Before CLEMENT, DUNCAN, and OLDHAM, Circuit Judges. STUART KYLE DUNCAN, Circuit Judge: David Cruson and other Texas residents brought a putative class action lawsuit against Jackson, a life insurance company that sells annuities, claiming Jackson overcharged them by miscalculating early-withdrawal fees in breach of the annuities contracts. The district court certified a nationwide class of similarly-situated investors and determined that Jackson had waived its personal jurisdiction defense as to any non-Texas class members. On interlocutory appeal, we conclude that the district court’s order is flawed in the following respects. First, we conclude that Jackson did not waive its personal jurisdiction defense as to non-Texas class members. Second, we conclude that the district court erred in its predominance analysis—specifically, by failing to assess how state-law variations may impact adjudication of the breach question and also by failing to consider the individualized evidence relevant to Jackson’s affirmative defenses of waiver and ratification. Third, we conclude that the plaintiffs failed to offer a damages model adequate to support class treatment, an issue they virtually conceded at oral argument. We therefore vacate the class certification order and remand for further proceedings. I. Jackson National Life Insurance Company, a Michigan corporation, sells variable annuities to customers nationwide through a network of brokers and other intermediaries. Because the annuities are structured as long-term investments, a customer who withdraws money early incurs charges meant to compensate Jackson for up-front costs, such as commissions paid to brokers and enhancements added to the annuities. Different charges, collectively called “surrender charges,” are calculated according to a schedule of percentages that decrease with the annuity’s age.[1] In other words, the longer a customer has held an annuity, the lower the surrender charge. In November 2016, fourteen Texas residents sued Jackson in federal district court for breach of contract, breach of fiduciary duty, and negligent misrepresentation. The plaintiffs—who had all bought Jackson annuities through a Texas broker, Tim Hightower—alleged Jackson had miscalculated their surrender charges in violation of the annuity contracts, resulting in lost income for them and a windfall for Jackson. The plaintiffs also alleged their living and death benefits under the annuities had been improperly reduced due to the inflated surrender charges. They sought compensatory and punitive damages, as well as injunctive relief. They also sought to bring claims on behalf of a nationwide class, consisting of all Jackson customers who had incurred surrender charges. A complex procedural history followed. Jackson moved to dismiss the complaint under Rules 12(b)(1) and 12(b)(6) for lack of standing and failure to state a claim. Plaintiffs then filed an amended complaint, and Jackson again moved to dismiss under Rules 12(b)(1) and 12(b)(6). The district court denied Jackson’s motion in large part but did dismiss the claims of three plaintiffs on standing grounds because they had admittedly never incurred surrender charges.[2] In neither of Jackson’s Rule 12 motions did it raise lack of personal jurisdiction as a defense. In its subsequent answer, however, Jackson stated that, “[t]o the extent that a class outside of Texas is certified, [it was] denied that [the district court] has personal jurisdiction over Jackson for the remaining two-thirds of the putative class members residing outside of Texas.” Plaintiffs then moved to certify a nationwide class under Rule 23(b)(3), alleging Jackson’s inflated surrender charges had harmed roughly 150,000 people. In support, plaintiffs introduced an expert report which proposed a damages calculation for inflated surrender charges but not for damages associated with living and death benefits.[3] Jackson opposed class certification, arguing inter alia that contract formation issues, affirmative defenses, and damages calculations would require individual determinations that would predominate over common issues. Moreover, Jackson again raised lack of personal jurisdiction, arguing that specific jurisdiction over Jackson as to claims by non-Texas residents was foreclosed by the Supreme Court’s recent decision in Bristol-Myers Squibb Co. v. Superior Court of California, 137 S. Ct. 1773 (2017).[4] On the same day that Jackson filed its response opposing class certification, Jackson also moved for summary judgment. The district court subsequently held a class certification hearing in March 2018. After the hearing, but before any certification ruling, all but one of the original plaintiffs moved to voluntarily dismiss their claims, stating they would seek relief in state court. While that motion was pending, plaintiffs moved to amend their complaint to add an additional plaintiff, John Denman, as a potential class representative, to join David Cruson (the one original plaintiff not moving to dismiss his claims). While all initial plaintiffs had been clients of Tim Hightower, Denman had bought annuities through a different broker. Jackson opposed this addition, claiming it was too late to add a named party and that doing so would prejudice Jackson. On May 9, 2018, the district court issued an order addressing three issues. First, the court granted plaintiffs’ motion to add Denman. Second, the court held that Jackson had waived any personal jurisdiction defense by failing to raise it in its Rule 12 motions and, alternatively, by litigating on the merits of plaintiffs’ claims. Third, the district court certified a nationwide class composed of “[a]ll persons who, within the applicable statute of limitations, purchased [various] variable annuity products from [Jackson] or its affiliates, and incurred a Surrender Charge during their ownership of such product.”