{¶ 1} State Farm Mutual Automobile Insurance Company appeals from a judgment of the Eighth District Court of Appeals affirming class certification of claims brought by Michael Cullen alleging that State Farm failed to disclose all benefits available to policyholders who made claims for damaged windshields. This case clarifies the standards to apply when an appellate court reviews certification of a class action pursuant to Civ.R. 23.
{¶ 2} A class action is an exception to the general rule that litigation is conducted by and on behalf of the named parties only, and therefore, to justify a departure from this rule, the representative of the putative class is required .to affirmatively demonstrate that each requirement of Civ.R. 23 has been satisfied. When determining whether to certify a class, a trial court must conduct a rigorous analysis, and it may grant certification only after resolving all relevant factual disputes and finding that sufficient evidence proves that all requirements of Civ.R. 23 have been satisfied.
{¶ 3} Here, the appellate court affirmed certification of the class pursuant to Civ.R. 23(B)(2) and (3). However, because the declaratory relief at issue here is incidental to an individualized claim for monetary damages, Cullen has not met the requirement for certification set forth in Civ.R. 23(B)(2). In addition, because individual questions predominate over the questions common to the proposed class, Cullen has not proven that this action satisfies Civ.R. 23(B)(3). Accordingly, we reverse the judgment of the court of appeals and remand the matter to the trial court for further proceedings consistent with this opinion.
Facts and Procedural History
{¶ 4} In 2003, Michael Cullen contacted his automobile insurance carrier, State Farm, to report damage to the windshield of his car. After speaking with his agent, Cullen spoke to a representative from Lynx Services, L.L.C., a company that began handling windshield claims for State Farm in 1996. As a result of that conversation, Twinsburg Glass & Mirror repaired his windshield.
{¶ 5} In 2005, Cullen sued State Farm, requesting class certification and a declaratory judgment that State Farm’s practices were illegal and violated obligations owed by fiduciaries pursuant to Ohio law. In addition, Cullen asserted claims for breach of contract, bad faith, and breach of fiduciary duty and sought compensatory and punitive damages. The complaint defined the class to include all State Farm policyholders on or after February 18, 1990, and alleged that State Farm had denied them full payment on windshield claims because, instead of replacing windshields, it repaired some windshields with a chemical compound that it knew or should have known was “only temporary, not entirely translucent, and incapable of restoring the windshield to its preaccident condi*375tion” and that State Farm was not “paying the insured to replace the glass,” less any deductible. As a result of State Farm’s alleged breach, Cullen and the class demanded “damages in an amount to be determined at trial under principles of Ohio common law” or that State Farm be ordered to “tender benefits sufficient to replace the windshields in accordance with policy requirements.”
{¶ 6} The trial court held a hearing on the certification issue. Although Cullen could not remember speaking to a representative of..Lynx or whether he had been offered the choice of replacing his windshield or receiving the cash value of a new windshield, less his deductible, he asserted that State Farm agents and representatives failed to disclose to policyholders making “glass-only” claims that the policies contained a benefit referred to as the “cash-out” option. According to Cullen, although the State Farm policy promised its insureds the option of receiving a cash payment of the replacement cost of the windshield, less any deductible, it prepared a script for representatives to induce policyholders to repair their windshields without disclosing the cash-out option. State Farm referred to the outline it provided to its agents and representatives handling glass-only claims as a “word track,” arguing that it gave them discretion to respond to questions asked by insureds.
{¶ 7} The trial court concluded that Cullen and the class satisfied the requirements of Civ.R. 23(A) and (B)(2) and (3), specifically finding that the class satisfied Civ.R. 23(B)(2) because “it appeared] that the same practices which the Named Plaintiff experienced [were] still ongoing” and it further determined that a declaratory judgment and injunctive relief were “potential available remedies which can be issued on a class wide [sic] basis in the event that [Cullen] prevails upon the merits of his claim.” In finding that Cullen and the class met the requirements for certification pursuant to Civ.R. 23(B)(3), the trial court explained:
The Court is sufficiently convinced that Plaintiffs’ claims for relief are founded squarely upon standardized policies and practices which had been adopted and employed by State Farm throughout Ohio on a systematic basis during the Class Period. Given that the maximum individual recoveries will be relatively modest, separate lawsuits are not realistic. And it is doubtful that the Ohio judicial system could afford full and fair relief to thousands of aggrieved insureds on a case-by-case basis. A class action is thus the most preferable and superior method for adjudicating the common questions of law and fact, which the Court concludes, predominate over any individual questions which may exist.
