the initial stage should convert final review to a mere formality. Sometimes objectors may bring a flaw to the court’s attention at the final stage—one the court didn’t catch at the initial stage. Other times, further factual development *1037between the initial and final stages may cause the court to conclude that the agreement is unfair after all. But by scrutinizing the agreement carefully at the initial stage and identifying any flaws that can be identified, the court allows the parties to decide how to respond to those flaws (whether by fixing them or opting not to settle) before they waste a great deal of time and money in the notice and opt-out process.
In sum, district courts should review class action settlements just as carefully at the initial stage as they do at the final stage. At the initial stage, the inquiry should be whether the settlement is “fair, reasonable, and adequate,” based on any information the district court receives from the parties or can obtain through its own research. If the parties don’t provide enough ’ information to allow the district court to carefully evaluate the strength of the claims, the risks of litigating those claims all the way through, and the value of the relief each class member will receive from the settlement, the court should deny the motion for preliminary approval or demand supplemental briefing, rather than kicking the can down the road. See, e.g., Eddings v. D.S. Servs. of Am., Inc., No. 15-cv-02576-VC, 2016 WL 3390477 (N.D.Cal. May 20, 2016). And perhaps it ought to be called something other than “preliminary approval,” to avoid the implication that the first round of review is cursory.
III.
A.
The new settlement agreement in the Cotter case adequately addresses the flaws identified in the Court’s prior ruling. The parties have increased the total settlement amount from $12.25 to $27 million, primarily to account for the error made by plaintiffs’ counsel in estimating the value of the reimbursement claim. The parties have also allocated $1 million of the settlement amount to the PAGA claims, which will result in a $750,000 payment to the State of California, with the remaining $250,000 going into the settlement fund. Although $1 million might seem low when compared to the maximum PAGA penalties that would accrue-in the event the plaintiffs won on the merits of their claims that Lyft violated the Labor Code by mis-classifying them as independent contractors, it’s quite high compared to the typical amount apportioned to claims for PAGA penalties in wage and hour settlements. More importantly, courts have discretion to reduce an award of PAGA penalties when it would be “unjust, arbitrary and oppressive, or confiscatory” to impose the full amount. Cal. Lab. Code § 2699(e)(2). A significant reduction would be appropriate here. This is not a case where a company has deliberately evaded a clear legal obligation to provide legally required pay and benefits to its employees. Nor does this appear to be a case where a company negligently failed to learn about its obligations under the wage and hour laws. Rather, as the prior rulings in this case have discussed, Lyft’s obligations to its drivers are genuinely unclear (and will likely remain so absent legislative or executive action) because there is no straight answer .to the question whether those drivers must be classified as employees or independent contractors under California law. On these facts, if -the case went to judgment and if the drivers prevailed, the Court would very likely reduce the award of PAGA penalties by a substantial amount. The $1 million settlement figure for the PAGA claims is reasonable.2
*1038B.
The Zamora plaintiffs object only to the fact that the agreement in Cotter releases the Lyft drivers’ newly identified Prime Time gratuity claims based on California Labor Code section 351. The Zamora plaintiffs argue that Lyft and the Cotter plaintiffs should agree to either: (i) carve out these claims from the release language of the settlement agreement; or (ii) add money to the settlement total to account for the value of these claims'for1 the class. They contend (implicitly at least) that the agreement should be rejected if one of those two things does not happen.
The Cotter plaintiffs and Lyft complain that the objection by the Zamora plaintiffs has come too late. They note that the Zamora plaintiffs could have made this objection months earlier, in response to the motion for preliminary approval of the first settlement agreement. And they point out that if the Zamora plaintiffs truly had strong gratuity claims, they could have brought those claims years ago. The issue of timeliness may well be relevant to whether the Court should grant the Zamora plaintiffs’ motion to intervene in the case—a motion that will be’ addressed in a separate ruling. But on a motion for preliminary approval of a class action settlement, a district court’s obligation is to assess the agreement to ensure that the class members are being treated fairly. Even if the court receives information relevant to the fairness question at the eleventh hour, the court should consider it.
On the question whether the failure by counsel for the Cotter plaintiffs to consider the Zamora gratuity claims should tank the settlement as currently drafted, it bears repeating that district courts must review “the settlement taken as a whole, rather than the individual component parts,” for fairness under Rule 23(e)(2). Hanlon, 150 F.3d at 1026. Accordingly, even though the Cotter plaintiffs didn’t consider these particular claims when they filed suit, and even though the lawyers for the. Cotter plaintiffs concede they neglected to assess the value of the claims when they negotiated the agreement on behalf of the proposed class, this does not automatically invalidate the settlement they reached. If the unconsidered claims are not particularly strong or valuable, such that they’re not likely to have materially influenced the overall settlement, counsel’s failure to consider the claims would not be a basis for rejecting the agreement. Nor is there anything necessarily unseemly (or unusual) about a class action settlement agreement that releases claims the plaintiff did not originally bring. What matters is whether the released claims arise from the same facts as those alleged in the lawsuit, and whether the settlement as a whole is reasonable in light of the strength and value of all the claims being released. See, e.g., Reyn’s Pasta Bella, LLC v. Visa USA Inc., 442 F.3d 741, 748 (9th Cir.2006) (quoting Class Plaintiffs v. City of Seattle, 955 F.2d 1268, 1287 (9th Cir.1992)). On the other hand, if the claims that plaintiffs’ counsel failed to consider are strong and valuable, counsel’s failure to include those claims in the settlement calculus could require the Court to reject the agreement, because it could mean that the overall settlement is inadequate within the meaning of Rule 23(e)(2).