[5]The court found that determining the correct calculation of surrender charges under Jackson’s contracts, an issue common to all plaintiffs, would “predominate over any questions affecting only individual members.” See Fed. R. Civ. P. 23(b)(3). The court further held that predominance was not defeated by issues pertaining to affirmative defenses or damages calculations. We granted Jackson permission to appeal the district court’s class certification order under Federal Rule of Civil Procedure 23(f). On appeal, Jackson challenges the certification order on three fronts. First, Jackson reasserts its claim that the district court lacks personal jurisdiction over it with respect to the claims of potential class members outside Texas, and argues that the court erred in holding that Jackson waived its personal jurisdiction defense. Second, Jackson contends that the district court erred in concluding that issues common to the class predominated over individualized issues. Finally, Jackson argues that the district court erred in certifying a class when plaintiffs failed to provide a damages model.[6] II. Our court has not settled the standard for reviewing a district court’s finding of waiver of personal jurisdiction. See In re Depuy Orthopaedics, Inc., 870 F.3d 345, 351 n.43 (5th Cir. 2017) (stating “[w]e need not decide th[e] standard-of-review question” for personal jurisdiction waiver); id. at 359 (Jones, J., concurring in part and dissenting in part) (noting majority “takes no firm position on . . . [whether] waiver of personal jurisdiction should be reviewed for abuse of discretion”). Several circuits review waiver for abuse of discretion, reasoning that the finding pertains to the “[district court's] broad duties in managing the conduct of cases pending before it.” Lechoslaw v. Bank of Am., N.A., 618 F.3d 49, 55-56 (1st Cir. 2010) (quoting United States v. Ziegler Bolt & Parts Co., 111 F.3d 878, 882 (Fed. Cir. 1997)); see also, e.g., InfoSpan, Inc. v. Emirates NBD Bank PJSC, 903 F.3d 896, 900 (9th Cir. 2018); Hamilton v. Atlas Turner, Inc., 197 F.3d 58, 60 (2d Cir. 1999). By contrast, the Tenth Circuit treats the issue as a mixed question of law and fact, reviewing pertinent factual findings for clear error but legal conclusions de novo. Am. Fid. Assurance Co. v. Bank of N.Y. Mellon, 810 F.3d 1234, 1237 (10th Cir. 2016). Much recommends the Tenth Circuit’s nuanced approach, given the presence of personal jurisdiction is itself a legal question reviewed de novo. See Patin v. Thoroughbred Power Boats Inc., 294 F.3d 640, 652 (5th Cir. 2002) (“Whether in personam jurisdiction can be exercised over a defendant is a question of law subject to de novo review by this court.”); see also, e.g., Halliburton Energy Servs., Inc. v. Ironshore Specialty Ins. Co., 921 F.3d 522, 539 (5th Cir. 2019) (reviewing de novo “a district court’s decision to grant a motion to dismiss for lack of personal jurisdiction”). We need not choose sides in this debate, however. As explained below, even under an abuse of discretion standard we conclude the district court committed reversible error in finding Jackson waived the defense of personal jurisdiction. We review the district court’s class certification order for abuse of discretion. Crutchfield v. Sewerage & Water Bd. of New Orleans, 829 F.3d 370, 375 (5th Cir. 2016); Union Asset Mgmt. Holding A.G. v. Dell, Inc., 669 F.3d 632, 638 (5th Cir. 2012). We review the legal conclusions in the order de novo, “[b]ecause . . . a court by definition abuses its discretion when it applies an incorrect legal standard.” Steering Comm. v. Exxon Mobil Corp., 461 F.3d 598, 601 (5th Cir. 2006). While retaining broad discretion to certify a class, “[a] district court must rigorously analyze Rule 23′s prerequisites before certifying a class.” Spence v. Glock, Ges.m.b.H., 227 F.3d 308, 310 (5th Cir. 2000) (citing Gen. Tel. Co. v. Falcon, 457 U.S. 147, 161 (1982); Castano v. Am. Tobacco Co., 84 F.3d 734, 740 (5th Cir. 1996)). Finally, “if the district court has committed legal error in the predominance inquiry, reversal is required.” Id. at 311. III. A. We first consider whether Jackson waived its defense of personal jurisdiction, either by failing to raise the defense in its Rule 12 motions or else by litigating the merits. Jackson denies both conclusions. It argues that the personal jurisdiction defense was not “available” as to the putative class of non- Texas residents (and so could not be waived) when it filed its Rule 12 motions. Jackson also argues that its litigation conduct was consistent with preserving its objection to personal jurisdiction, which it raised in both its answer and opposition to class certification. We agree with Jackson that it did not waive its personal jurisdiction defense.