{¶ 8} The trial court defined the class as follows:
*376All persons and business entities covered under an Ohio motor vehicle insurance policy issued by [State Farm] who made a “Glass Only” physical damage comprehensive coverage claim on or after January 1, 1991 for cracked, chipped or damaged windshields and received a chemical filler or patch repair, or payment thereof, instead of a higher amount for actual cash value or replacement cost of the windshield. The lesser of the amount of the actual cash value or the replacement cost of the windshield for each claim must exceed the insured’s applicable deductible.
The trial court further divided the class into two subclasses: those insureds who had their claims administered by Lynx and those who did not.
{¶ 9} State Farm appealed, and the Eighth District Court of Appeals affirmed the order certifying the class pursuant to Civ.R. 23(B)(2) and (3), but it reversed that portion of the decision defining the class and remanded with instructions to the trial court to redefine it. The court of appeals explained:
For claims handled using a common script or word track, the trial court did not err in certifying the class in this case. Individual questions do not predominate because the script used by Lynx and developed by State Farm establishes class-wide treatment under Cullen’s theory that State Farm breached its contracts with insureds by dissuading individuals from replacing their windshields and not informing them of their option to receive a check for the value of the windshield less their deductible. For claims made prior to the use of a common script, Cullen argues that the policy language simplifies the case to a showing that the policy in question required State Farm to restore vehicles to their preloss condition and that a windshield repair cannot do so. The theory, while dubious, does provide a means to resolve the case on a class-wide basis for these members. Therefore, the trial court did not err in certifying this class. However, the class definition must be restricted to exclude those who had their windshields replaced after repair.
2011-Ohio-6621, 970 N.E.2d 1043, at ¶ 56.
{¶ 10} We accepted State Farm’s discretionary appeal regarding the class certification and the standards that apply to a review of an order certifying a class pursuant to Civ.R. 23(B)(2) and (B)(3).
*377Class Actions
{¶ 11} A class action is “an exception to the usual rule that litigation is conducted by and on behalf of the individual named parties only,” Califano v. Yamasaki, 442 U.S. 682, 700-701, 99 S.Ct. 2545, 61 L.Ed.2d 176 (1979), and “[t]o come within the exception, a party seeking to maintain a class action ‘must affirmatively demonstrate his compliance’ with Rule 23,” Comcast Corp. v. Behrend, — U.S. -, 133 S.Ct. 1426, 1432, 185 L.Ed.2d 515 (2013), quoting Wal-Mart Stores, Inc. v. Dukes, 564 U.S.-, 131 S.Ct. 2541, 2551-2552, 180 L.Ed.2d 374 (2011).
{¶ 12} Civ.R. 23 provides seven requirements for maintaining a class action:
“(1) an identifiable class must exist and the definition of the class must be unambiguous; (2) the named representatives must be members of the class; (3) the class must be so numerous that joinder of all members is impracticable; (4) there must be questions of law or fact common to the class; (5) the claims or defenses of the representative parties must be typical of the claims or defenses of the class; (6) the representative parties must fairly and adequately protect the interests of the class; and (7) one of the three Civ.R. 23(B) requirements must be met.”
Stammco, L.L.C. v. United Tel. Co. of Ohio, 125 Ohio St.3d 91, 2010-Ohio-1042, 926 N.E.2d 292, at ¶ 6, quoting Hamilton v. Ohio Sav. Bank, 82 Ohio St.3d 67, 71, 694 N.E.2d 442 (1998), citing Civ.R. 23(A) and (B) and Warner v. Waste Mgt., 36 Ohio St.3d 91, 521 N.E.2d 1091 (1988).
{¶ 13} This appeal does not challenge the determination of the trial court that Cullen and the class met the requirements of Civ.R. 23(A), but rather focuses on whether the class met the requirements of Civ.R. 23(B)(2) and (3), which allow a class action in either of the following situations:
(2) the party opposing the class has acted or refused to act on grounds generally applicable to the class, thereby making appropriate final injunctive relief or corresponding declaratory relief with respect to the class as a whole; or
(3) the court finds that the questions of law or fact common to the members of the class predominate over any questions affecting only individual members, and that a class action is superior to other available methods for the fair and efficient adjudication of the controversy. The matters pertinent to the findings include: (a) the interest of members of the class in individually controlling the prosecution or defense of separate *378actions; (b) the extent and nature of any litigation concerning the controversy already commenced by or against members of the class; (c) the desirability or undesirability of concentrating the litigation of the claims in the particular forum; (d) the difficulties likely to be encountered in the management of a class action.
