KETHLEDGE, J., delivered the opinion of the court, in which THAPAR, D. J., joined. COLE, J. (pp. 723-24), delivered a separate dissenting opinion.
OPINION
KETHLEDGE, Circuit Judge.Class-action settlements are different from other settlements. The parties to an ordinary settlement bargain away only their own rights — which is why ordinary settlements do not require court approval. In contrast, class-action settlements affect not only the interests of the parties and counsel who negotiate them, but also the interests of unnamed class members who by definition are not present during the negotiations. And thus there is always the danger that the parties and counsel will bargain away the interests of unnamed class members in order to maximize their own.
This case illustrates these dangers. The class is made up of consumers who purchased certain kinds of Pampers diapers between August 2008 and October 2011. The parties and their counsel negotiated a settlement that awards each of the named plaintiffs $1000 per “affected child,” awards class counsel $2.73 million, and provides the unnamed class members with nothing but nearly worthless injunctive relief. The district court found that the settlement was fair and certified the settlement class. We disagree on both points, and reverse.
I.
The defendant Procter & Gamble Company (“P & G”) manufactures Pampers brand diapers. In March 2010, P & G began marketing Pampers with so-called “Dry Max technology.” Two months later, the Consumer Product Safety Commission began investigating whether Dry Max diapers tend to cause severe diaper rash. The Clark lawsuit and 11 others were filed very soon thereafter. The district court consolidated all 12 cases.
In August 2010, the Commission — along with a Canadian agency, Health Canada— released the results of its investigation. Based upon a review of 4,700 incident reports, the Commission found no connection *716between the use of Dry Max diapers and diaper rash.
P & G then filed a motion to dismiss. Before Plaintiffs responded to the motion, however — and indeed before any formal discovery in the case — the parties began discussing settlement. By March 2011 they had reached a deal. Its essential terms were as follows: Although the plaintiffs had sought class certification under Federal Rule of Civil Procedure 28(b)(1) and (b)(3) — under which class members are free to opt out of the class, and thus out of the deal negotiated for them by class counsel — the parties agreed to seek certification of a class under Rule 23(b)(2), under which absent class members cannot opt out of the deal. The class was defined as follows:
All persons in the United States and its possessions and territories, who purchased or acquired (including by gift) Pampers brand diapers containing “Dry Max Technology” from August 2008 through Final Judgment. All federal judges to whom this case is assigned and members of their families within the first degree of consanguinity, and officers and directors of Procter & Gamble, are excluded from the class definition.
P & G agreed to reinstate, for one year, a refund program that P & G had already made available to its customers from July 2010 to December 2010. The program limits refunds to one box per household, and requires consumers to provide an original receipt and UPC code clipped from a Pampers box. P & G also agreed, for a period of two years, to add to its Pampers box-label a single sentence suggesting that consumers “consult Pampers.com or call 1-800-Pampers” for “more information on common diapering questions such as choosing the right Pampers product for your baby, preventing diaper leaks, diaper rash, and potty training[.]” P & G similarly agreed, for a period of two years, to add to the Pampers website some rudimentary information about diaper rash {e.g., “[draper rash is usually easily treated and improves within a few days after starting treatment”) and a suggestion to “[s]ee your child’s doctor” if certain severe symptoms develop {e.g., “pus or weeping discharge”), along with two links to other websites. P & G also agreed to contribute $300,000 to a pediatric resident training program — the recipient program is not identified in the agreement — and $100,000 to the American Academy of Pediatrics to fund a program “in the area of skin health.”
The agreement treats named plaintiffs differently than other class members. Named plaintiffs release all of their Pampers-related claims against P & G and receive an “award” of $1000 “per affected child.” (Thus, for example, a named plaintiff with two “affected children” would receive $2000.) Unnamed class members do not receive any award, and benefit only from the labeling and website changes and the one-box refund program (to the extent they have not done so already and have their original receipts and UPC codes). Unnamed plaintiffs are also forced to release their “equitable” claims against P & G, and are “permanently barred and enjoined from seeking to use the class action procedural device in any future lawsuit against” P & G. But in theory, at least, unnamed class members retain the right to file individual lawsuits for “personal injury” or “actual damages” resulting from their children’s use of Dry Max diapers.