[7] Federal Rule of Civil Procedure 12(h)(1) relevantly provides that a party waives the defense of lack of personal jurisdiction by “omitting it from a motion in the circumstances described in Rule 12(g)(2).” Fed. R. Civ. P. 12(h)(1)(A). Rule 12(g)(2) prohibits a party from a raising a defense “that was available to the party but omitted from its earlier [Rule 12] motion.” Fed. R. Civ. P. 12(g)(2). Thus, lack of personal jurisdiction is waived if a party omits the defense from a Rule 12 motion and the defense was “available.” See Jackson v. FIE Corp., 302 F.3d 515, 523 (5th Cir. 2002); Cactus Pipe & Supply Co. v. M/V Montmartre, 756 F.2d 1103, 1107-08 (5th Cir. 1985). See generally 5C Charles Alan Wright & Arthur R. Miller, Federal Practice and Procedure § 1391 (3d ed. 2004) [hereinafter Wright & Miller] (discussing waiver under Rule 12(h)(1)). A defense is not “available” under Rule 12(g)(2), for instance, if “its legal basis did not [then] exist,” Gilmore v. Palestinian Interim Self-Gov’t Auth., 843 F.3d 958, 964 (D.C. Cir. 2016), or if the defense would have been “futile” in the sense it was precluded by controlling precedent, In re Micron Tech., Inc., 875 F.3d 1091, 1097-98 (Fed. Cir. 2017) (collecting decisions). See generally 5C Wright & Miller § 1388 (discussing “available” factor under Rule 12(g)). The issue, then, is whether the personal jurisdiction defense was “available” under Rule 12(g)(2) when Jackson filed its Rule 12 motions. We conclude it was not. Jackson’s objection to personal jurisdiction concerned only class members who were non-residents of Texas. Those members, however, were not yet before the court when Jackson filed its Rule 12 motions. What brings putative class members before the court is certification: “Certification of a class is the critical act which reifies the unnamed class members and, critically, renders them subject to the court’s power.” In re Checking Account Overdraft Litig., 780 F.3d 1031, 1037 (11th Cir. 2015); accordMolock, 2020 WL 1146733, at *3 (“Putative class members become parties to an action—and thus subject to dismissal—only after class certification.” (citing In re Bayshore Ford Trucks Sales, Inc., 471 F.3d 1233, 1245 (11th Cir. 2006))).[8] When Jackson filed its pre-certification Rule 12 motions, however, the only live claims belonged to the named plaintiffs, all Texas residents as to whom Jackson conceded personal jurisdiction. See, e.g., Police & Fire Ret. Sys. of Detroit v. IndyMac MBS, Inc., 721 F.3d 95, 112 n.22 (2d Cir. 2013) (noting “until certification there is no class action but merely the prospect of one; the only action is the suit by the named plaintiffs” (quoting Morlan v. Universal Guar. Life Ins. Co., 298 F.3d 609, 616 (7th Cir. 2002))).[9] Thus, at that time, a personal jurisdiction objection respecting merely putative class members was not “available,” as Rule 12(g)(2) requires for waiver. We find additional support for this conclusion in an analogous Eleventh Circuit decision, In re Checking Account Overdraft Litigation. That case addressed whether a defendant, who had waived arbitration against named plaintiffs prior to class certification, should also be “preclude[d] . . . from compelling arbitration of the unnamed putative class members’ claims.” 780 F.3d at 1034. The district court had extended the waiver to the putative class members, id. at 1036, but the Eleventh Circuit reversed. It explained that, “[a]bsent class certification, there is no justiciable controversy between [the defendant] and the unnamed putative class members,” who “[were] not yet before the court.” Id. at 1037. Consequently, any ruling by the district court purporting to foreclose arbitration of “hypothetical” claims by “future” class members was, the court concluded, “an impermissible ‘advisory opinion on an abstract proposition of law.’” Id. (quoting Hall v. Beals, 396 U.S. 45, 48 (1969) (per curiam)) (cleaned up). Applying the Eleventh Circuit’s reasoning here, it is evident why a personal jurisdiction objection was not “available” with respect to the putative claims of unnamed Texas non-residents. Prior to certification, those nonresidents were “not yet before the [district] court,” their possible “future” claims against Jackson were “hypothetical,” and so there was no “justiciable controversy between [Jackson] and [them].” Id. To rule that Jackson was required, on pain of waiver, to raise a personal jurisdiction objection against those putative class members would validate “the novel and surely erroneous argument that a nonnamed class member is a party to the class-action litigation before the class is certified.” Bayer Corp., 564 U.S. at 313 (quoting Devlin, 536 U.S. at 16 n.1 (Scalia, J., dissenting)). That we decline to do.