{¶ 14} Because Civ.R. 23 is virtually identical to Fed.R.Civ.P. 23, we have recognized that “federal authority is an appropriate aid to interpretation of the Ohio rule.” Marks v. C.P. Chem. Co., Inc., 31 Ohio St.3d 200, 201, 509 N.E.2d 1249 (1987). In Marks, we reviewed certification of a class action against the manufacturer, distributor, and installers of foamed-in-place insulation that allegedly emitted dangerous levels of formaldehyde. In the complaint, the class representative sought injunctive and declaratory relief ordering future diagnostic testing for every class member. The trial court denied certification pursuant to Civ.R. 23(B)(2) and (B)(3), and the court of appeals reversed. We reinstated the trial court’s decision that Civ.R. 23(B)(2) did not permit class certification, relying on decisions of federal circuit courts construing Fed.R.Civ.P. 23(b)(2) and holding that “this provision is inapplicable where the primary relief requested is damages.” Marks at 203, citing Eisen v. Carlisle & Jacquelin, 391 F.2d 555 (2d Cir.1968), vacated on other grounds, 417 U.S. 156, 94 S.Ct. 2140, 40 L.Ed.2d 732 (1974), and School Dist. of Lancaster v. Lake Asbestos of Quebec, Ltd., 789 F.2d 996 (3d Cir.1986). We also reversed the appellate court’s determination that the class action could proceed pursuant to Civ.R. 23(B)(3), because individual questions — such as the products used, the manner of their storage and installation, and defects in the structures where installed, which affected the emission of formaldehyde — predominated over common ones.
The Rigorous Analysis of the Trial Court
{¶ 15} In Comcast, 133 S.Ct. at 1432, 185 L.Ed.2d 515, quoting Wal-Mart, 131 S.Ct. at 2551-2552, 180 L.Ed.2d 374, the court stated that Civ.R. 23 “ ‘does not set forth a mere pleading standard.’ * * * Rather, a party must not only ‘be prepared to prove that there are in fact sufficiently numerous parties, common questions of law or fact,’ typicality of claims or defenses, and adequacy of representation, as required by Rule 23(a). * * * The party must also satisfy through evidentiary proof at least one of the provisions of Rule 23(b).” And this court has held that a party seeking certification pursuant to Civ.R. 23 bears the burden of demonstrating by a preponderance of the evidence that the proposed class meets each of the requirements set forth in the rule. Warner v. Waste Mgt., 36 Ohio St.3d at 94, 521 N.E.2d 1091; see also Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 614, 117 S.Ct. 2231, 138 L.Ed.2d 689 (1997).
*379{¶ 16} Moreover, a trial court must conduct a rigorous analysis when determining whether to certify a class pursuant to Civ.R. 23 and may grant certification only after finding that all of the requirements of the rule are satisfied; the analysis requires the court to resolve factual disputes relative to each requirement and to find, based upon those determinations, other relevant facts, and the applicable legal standard, that the requirement is met. State ex rel. Davis v. Pub. Emps. Retirement Bd., 111 Ohio St.3d 118, 2006-Ohio-5339, 855 N.E.2d 444, at ¶ 20, citing Hamilton v. Ohio Sav. Bank, 82 Ohio St.3d at 70, 694 N.E.2d 442; see also Comcast, 133 S.Ct. at 1432, 185 L.Ed.2d 515, Wal-Mart, 131 S.Ct. at 2551, 180 L.Ed.2d 374, and Initial Pub. Offering Sec. Litigation, 471 F.3d 24, 40-41 (2d Cir.2006) (“In re IPO ”).
{¶ 17} In Ojalvo v. Bd. of Trustees of Ohio State Univ., 12 Ohio St.3d 230, 233, 466 N.E.2d 875 (1984), we stated, “Class action certification does not go to the merits of the action.” (Emphasis sic.) However, deciding whether a claimant meets the burden for class certification pursuant to Civ.R. 23 requires the court to consider what will have to be proved at trial and whether those matters can be presented by common proof. 7AA Wright, Miller & Kane, Federal Practice and Procedure, Section 1785 (3d Ed.2005). Thus, Wal-Mart instructs that in resolving a factual dispute when a requirement of Civ.R. 23 for class certification and a merit issue overlap, a trial court is permitted to examine the underlying merits of the claim as part of its rigorous analysis, but only to the extent necessary to determine whether the requirement of the rule is satisfied. Wal-Mart at 2551-2552; Ellis v. Costco Wholesale Corp., 657 F.3d 970, 981 (9th Cir.2011).