Meanwhile, the agreement provides that Plaintiffs’ class counsel will receive a fee award of $2.73 million.
The parties thereafter moved for the district court to certify the class and to approve the settlement agreement. Daniel Greenberg and two other class members objected. Greenberg’s objections ran *71738 pages and were numerous, detailed, and substantive. Among other objections, Greenberg argued that certification of the putative class under Rule 23(b)(2) would be improper because the plaintiffs’ claims are predominantly monetary and because the class members have no ongoing relationship with P & G; that the agreement affords inadequate notice to unnamed class members; that the agreement violates the due-process rights of unnamed class members by depriving them of their rights to seek class-wide monetary relief while denying them the ability to opt-out of the settlement class; that the class representatives and class counsel had not adequately represented the interests of unnamed class members within the meaning of Rule 23(a)(4); and that the settlement was unfair to unnamed class members, in part because of the size of the $2.73 million fee award in comparison to the utility of the injunctive relief (ie., the one-box refund program and labeling and website changes) to unnamed class members.
On September 28, 2011, the district court held a fairness hearing in which it considered whether to certify the class, whether the settlement agreement was fair, whether to approve the request of class counsel for $2.73 million in fees, and whether to approve the “incentive awards” (of $1000 per child) for each named plaintiff. The hearing lasted less than an hour. Counsel for the plaintiffs and for P & G presented argument with little interruption from the district court. Counsel for Greenberg likewise presented argument with virtually no questions or comment from the district court. Near the end of the hearing, the court stated that it would certify the class, that it would approve the $2.73 million fee award and the incentive awards, and that the settlement was fair. The court ventured no response at all to any of Greenberg’s objections, other than to say that they had been “rebutted thoroughly by the parties’ briefs.” Hearing Tr. 34. The court did state, however, that it would “put on a significant written final approval order and final judgment[.]” Id. at 35.
The district court entered its “Final Approval Order and Final Judgment” later that afternoon. With the exception of a few typographical changes, the order was a verbatim copy of a proposed order that the parties had submitted to the court before the hearing. The order was conclusory, for the most part merely reciting the requirements of Rule 23 in stating that they were met. About Greenberg’s objections, the order had nothing to say.
Greenberg’s appeal followed.
II.
We review the district court’s certification of the class and approval of the settlement for an abuse of discretion. Int’l Union, UAW v. Gen. Motors Corp., 497 F.3d 615, 625 (6th Cir.2007).
A.
In class-action settlements, the adversarial process — or what the parties here refer to as their “hard-fought” negotiations — extends only to the amount the defendant will pay, not the manner in which that amount is allocated between the class representatives, class counsel, and unnamed class members. For “the economic reality [is] that a settling defendant is concerned only with its total liabilify[,]” Strong v. BellSouth Telecomms., Inc., 137 F.3d 844, 849 (5th Cir.1998); and thus a settlement’s “allocation between the class payment and the attorneys’ fees is of little or no interest to the defense.” In re Gen. Motors Corp. Pick-Up Truck Fuel Tank Prods. Liab. Litig. (“Gen. Motors Pick-Up Litig.”), 55 F.3d 768, 820 (3d Cir.1995) (internal quotation marks omitted)'.' Hence-unlike in virtually every oth*718er kind of case — in class-action settlements the district court cannot rely on the adversarial process to protect the interests of the persons most affected by the litigation — namely, the class. Instead, the law relies upon the “fiduciary obligation[s]” of the class representatives and, especially, class counsel, to protect those interests. Creative Montessori Learning Ctrs. v. Ashford Gear LLC, 662 F.3d 913, 917 (7th Cir.2011). And that means the courts must carefully scrutinize whether those fiduciary obligations have been met.