[10] Alternatively, the district court ruled that Jackson waived the defense “by delaying in challenging personal jurisdiction and substantially litigating on the merits.” In particular, the district court faulted Jackson for “waiting exactly seven months after Bristol-Myers” to raise a personal jurisdiction objection. This reasoning is flawed. Jackson did not “delay” in raising an objection to personal jurisdiction—it raised the defense in its answer and in its opposition to class certification. Prior to class certification, we have already explained, a personal jurisdiction defense as to putative non-resident class members was not “available” under Rule 12. Although Bristol-Myers provided new legal support for Jackson’s objection, the Supreme Court’s decision did not make the objection “available.” Certification did. Furthermore, nothing in Jackson’s subsequent litigation conduct was inconsistent with maintaining an objection to personal jurisdiction. We have “long held that a non-resident defendant may participate in litigation without submitting to the court’s jurisdiction so long as it maintains its objection to personal jurisdiction.” Halliburton, 921 F.3d at 540. The district court focused on Jackson’s summary judgment motion, but the same day that was filed Jackson also filed its opposition to class certification, which raised lack of personal jurisdiction as to “non-Texas putative class members.” In other words, Jackson’s summary judgment motion did not waive its personal jurisdiction defense because its contemporaneous litigation conduct “reflect[ed] a continuing objection to the power of the court to act over the defendant’s person.” PaineWebber Inc. v. Chase Manhattan Private Bank (Switz.), 260 F.3d 453, 460 (5th Cir. 2001) (quoting Alger v. Hayes, 452 F.2d 841, 844 (8th Cir. 1972)); see also Halliburton, 921 F.3d at 540 (explaining a non-resident defendant may “simultaneously protest personal jurisdiction while vigorously advocating the merits of his case” (quoting Toshiba Int’l Corp. v. Fritz 993 F. Supp. 571, 573-74 (S.D. Tex. 1998))). In sum, we conclude the district court legally erred, and thus abused its discretion, in finding Jackson waived its personal jurisdiction defense. B. We next consider the propriety of the district court’s order certifying a nationwide class of Jackson customers who have incurred surrender charges. Jackson argues that the district court erred in its Rule 23(b)(3) analysis by concluding that common issues would predominate. According to Jackson, individual issues predominate both in determining breach and in evaluating affirmative defenses. Although we recognize that cases involving form contracts are often prime candidates for class certification because of the common interpretive questions at stake, we agree with Jackson that the district court’s predominance analysis was deficient. “To obtain class certification, parties must satisfy [Federal Rule of Civil Procedure] 23(a)’s four threshold requirements, as well as the requirements of Rule 23(b)(1), (2), or (3).” Ludlow v. BP, P.L.C., 800 F.3d 674, 682 (5th Cir. 2015) (alteration in original) (quoting Funeral Consumers All., Inc. v. Serv. Corp. Int’l, 695 F.3d 330, 345 (5th Cir. 2012)). See generally Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 613-14 (1997). After finding Rule 23(a) satisfied,[11] the district court certified the nationwide class under Rule 23(b)(3), which embodies two requirements: (1) “[c]ommon questions must ‘predominate over any questions affecting only individual members’”; and (2) “class resolution must be ‘superior to other available methods for the fair and efficient adjudication of the controversy.’” Amchem, 521 U.S. at 615 (quoting FED. R. CIV. P. 23(b)(3)). Jackson’s arguments on appeal address only the district’s court’s findings on the “predominance” prong, which Jackson contends are flawed in two ways. First, Jackson argues the court failed to recognize that individual issues would predominate over common issues of contract interpretation. Second, Jackson argues the court misconstrued the law applicable to Jackson’s affirmative defenses and thus failed to appreciate the individual issues that would be raised by those defenses. Before addressing those issues, we summarize the standards that guide our analysis. “Where the plaintiff seeks to certify a class under Rule 23(b)(3), the Rules demand ‘a close look at the case before it is accepted as a class action.’” Madison v. Chalmette Refining, L.L.C., 637 F.3d 551, 554 (5th Cir. 2011) (quoting Amchem, 521 U.S. at 615). We have long cautioned that “[a] district court must conduct a rigorous analysis of the rule 23 prerequisites before certifying a class.” Castano, 84 F.3d at 740. Furthermore, “[t]he plain text of Rule 23 requires the court to ‘find,’ not merely assume, the facts favoring class certification.” Unger v. Amedisys Inc., 401 F.3d 316, 321 (5th Cir. 2005) (quoting FED. R. CIV. P. 23(b)(3)). “[I]t is the party seeking certification who bears the burden of establishing that the requirements of Rule 23 have been met.” Bell Atl. Corp. v. AT&T Corp., 339 F.3d 294, 301 (5th Cir. 2003). Our focus here is the Rule 23(b)(3) predominance requirement, which “tests whether proposed classes are sufficiently cohesive to warrant adjudication by representation.” Torres v. S.G.E. Mgmt., L.L.C., 838 F.3d 629, 636 (5th Cir. 2016) (en banc) (quoting Amchem, 521 U.S. at 623). Predominance is a “far more demanding” hurdle than Rule 23(a)’s commonality requirement. Amchem, 521 U.S. at 623-24; see also, e.g., Maldonado v. Ochsner Clinic Found., 493 F.3d 521, 525 (5th Cir. 2007) (predominance