{¶ 18} In Comcast, the Supreme Court reiterated that “it ‘may be necessary for the court to probe behind the pleadings before coming to rest on the certification question,’ and that certification is proper only if ‘the trial court is satisfied, after a rigorous analysis, that the prerequisites of Rule 23(a) have been satisfied.’ ” Comcast, 133 S.Ct. at 1432, 185 L.Ed.2d 515, quoting Wal-Mart, 131 S.Ct. at 2541, quoting Gen. Tel. Co. of the Southwest v. Falcon, 457 U.S. 147, 160-161, 102 S.Ct. 2364, 72 L.Ed.2d 740 (1982). The court noted that “[t]he same analytical principles govern Rule 23(b)” and explained that “[s]uch an analysis will frequently entail ‘overlap’ with the merits of the plaintiffs underlying claim” because “ ‘ “class determination generally involves considerations that are enmeshed in the factual and legal issues comprising the plaintiffs cause of action.” ’ ” Comcast at 1432, quoting Wal-Mart at 2551, quoting Gen. Tel. Co. of the Southwest at 160.
Standard of Review
{¶ 19} “A trial judge has broad discretion in determining whether a class action may be maintained and that determination will not be disturbed absent a showing of an abuse of discretion.” Marks, 31 Ohio St.3d 200, 509 N.E.2d 1249, syllabus; *380see also Howland v. Purdue Pharma L.P., 104 Ohio St.3d 584, 2004-Ohio-6552, 821 N.E.2d 141, at ¶ 17. A trial court, however, abuses its discretion when its decision is “unreasonable, arbitrary, or unconscionable.” Wilson v. Brush Wellman, Inc., 103 Ohio St.3d 538, 2004-Ohio-5847, 817 N.E.2d 59, at ¶ 30. This standard applies to the ultimate decision of the trial court, Marks, syllabus, as well as to its determination regarding each requirement of the rule. See In re IPO, 471 F.3d at 40-41.
{¶ 20} However, as we clarified in Eastley v. Volkman, 132 Ohio St.3d 328, 2012-Ohio-2179, 972 N.E.2d 517, “[i]n a civil case, in which the burden of persuasion is only by a preponderance of the evidence, rather than beyond a reasonable doubt, evidence must still exist on each element (sufficiency) and the evidence on each element must satisfy the burden of persuasion (weight).” Id. at ¶ 19.
Certification of the Class
Civ.R. 23(B)(2)
{¶ 21} Civ.R. 23(B)(2) provides that class treatment may be maintained when “the party opposing the class has acted or refused to act on grounds generally applicable to the class, thereby making appropriate final injunctive relief or corresponding declaratory relief with respect to the class as a whole.” Claims for individualized relief are not compatible with Civ.R. 23(B)(2), because the relief sought must affect the entire class at once. Wal-Mart, 131 S.Ct. at 2557-2558, 180 L.Ed.2d 374. Thus, the Supreme Court clarified in Wal-Mart that Fed. R.Civ.P. 23(b)(2)
applies only when a single injunction or declaratory judgment would provide relief to each member of the class. It does not authorize class certification when each individual class member would be entitled to a different injunction or declaratory judgment against the defendant. Similarly, it does not authorize class certification when each class member would be entitled to an individualized award of monetary damages.
(Emphasis sic.) 131 S.Ct. at 2557.
{¶ 22} For these reasons, certification depends on “what type of relief is primarily sought, so where the injunctive relief is merely incidental to the primary claim for money damages, Civ.R. 23(B)(2) certification is inappropriate.” Wilson v. Brush Wellman, Inc., 103 Ohio St.3d 538, 2004-Ohio-5847, 817 N.E.2d 59. at ¶ 17.