Based on the evidence currently in the record, the new gratuity claims asserted by the Zamora plaintiffs do not seem very strong. To reiterate, the Zamora lawsuit alleges that, since August 2014, Lyft has *1039required riders to pay a “Prime Time” surcharge, telling riders that the surcharge is entirely for the drivers while in fact taking 20 percent of the surcharge for itself. The suit alleges that these actions violated, among other things, Labor Code section 351, which prohibits an employer from taking gratuities given to employees. But based on the record developed so far, it appears that for most or all of the relevant time period, Lyft did not explicitly describe the Prime Time surcharge as a “tip” or a “gratuity.” Rather, after August 2014, it appears that Lyft’s app displayed a message to riders, when they initially “hailed” a ride on the app, that read something like the following: “It’s busy! 100% will be added to your total for the driver.” And it appears that after the ride, the app separately prompted riders to “tip” the driver. In this context, the idea that a reasonable rider would view the Prime Time surcharges (after the August 2014 change) as a mandatory gratuity, as opposed to simply a price increase, is questionable. This makes the Lyft drivers’ Prime Time gratuity claims far weaker than the gratuity claim asserted by the Uber drivers in O’Connor v. Uber Technologies, Inc., No. C-13-3826-EMC, 2013 WL 6354534, at *7-10 (N.D.Cal. Dec. 5, 2013). And it makes the Lyft drivers’ Prime Time gratuity claim far weaker than their claim for mileage reimbursement in this case.3
Moreover, it bears recalling that the new section 351 gratuity claims released by the current settlement agreement could only succeed on the merits if a jury ultimately concluded that Lyft drivers are employees rather than independent contractors under California law. It is therefore subject to all the risks discussed in the-Court’s prior ruling on the first motion for preliminary approval. See Cotter, 2016 WL 1394236, at *10-11. This is in contrast to the other claims being pursued by the Zamora plaintiffs, which seek to vindicate the same alleged wrongdoing but which are not released by the settlement agreer ment (at least to the extent they rely on theories that do not require a finding that the drivers are employees).
In addition to not appearing very strong, these new gratuity claims do not appear very valuable, at least relative to the reimbursement claim that has driven settlement negotiations in the Cotter ease. According to the evidence received so far, while the reimbursement claim has a maximum value of $156 million, the total restitution value of the Zamora claims as to the California class is just over $10 million.4 The new settlement reasonably contemplates payment of roughly 17 percent of the value of the $156 million reimbursement claim; it would have been reasonable for the plaintiffs to agree to a much larger discount on the much weaker $10 million gratuity claim. And even this may overstate the significance of releasing the new gratuity claims. Recall that the. Cotter set*1040tlement would not preclude the Zamora plaintiffs from pursuing (potentially on behalf of a proposed class of drivers) the claims based on this same alleged wrongdoing that do not depend on a finding that Lyft drivers are “employees” rather than “independent contractors.” In other words, notwithstanding this settlement, drivers may still attempt to collect on the commissions that Lyft allegedly kept from the Prime Time premiums.
In sum, the relative weakness of the section 351 claims, the low value of those claims relative to the value of the reimbursement claims, and the fact that drivers might still be able to pursue reimbursement of the Prime Time commissions under other legal theories, combine to lead to the conclusion that, even though counsel for the Cotter plaintiffs ought to have identified and considered the Prime Time gratuity claims before negotiating a settlement agreement, their failure to do so does not render the current settlement unfair, unreasonable, or inadequate within the meaning of Rule 23(e)(2).
IV.
The new proposed settlement agreement fixes the monetary flaws the Court previously identified and enhances the nonmonetary benefits at least to some degree. Considering the risks the plaintiffs would face in taking this case to trial, together with the value of all of the claims being released and the value of the proposed settlement (both monetary and non-monetary) to the class members, the Court concludes that the settlement, on the current record, is “fair, reasonable, and adequate” within the meaning of Rule 23(e)(2). See also Hanlon, 150 F.3d at 1026. Accordingly, the motion- for preliminary approval is granted. Within 7 days of this order, the parties should submit a revised proposal for notifying the class members of the settlement. The revised notice should include a description of the Zamora lawsuit, and the fact that the current agreement releases the section 351 claims being pursued by the Zamora plaintiffs (without releasing other claims based on the same facts).
IT IS SO ORDERED.
. The proposed waiver would not apply to claims that arise after the date of preliminary approval, whether or not based on the theory that the drivers should be classified as employees. See Cotter, 176 F.Supp.3d at 934-35, 2016 WL 1394236, at *4.
. The parties have also enhanced the non-monetary benefits by extending the optional pre-arbitration process to all drivers, whether they have been deactivated or not. And Lyft has agreed to remove a provision from its contracts with drivers that purports to permit Lyft to deactivate drivers who "create liabili*1038ty” for Lyft or cause Lyft to be regulated as a transportation carrier or provider of taxi service. While neither of these additional benefits is particularly groundbreaking, they do marginally increase the value of the nonmonetary benefits conferred on the class as compared to the prior settlement agreement.
. The Zamora plaintiffs allege in their complaint that, from August 2014 until approximately August 2015, Lyft actually did tell riders that the Prime Time surcharge was a "tip” that went "entirely” to the driver. However, while the record is not entirely clear and might require further development, at this stage Lyft has presented evidence that appears to refute the Zamora plaintiffs’ allegation about the August 2014-August 2015 time period. It appears that: (i) before August 2014, the Lyft app told riders that the Prime Time surcharge was a "tip” that was "entirely” for the drivers, when in fact that was true; and (ii) in August 2014, Lyft updated the app to coincide with its decision to take a 20 percent cut of the Prime Time surcharge, so that riders were no longer told the surcharge was a "tip” that went "entirely” to drivers; but (iii) some riders may have delayed in updating their apps, which may have caused them to continue to receive the old message beyond August 2014.'
. Again, it’s possible that new evidence could alter this assessment.