is “more demanding than the commonality requirement of Rule 23(a)” (quoting Bell Atl., 339 F.3d at 301)). Assessing predominance “calls upon courts to give careful scrutiny to the relation between common and individual questions in a case.” Tyson Foods, Inc. v. Bouaphakeo, 136 S. Ct. 1036, 1045 (2016). The Supreme Court has distinguished common and individual questions this way: An individual question is one where “members of a proposed class will need to present evidence that varies from member to member,” while a common question is one where “the same evidence will suffice for each member to make a prima facie showing [or] the issue is susceptible to generalized, class-wide proof.” Id. (alteration in original) (quoting 2 NEWBERG § 4:50); see also, e.g., Crutchfield, 829 F.3d at 376-78 (discussing individual versus common questions).[12] When “one or more of the central issues in the action are common to the class and can be said to predominate, the action may be considered proper under Rule 23(b)(3) even though other important matters will have to be tried separately, such as damages or some affirmative defenses peculiar to some individual class members.” Tyson Foods, 136 S. Ct. at 1045 (quoting 7AA WRIGHT & MILLER § 1778). “At bottom, the [predominance] inquiry requires the trial court to weigh common issues against individual ones and determine which category is likely to be the focus of a trial.” Crutchfield, 829 F.3d at 376; see also, e.g., Sandwich Chef of Tex., Inc. v. Reliance Nat’l Indem. Ins. Co., 319 F.3d 205, 218 (5th Cir. 2003) (court considers “how a trial on the merits would be conducted if a class were certified”). The court must “go[ ] beyond the pleadings” to “understand the claims, defenses, relevant facts, and applicable substantive law.” Madison, 637 F.3d at 555 (quoting Unger, 401 F.3d at 321). “This, in turn, ‘entails identifying the substantive issues that will control the outcome, assessing which issues will predominate, and then determining whether the issues are common to the class,’ a process that ultimately ‘prevents the class from degenerating into a series of individual trials.’” Id. (quoting O’Sullivan v. Countrywide Home Loans, Inc., 319 F.3d 732, 738 (5th Cir. 2003)). Having set out the governing standards, we turn to Jackson’s arguments. 1. Jackson first argues that, contrary to the district court’s ruling, individual issues will predominate over common issues in determining whether Jackson breached the annuity contracts by miscalculating surrender charges. The court relied on Jackson’s stipulations that there are “no material differences” in the pertinent contract language and that Jackson calculates surrender charges “uniformly”; it also noted that “neither party asserts that the contract is ambiguous.” The court thus reasoned that the only evidence necessary to prove breach is “whether class members purchased an annuity policy with Jackson and whether they suffered a surrender charge.” The only disputed issue, the court continued, was whether “Jackson’s method of calculating surrender charges is permitted by Jackson’s contracts” and that could be done on a class-wide basis: “For every class member, the answer will uniformly be either ‘yes’ or ‘no.’” Jackson argued that, to the contrary, determining breach would involve consideration of extrinsic evidence—for example, prospectuses accompanying the annuity contracts and individual brokers’ advice illustrating the surrender charge calculations—that must be assessed on a case-by-case basis according to varying state-law standards. The court rejected Jackson’s argument, reasoning that courts “routinely” certify class actions involving “form contracts” like Jackson’s, and that it was “hard to see what extrinsic evidence would be relevant to interpreting [such] unambiguous form contract[s].” Jackson renews its arguments on appeal. We agree with Jackson that the district court’s predominance inquiry failed to rigorously analyze the potential impact of state-law variations on the question of breach, as our precedents require. Where a proposed class stretches across multiple jurisdictions, we have underscored that a Rule 23(b)(3) analysis must carefully consider variations in pertinent state laws: [I]n a class action governed by the laws of multiple states, . . . “variations in state law may swamp any common issues and defeat predominance.” The party seeking certification of a nationwide class must therefore “provide an ‘extensive analysis’ of state law variations to reveal whether these pose ‘insuperable obstacles.’” And the district court must then “consider how [variations in state law] affect predominance.” Failure to engage in an analysis of state law variations is grounds for decertification. Cole v. Gen. Motors Corp., 484 F.3d 717, 724 (5th Cir. 2007) (citations omitted) (first quoting Castano, 84 F.3d at 741; then quoting Spence, 227 F.3d at 313; and then quoting Castano, 84 F.3d at 740). Here, the district court provided only a cursory analysis on this key point. That is reversible error. See, e.g., Castano, 84 F.3d at 742 (vacating predominance finding because court “conducted a cursory review of state law variations and gave short shrift to the defendants’ arguments concerning [state-law] variations”).