*381{¶ 23} In the instant case, the court of appeals appears to have relied on our decision in Hamilton v. Ohio Sav. Bank, 82 Ohio St.3d 67, 694 N.E.2d 442, which it viewed as holding that when the Civ.R. 23(A) prerequisites have been met and injunctive or declaratory relief has been requested, the action should be allowed to proceed pursuant to subdivision (B)(2), without the need for conducting a rigorous analysis. However, Hamilton does not allow a court to dispense with the more rigorous analysis of whether a class should be certified. Instead, Hamilton directs that “[t]he trial court is required to carefully apply the class action requirements and conduct a rigorous analysis into whether the prerequisites of Civ.R. 23 have been satisfied.” Id. at 70. We did not create an exception to this requirement for actions brought pursuant to Civ.R. 23(B)(2), but rather we emphasized that certification of a class is not permitted by Civ.R. 23(B)(2) when the primary relief sought is damages, not injunctive or declaratory relief.
{¶ 24} Furthermore, Hamilton is distinguishable from this case on its facts. There, the class sought to enjoin the practice of overcharging interest and misamortizing loans. We concluded that without injunctive or declaratory relief, the class would not be able to recover for ongoing injuries caused to each class member by continuing practices. In contrast, the proposed Cullen class seeks a declaration “establishing that State Farm’s practices as herein described are illegal and/or violative of the terms of the standard policies and the obligations owed by fiduciaries under Ohio law,” as well as one “establishing the damages and remedies that are due to them.” This does not allege that any ongoing practice continues to injure all class members, some of whom, like Cullen himself, are no longer State Farm policyholders and could not be injured by future actions taken by State Farm. And for any current policyholders to be harmed by this practice, they necessarily would have to suffer another damaged windshield that State Farm repaired rather than replaced.
{¶ 25} As the Supreme Court explained in Wal-Mart, “The key to the (b)(2) class is ‘the indivisible nature of the injunctive or declaratory remedy warranted — the notion that the conduct is such that it can be enjoined or declared unlawful only as to all of the class members or as to none of them.’ ” (Emphasis added.) 131 S.Ct. at 2557, 180 L.Ed.2d 374, quoting Nagareda, Class Certification in the Age of Aggregate Proof, 84 N.Y.U.L.Rev. 97, 132 (2009). In this case, claimants have not demonstrated that all class members would benefit from the declaratory relief sought because, for example, some of the class members are not currently policyholders.
{¶ 26} The appellate court also relied on Allison v. Citgo Petroleum Corp., 151 F.3d 402 (5th Cir.1998), for the proposition that certification of a class pursuant to Civ.R. 23(B)(2) is permissible when “the relief sought includes money damages *382that require individualized analyses as to the proper amount, but that relief flows from the declaratory judgment sought.” Cullen, 2011-Ohio-6621, 970 N.E.2d 1043, at ¶ 50. But in Wal-Mart, the court questioned the Fifth Circuit’s conclusion that the federal analogue to Civ.R. 23(B)(2) permits certification when the monetary damages sought are “incidental to requested injunctive or declaratory relief.” 131 S.Ct. at 2560.
{¶ 27} Even if Civ.R. 23(B)(2) did allow certification when the monetary damages are only incidental to the declaratory relief, the damages in this case are not merely incidental to the declaratory relief but, rather, are the primary relief sought. The effect of a declaration on members of the proposed class could establish liability, thereby allowing an individualized award of monetary damages to each class member. However, claims for declaratory relief that merely lay a foundation for subsequent determinations regarding liability or that facilitate an award of damages do not meet the requirement for certification as set forth in Civ.R. 23(B)(2). Kartman v. State Farm Mut. Auto. Ins. Co., 634 F.3d 883, 893 (7th Cir.2011) (“this case is unsuitable for class certification: An injunction would not provide ‘final’ relief as required by Rule 23(b)(2). An injunction is not a final remedy if it would merely lay an evidentiary foundation for subsequent determinations of liability”); Richards v. Delta Air Lines, Inc., 453 F.3d 525, 530 (D.C. Cir.2006).
{¶ 28} Accordingly, Cullen’s action seeking a declaration that State Farm’s practices are illegal and violated fiduciary obligations merely lays a foundation for a subsequent individual determination of liability and does not satisfy the requirements for class certification pursuant to Civ.R. 23(B)(2).
Civ.R. 23(B)(3)
{¶ 29} Certification pursuant to Civ.R. 23(B)(3) requires the trial court to make two findings: first, “that the questions of law or fact common to the members of the class predominate over any questions affecting only individual members” and, second, “that a class action is superior to other available methods for the fair and efficient adjudication of the controversy.” This inquiry requires a court to balance questions common among class members with any dissimilarities between them, and if the court is satisfied that common questions predominate, it then should “consider whether any alternative methods exist for resolving the controversy and whether the class action method is in fact superior.” Ealy v. Pinkerton Govt. Servs., Inc., 4th Cir. No. 12-1252, 2013 WL 980035, *7 (Mar. 14, 2013).