[13] As noted, the district court’s decision not to delve into state-law variations rested on its conclusion that the central interpretive issue in the case could be resolved by parsing the contract language alone. The court thought it unlikely that extrinsic evidence was “relevant to interpreting unambiguous form contract terms.” Even if that is right, the court needed to conduct an “extensive analysis” of whether any state-law variations existed and, if they did, determine whether those variations defeated predominance. Cole, 484 F.3d at 724. Whether the district court’s conclusion is ultimately correct or not, its failure to conduct a thorough analysis was reversible error. We recognize that suits involving form contracts often lend themselves to class treatment. See, e.g., Sacred Heart Health Sys., Inc. v. Humana Military Healthcare Servs., Inc., 601 F.3d 1159, 1171 (11th Cir. 2010) (observing “[i]t is the form contract, executed under like conditions by all class members, that best facilitates class treatment”); Kolbe v. BAC Home Loans Servicing, LP, 738 F.3d 432, 441 (1st Cir. 2013) (explaining that class certification is generally proper where the outcome will not depend on extrinsic evidence “that would be different for each putative class member”); 2 NEWBERG § 3.24 (explaining that claims arising out of “form contracts” are often “particularly appropriate for class action treatment”). But this is not always so. Predominance in form contract cases may be defeated, for instance, “if individualized extrinsic evidence bears heavily on the interpretation of the class members’ agreements,” or if “there may be considerable variation in the state law under which any extrinsic evidence would have to be scrutinized.” Sacred Heart Health Sys., 601 F.3d at 1176-77, 1180. Because of this possibility, it is incumbent on the district court to undertake a “rigorous analysis” of the predominance requirements, “including the question of whether variations in state law will destroy predominance.” Id. at 1180; see also, e.g., Spence, 227 F.3d at 313 (noting “the court’s duty to determine whether the plaintiffs have borne their burden where a class will involve multiple jurisdictions and variations in state law”); cf., e.g., McKeage v. TMBC, LLC, 847 F.3d 992, 999 (8th Cir. 2017) (predominance requirement met in nationwide form contract class where, inter alia, all claims were governed by Missouri law). One such rigorous predominance analysis appears in Steinberg v. Nationwide Mutual Insurance Co., involving a 46-state putative class claiming breach of standard- form car insurance contracts. 224 F.R.D. 67, 76 (E.D.N.Y. 2004). Assisted by the plaintiffs’ extensive “50-state breach of contract analysis,” the district court concluded that states could be grouped into four categories based on their varying contract interpretation laws. Id. at 77-78. The court found that these state-law differences would be manageable in the class context or, alternatively, that sub-classes could be created based on the four categories. Id. at 78. This is the kind of “extensive analysis,” Cole, 484 F.3d at 724, of the potential impact of state-law variations on predominance that the district court should conduct on remand. Jackson next argues that predominance is also defeated by its affirmative defenses of waiver and ratification. Those defenses, Jackson urges, depend on individual class members’ knowledge of the facts surrounding formation of the annuity contracts, thus privileging individual issues over common ones. The district court disagreed, rejecting the salience of Jackson’s affirmative defenses to predominance. We agree with Jackson that the district court erred in assessing the impact of Jackson’s affirmative defenses on predominance. “We have noted that the ‘predominance of individual issues necessary to decide an affirmative defense may preclude class certification.’” Gene & Gene LLC v. BioPay LLC, 541 F.3d 318, 327 (5th Cir. 2008) (quoting In re Monumental Life Ins. Co., 365 F.3d 408, 420 (5th Cir. 2004)). Even in form contract cases, “[t]he risk of voluminous and individualized extrinsic proof [defeating predominance] runs particularly high where a defendant raises substantial affirmative defenses to breach.” Sacred Heart Health Sys., 601 F.3d at 1177. Here the pertinent defenses are waiver and ratification. The gravamen of Jackson’s waiver defense is that individual class members waived any objection to the calculation of surrender charges by knowingly accepting charges calculated according to Jackson’s formula. See, e.g., Jernigan v. Langley, 111 S.W.3d 153, 156 (Tex. 2003) (under Texas law, waiver is “an intentional relinquishment of a known right or intentional conduct inconsistent with claiming that right” (quoting Sun Expl. & Prod. Co. v. Benton, 728 S.W.2d 35, 37 (Tex. 1987))). Similarly, the gravamen of Jackson’s ratification defense is that individual class members ratified Jackson’s calculation of surrender charges by accepting those charges with full knowledge of the surrounding facts. See, e.g., Fortune Prod. Co. v. Conoco, Inc., 52 S.W.3d 671, 678 (Tex. 2000) (“Ratification is the adoption or confirmation by a person with knowledge of all material facts of a prior act which did not then legally bind him and which he had the right to repudiate.” (quoting Spellman v. Am. Universal Inv. Co., 687 S.W.2d 27, 29 (Tex. App.—Corpus Christi 1984, writ refd n.r.e.))). Jackson argues that assessing the validity of these affirmative defenses “will require individual inquiries into the knowledge and actions of each class member, overwhelming predominance.” The district court rejected this argument. It reasoned that these affirmative defenses would require class members to know not only the facts informing contract formation, such as the advice given by brokers or the contents of prospectuses annexed to the contracts, but also the “proper interpretation” of the surrender charge provisions and the “unlawfulness” of Jackson’s conduct. Because customers could not have been aware of this knowledge “until either liability is established or Jackson admits that its actions were unlawful,” the district court held that Jackson’s affirmative defenses could not defeat predominance. Jackson argues that the district court legally erred in requiring this kind of legal knowledge to trigger waiver or ratification. We agree. Generally speaking, the knowledge contemplated in the defenses of waiver or ratification is knowledge of the essential facts of a transaction, not the legal effects of those facts. See, e.g., RESTATEMENT (SECOND) OF CONTRACTS § 93 (1981) (making various promises non-binding unless promisor “knew or had reason to know the essential facts . . . but his knowledge of the legal effects of the facts is immaterial”); id. § 84 cmt. b (rejecting for waiver purposes “the incorrect inference that the promisor must know his legal rights,” while explaining “it is sufficient if he has reason to know the essential facts”); WILLISTON ON CONTRACTS § 39:22 (4th ed. 1999) (explaining “a waiver made with knowledge of facts which would put an ordinary person on inquiry is sufficient” but “it is not necessary that the party charged with a waiver be aware of the exact legal nature or scope of the right being relinquished or of the legal effect of the right at issue”). Based on these authorities, one well- reasoned Texas decision explains that “[p]roof regarding a party’s actual understanding of the legal consequences of those [essential] facts is not required” to establish waiver, because “parties are presumed to know and understand the legal effect of their contracts and waivers.” Trelltex, Inc. v. Intecx, LLC, 494 S.W.3d 781, 792 & n.10 (Tex. App.—Houston [14th Dist.] 2016, no pet.) (and collecting decisions). Taking this proper view of the knowledge requirement, a defense of waiver or ratification would be based on customers’ knowledge of facts, not their knowledge of the legal effect of those facts, such as the proper interpretation of the surrender charge provisions or the unlawfulness of Jackson’s conduct. The district court cited no authority supporting its view that Jackson’s affirmative defenses require class members’ knowledge of legal conclusions. The court cited In re Checking Account Overdraft Litigation, 307 F.R.D. 630, 650 (S.D. Fla. 2015), but that decision simply states the general rule that affirmative defenses like ratification and waiver “require a party to have had full knowledge of the circumstances in order for the defense to prevail.” In that case predominance was not defeated by those defenses, however, because the defendant had allegedly engaged in widespread concealment of its illegal scheme—meaning that plaintiffs could “undercut [its] defenses through the use of common evidence.” Id. at 651. No such allegations are present here. The court also cited Allen v. Holiday Universal, 249 F.R.D. 166, 174 (E.D. Pa. 2008), but that decision again recites the general rule that a “party charged with ratification must have acted . . . with full knowledge of all material facts,” while saying nothing to support the district court’s view about legal consequences.[14] In sum, the district court legally erred, and therefore abused its discretion, in assessing the potential impact of Jackson’s affirmative defenses on the predominance inquiry. C. Finally, Jackson argues that plaintiffs failed to offer a damages model adequate to support class treatment. Because plaintiffs only put forth a model for calculating one aspect of the damages alleged, we agree with Jackson. Even where plaintiffs seeking class certification show that common issues predominate on questions of liability, they must also present a damages model “establishing that damages are capable of measurement on a classwide basis.” Comcast Corp. v. Behrend, 569 U.S. 27, 34 (2013). The issue of damages may defeat predominance “where the calculation of damages is not susceptible to a mathematical or formulaic calculation.” Bell Atl., 339 F.3d at 307. Here, plaintiffs seek two kinds of damages linked to Jackson’s allegedly inflated surrender charges: (1) the excessive charges themselves and (2) the effect of those charges on living and death benefits paid under the annuity contracts. At the certification stage, plaintiffs offered an expert opinion that excessive charges could be uniformly calculated through “relevant mathematical calculations appropriately applied” to Jackson’s own records. But the report offered no formula for calculating