{¶ 30} “For common questions of law or fact to predominate, it is not sufficient that such questions merely exist; rather, they must present a significant aspect of the case. Furthermore, they must be capable of resolution for all members in a single adjudication.” Marks, 31 Ohio St.3d at 204, 509 N.E.2d 1249. “ ‘To meet *383the predominance requirement, a plaintiff must establish that issues subject to generalized proof and applicable to the class as a whole predominate over those issues that are subject to only individualized proof.’ ” Young v. Nationwide Mut. Ins. Co., 693 F.3d 532, 544 (6th Cir.2012), quoting Randleman v. Fid. Natl. Title Ins. Co., 646 F.3d 347, 352-353 (6th Cir.2011), citing Beattie v. CenturyTel, Inc., 511 F.3d 554, 564 (6th Cir.2007).
{¶ 31} Rather than determining that Cullen proved compliance with Civ.R. 23(B)(3) by a preponderance of the evidence, the appellate court hypothesized that “if Cullen’s theory of the case is believed, the use of a common plan to steer claimants to opt for repair rather than replacement or disclosure of a cash payment for the value of the glass, less deductible, is a significant class-wide issue.” (Emphasis added.) 2011-Ohio-6621, 970 N.E.2d 1043, at ¶ 21. Thus, in reviewing the question of whether State Farm violated a duty to fully disclose all pertinent benefits and coverages pursuant to Ohio Adm.Code 3901-1-54(E)(1), the court did not evaluate the evidence presented to the trial court but noted only that “Cullen argues that this was not done.” (Emphasis added.) Id. at ¶ 22. And although the appellate court recognized that these issues were “hotly contested by the parties,” it concluded that “none of these issues need be decided at this time because class certification is not akin to a motion for summary judgment.” Id. at ¶ 24.
{¶ 32} Similarly, when deciding whether the pre-1997 policies obligated State Farm to restore the claimant’s vehicle to preloss condition and whether expert testimony could establish that repair of a windshield will never return it to preloss condition, the appellate court declined to consider the merits of the claim — even though the court described this theory of recovery as “dubious” — but presumed that this theory “provide[s] a means to resolve the case on a class-wide basis for these members.” Id. at ¶ 56.
{¶ 33} Most importantly, the crux of Cullen’s class action is that the insurance policies at issue provided policyholders with the option of a cash payment in lieu of repair or replacement of the windshield, but the appellate court did not determine whether the policies provided any such benefit, deciding that “[t]his goes to the heart of the merits of the case and is inappropriate at this point. Class certification does not address the merits of the claim.” 2011-Ohio-6621, 970 N.E.2d 1043, at ¶ 55. The appellate court further noted, “The trial court examined these issues and determined that Cullen has raised a colorable claim sufficient to satisfy the Civ.R. 23 standards. That was not an abuse of discretion.” (Emphasis added.) Id. at ¶ 26.
{¶ 34} A colorable claim does not satisfy the requirements of Civ.R. 23. Nor can compliance with the rule be presumed from allegations in a complaint. Rather, in this instance, Cullen had to demonstrate, and the trial court had to *384find, that questions common to the class in fact predominate over individual ones, and proof of predominance necessarily overlaps with proof of the merits in this case. Because the appellate court rejected any consideration of the underlying merits, it did not review whether the trial court conducted a rigorous analysis or whether sufficient evidence supported the trial court’s findings.
{¶ 35} Further, the appellate court did not even consider whether any of State Farm’s defenses to liability raise individualized issues not susceptible of generalized proof. For example, State Farm asserts that it is not liable if an individual class member knowingly chose windshield repair — but individual consent and knowledge cannot be proven with common evidence. And if a windshield repair could return a vehicle to preloss condition — a question that neither the trial nor the appellate court resolved — State Farm’s liability would be subject to individual examinations of each vehicle, not common questions.
{¶ 36} Rather than remand this matter to the court of appeals to consider these issues, our review of the record reveals that individual issues overwhelm the questions common to the class, and the trial court therefore abused its discretion in certifying the class action.
{¶ 37} First, this putative class action is premised on State Farm’s alleged failure to disclose policy benefits through Lynx by using a script to steer insureds toward windshield repair. However, the class includes claims made beginning in 1991, even though Lynx began administering claims in 1997, and prior to that time individual State Farm agents handled windshield claims. But even after Lynx began to administer claims, policyholders had various individual, unscripted conversations with Lynx representatives, insurance agents, and repair-shop personnel, and there is no common proof of what any individual policyholder knew when consenting to windshield repair. Determining whether State Farm breached any obligations to insureds necessarily entails an individualized inquiry into each of these communications.
{¶ 38} Second, different versions of the policy covered putative class members on claims that span a period greater than 20 years. The policy in effect from 1991 to March 31, 1998, stated that State Farm would “pay for loss” to the policyholder’s vehicle and that it had the right to settle the loss by paying the actual cash value of the property at the time of the loss, as determined by agreement or appraisal. State Farm could also pay to repair or replace the damaged property, and the policy limited its liability to the lower of the actual cash value or the costs to repair or replace the property. But even if, as Cullen contends, this language allowed class members to elect to receive the actual cash value of the windshield, the policy provides that “[ajctual cash value is determined by the market value, age and condition at the time the loss occurred. Any deductible amount that applies is then subtracted.” Significant individual ques*385tions therefore exist concerning the market value, age, and condition of a particular windshield, the costs of repairing or replacing it, and the amount of the deductible in establishing State Farm’s liability to any given class member.
{¶ 39} Policies issued between April 1,1998, and August 31, 2005, contained the same language as the prior policies, but they also stated: “If we offer to pay for repair of the damaged windshield glass instead of replacement of the windshield and you agree to have such repair made, we will pay the full cost of repairing the windshield regardless of your deductible.” Thus, in addition to the individual questions regarding the actual cash value of the windshield compared to the costs of its repair or replacement, this version of the policy introduces new individualized questions concerning whether the policyholder knowingly chose and consented to repair in exchange for State Farm’s waiver of the deductible.
{¶ 40} State Farm later removed the deductible waiver for windshield repairs from subsequent policies, but individual questions regarding the actual cash value of a particular windshield and the costs to repair or replace it remain under this third variation of the policy.
{¶ 41} Third, the covered automobile, and therefore the value of the windshield compared to the cost of repairing or replacing it, varied. Multiple replacement windshields were available to glass shops over the decades-long period that the proposed class spans, and the costs of these windshields and the materials and labor needed to install them differed. State Farm adjusted the price it would pay based on the “market designation” assigned to each county, and these designations changed over time. The features of the covered windshield, such as tinting and rain sensors, also affected value and replacement cost. Thus, the costs to repair or replace a particular windshield varied by make, model, and year of the covered vehicle and by time and place of repair.
{¶ 42} Fourth, expert testimony presented in this case does not provide common proof that repairs failed to return all windshields to preloss condition. Cullen proffered the opinion of Craig Carmody that “[w]indshield repair fails to restore the windshield glass to an acceptable condition in terms of appearance and functionality in all cases.” But Carmody claimed to have examined only 17 to 22 repaired windshields, and it does not appear that his theories on glass repair have been thoroughly tested, peer-reviewed, evaluated for rate of error (including sampling error), or generally accepted in the scientific community. See generally Terry v. Caputo, 115 Ohio St.3d 351, 2007-Ohio-5023, 875 N.E.2d 72, ¶ 24-25, citing Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993). However, even if the trial court found Carmody qualified as an expert witness and his theories scientifically reliable, his opinion that a repair fails to restore the windshield to preloss condition does not establish that common questions predominate.
*386{¶ 43} Although Carmody testified that a repaired windshield will eventually fail, he could not say whether that failure would occur during the life of the vehicle or even during a ten-year period. Although he asserted that repaired windshields were more prone to failure, Carmody agreed that “you’d have to look at each of those windshields” to determine whether the PVB interlayer of laminated glass had been compromised and affected the function of the glass, causing it to fail. He also knew of no report of or statistics regarding windshield repairs causing delamination and spalling of glass, and he admitted that there is “no strong data” showing that the strength of repaired glass does not equal the original strength of laminated glass; moreover, he understood that one study found Ultra Bond resin to be as strong as glass and that the type of resin used in the repair “has some effect on the quality of the repair in addition to the other variables.”
{¶ 44} Thus, Carmody acknowledged “the huge variation that occurs in any repair, even in controlled conditions.” And notably, in examining Cullen’s windshield with a microscope, Carmody could not determine that the repair had failed or that the crack had spread, and his assertions that a repaired windshield could delaminate and spall did not apply to Cullen’s windshield, because the one-tenth-inch chip in it had not penetrated the PVB interlayer.