damages related to living and death benefits. Instead, it asserted that “[t]he precise details of how the calculation of these additional damages will be performed will depend on what information Jackson makes available.” Without separately addressing living and death benefits, the district court concluded that damages could be uniformly calculated based on Jackson’s records and, therefore, would not defeat predominance. We do not question the district court’s conclusion regarding excessive charges. The record persuades us that those damages can be reckoned by fairly straightforward math from Jackson’s files. The fact that this will require individualized calculations does not defeat predominance. See, e.g., Ibe v. Jones, 836 F.3d 516, 529 (5th Cir. 2016) (“Generally, individualized damages calculations will not preclude a finding of predominance.” (citing Tyson Foods, 136 S. Ct. at 1045)). The more complex damages related to living and death benefits are another matter, however. The expert report plaintiffs relied on offered no indication how those damages would be calculated and, indeed, candidly admitted that it lacked any basis for an opinion. This was inadequate to meet plaintiffs’ burden to show that these kinds of damages are appropriate for class treatment. To be sure, proposed damages “[c]alculations need not be exact.” Comcast Corp., 569 U.S. at 35. But here plaintiffs offered barely a “preliminary overview of how [these] damages might be calculated.” Piggly Wiggly Clarksville, Inc. v. Interstate Brands Corp., 100 F. App’x 296, 299 (5th Cir. 2004). On remand, the district court should reassess whether the calculation of damages related to living and death benefits would defeat predominance. We suspect this issue may disappear on remand, however. At oral argument, plaintiffs’ counsel announced that plaintiffs would no longer seek damages based on living and death benefits, because calculating those damages turned out to be “complicated.” OA Audio 40:40-41:10. We decline to speculate on the effect of counsel’s concession, which is better left to the district court on remand. See, e.g., Slade v. Progressive Sec. Ins. Co., 856 F.3d 408, 412-15 (5th Cir. 2017) (remanding for consideration of how class representative’s waiver of claims impacts the adequacy requirement of Rule 23(a)(4)). IV. In sum, we VACATE the district court’s order certifying the class and REMAND for further proceedings consistent with this opinion. EDITH BROWN CLEMENT, Circuit Judge, concurring: I concur in the opinion and write separately to explain why our resolution of the personal-jurisdiction issue is appropriate even though Jackson does not directly address one of the district court’s rulings: that Jackson waived its personal-jurisdiction defense by substantially litigating on the merits. The appellate rules require an appellant to state in his brief his “contentions and the reasons for them, with citations to the authorities and parts of the record on which the appellant relies.” FED. R. APP. P. 28(a)(8)(A). Failure to argue an issue often means that the appellant has abandoned that issue. E.g., United States v. Scroggins, 599 F.3d 433, 446-47 (5th Cir. 2010). But not always. Our rule of deeming unargued issues abandoned is “a prudential construct that requires the exercise of discretion.” United States v. Miranda, 248 F.3d 434, 443 (5th Cir. 2001). I would reach the personal- jurisdiction issue through the exercise of our discretion. The district court held that Jackson waived its personal-jurisdiction defense “[b]y waiting exactly seven months after Bristol-Myers to informally raise the personal[-]jurisdiction objection in its response to class certification . . . and then proceed[ing] to substantially litigate on the merits, including filing its motion for summary judgment.” The district court cited two motions that Jackson filed. Those motions, plus Jackson’s summary-judgment motion, were what the district court apparently thought constituted substantial litigation on the merits. Jackson does not directly state why this was wrong, but it was plainly wrong. Jackson likely thought that the district court’s error was so obvious that it needn’t be explained—if class certification, rather than Bristol-Myers, is what made available Jackson’s personal-jurisdiction defense as to the putative class members, Jackson could not be faulted for filing motions after Bristol- Myers came down but before certification. That is true, but only to the extent that filing motions could itself constitute substantial litigation on the merits. Yet Jackson might have forfeited the defense if, in those motions, it had substantially litigated issues involving the putative class members. Jackson does not address that possibility, and it obviously did not litigate any such issues. In those motions, Jackson neither concedes personal jurisdiction as to the putative class members nor argues the merits of issues involving them. Thus, the district court clearly erred in holding that Jackson forfeited its personal-jurisdiction defense through its litigation conduct. Because this error was clear and obvious, I would exercise our discretion to reach this issue despite Jackson’s failure to address it head-on. With that caveat, I concur.