{¶ 45} Carmody also criticized a repaired windshield as not being as transparent as new glass. But he acknowledged that the transparency of repaired glass could “depend on the repair” and could be affected by real-world conditions, such as windshield-wiper abrasion and exposure to UV light and chemicals in the air. He also agreed that repair can restore transparency to windshield damage, and he did not know whether tests had shown that repairs can restore windshields to the transparency required by safety standards.
{¶ 46} Cullen’s other proffered expert, Gary Derian, admitted that he had no expertise in windshield repair and relied on Carmody’s glass analysis in forming his opinion that repairing a windshield failed to restore it to preloss condition, both aesthetically and from a safety standpoint. But he acknowledged that determining whether the resin in a repaired windshield had degraded and failed required individually examining that windshield and that performance varied based on the type of resin used and conditions such as “temperature, humidity, cleanliness, preparation.” When asked “whether or not the color of a resin changes over time,” Derian indicated that a change could not be assumed: “One would want to inspect and determine how much in cases of that, that particular [resin].” And when asked how he would decide whether a repaired windshield violated established standards, Derian responded, “[I]t would have to be done on a car by car basis with an automobile engineer present at each repair.”
*387{¶ 47} Both Carmody and Derian asserted that the repair could not restore a windshield to preloss condition, but neither had sufficient evidentiary foundation for those opinions. Derian relied on Carmody’s report, and Carmody in turn based his opinion on an unscientific sample of at most 22 vehicles in addition to a single study questioning whether resin should be used to repair longer cracks. Neither Carmody nor Derian performed any tests on repaired windshields to quantify the strength, durability, and transparency of repairs, and neither knew of any tests or studies that proved their opinions and methods.
{¶ 48} Thus, this expert testimony raises more individual questions than it resolves, and deciding whether State Farm breached any duty to restore policyholders’ windshields to preloss condition will require an individual inspection of each class member’s windshield to determine the preloss and postrepair conditions, and these individualized issues necessarily predominate over any questions common to the class.
{¶ 49} Our analysis conforms with Avery v. State Farm Mut. Auto. Ins. Co., 216 Ill.2d 100, 138, 296 Ill.Dec. 448, 835 N.E.2d 801 (2005), where the Supreme Court of Illinois considered a claim for breach of contract that depended on restoring an automobile to its preloss condition and held that “[a] necessary first step in making this showing would be to examine each class member’s vehicle to determine its preloss condition.” It further explained that “the determination of the preloss condition of each subclass member’s vehicle would require the individual examination of hundreds of thousands, if not millions, of vehicles. Undoubtedly, these examinations would overwhelm any question common to the subclass, rendering it impossible for such questions to predominate.” Id. For this reason, the court concluded, “a claim for breach of the preloss condition promise cannot be maintained as a class action.” Id.
{¶ 50} In sum, the determination of preloss and postrepair condition, the preloss value and the costs to repair or replace a particular windshield, and the individual knowledge and consent of each class claimant entail inspection of tens of thousands of automobiles and an individualized assessment of the damages each class member sustained, if any. For these reasons, this action does not satisfy the predominance requirement of Civ.R. 23(B)(3).
Conclusion
{¶ 51} Review of the certification of a class action requires the appellate court to determine whether the trial court conducted a rigorous analysis that resolved all relevant factual disputes and found by a preponderance of the evidence that the requirements of Civ.R. 23 have been satisfied. In making this determination, some consideration of the underlying merits of the cause of action may be necessary.
*388{¶ 52} Here, this action does not satisfy the requirements for class certification pursuant to Civ.R. 23(B)(2), because the declaratory relief sought is at best only incidental to an award of monetary damages, and the trial court abused its discretion in granting class certification pursuant to Civ.R. 23(B)(3), because a rigorous analysis of the evidence presented by the parties demonstrates that individual questions predominate over issues common to the class.
{¶ 53} Accordingly, we reverse the judgment of the appellate court and remand this matter to the trial court for further proceedings consistent with this opinion.
Judgment reversed and cause remanded.
O’Connor, C.J., and Kennedy and McFarland, JJ., concur. Lanzinger, J., concurs in paragraph three of the syllabus and in the judgment. Pfeifer and O’Neill, JJ., dissent. Matthew W. McFarland, J., of the Fourth Appellate District, sitting for French, J.