“[I]n evaluating the fairness of a settlement,” therefore, we look in part “to whether the settlement gives preferential treatment to the named plaintiffs while only perfunctory relief to unnamed class members.” Vassalle v. Midland Funding LLC, 708 F.3d 747, 755 (6th Cir.2013) (internal quotation marks omitted). “[S]ueh inequities in treatment make a settlement unfair.” Id. The same is true of a settlement that gives preferential treatment to class counsel; for class counsel are no more entitled to disregard their “fiduciary responsibilities” than class representatives are. Gen. Motors Pick-Up Litig., 55 F.3d at 788. Most class counsel are honorable; but “settlement classes create especially lucrative opportunities for putative class attorneys to generate fees for themselves without any effective monitoring by class members who have not yet been apprised of the pendency of the action.” Id. “[T]he danger being that the lawyers might urge a class settlement at a low figure or on a less-than-optimal basis in exchange for red-carpet treatment on fees.” Weinberger v. Great N. Nekoosa Corp., 925 F.2d 518, 524 (1st Cir.1991); see also, e.g., Creative Montessori, 662 F.3d at 918 (“We and other courts have often remarked the incentive of class counsel” to “agree[ ] with the defendant to’ recommend that the judge approve a settlement involving a meager recovery-for the class but generous compensation for the lawyers”). Thus, if the “fees are unreasonably high, the likelihood is that the defendant obtained an economically beneficial concession with regard to the merits provisions, in the form of lower monetary payments to class members or less injunctive relief for the class than could otherwise have [been] obtained.” Staton v. Boeing Co., 327 F.3d 938, 964 (9th Cir.2003). Hence the “courts must be particularly vigilant” for “subtle signs that class counsel have allowed pursuit of their own self-interests and that of certain class members tó infect the negotiations.” Dennis v. Kellogg Co., 697 F.3d 858, 864 (9th Cir.2012) (internal quotation marks omitted).
The signs are not particularly subtle here. On the one hand, the settlement agreement awards class counsel a fee of $2.73 million — this, in a case where counsel did not take a single deposition, serve a single request for written discovery, or even file a response to P & G’s motion to dismiss. On the other hand, the agreement provides unnamed class members a medley of injunctive relief. We must scrutinize that relief to determine whether the fee award amounts to “preferential treatment” in comparison to it.
We begin with the one-box refund program. Consumers cannot benefit from the program unless they have retained their original receipt and Pampers-box UPC code, in some instances for diapers purchased as long ago as August 2008. Greenberg sensibly asks who does this sort of thing. We have no answer. Neither do the parties — or more precisely they have offered none. The omission is conspicuous, for -the refund program here is merely a rerun of the very same program that P & G had already offered to its customers from July 2010 to December 2010. P & G surely has data as to the numbers of consumers who obtained refunds during that time; P & G’s counsel *719conceded as much at oral argument on appeal. And yet — even after Greenberg called out the parties on this very point in his objections to the district court — P & G chose not to provide that data in arguing that the settlement is fair.
“The burden of proving the fairness of the settlement is on the proponents.” 4 Newberg on Class Actions § 11:42 (4th ed.); see also, e.g., Ault v. Walt Disney World Co., 692 F.3d 1212, 1216 (11th Cir.2012); In re Katrina Canal Breaches Litig., 628 F.3d 185, 196 (5th Cir.2010). Thus, to the extent the parties here argue that the settlement was fair because the refund program has actual value for consumers, it was the parties’ burden to prove the fact, rather than Greenberg’s burden to disprove it. The parties did not carry that burden — which again (to his credit) P & G’s counsel conceded at oral argument on appeal.
There is another reason to think the reinstated refund program brings little value to unnamed class members: most of them have already had access to it. The class includes consumers who bought Dry Max Pampers between August 2008 and September 28, 2011. P & G’s initial refund program ended in December 2010. Thus, before this settlement agreement was even reached, consumers who purchased Pampers during a 29-month period — of the 38 months encompassed by the class definition — had already had an opportunity to obtain their single-box refund, without the assistance of class counsel and without assigning away important rights as captive members of a settlement class. That is all the more reason to doubt the parties’ assertions of value. Cf. In re Aqua Dots Prods. Liab. Litig., 654 F.3d 748, 752 (7th Cir.2011) (“A representative who proposes that high transaction costs (notice and attorneys’ fees) be incurred at the class members’ expense to obtain a refund that already is on offer is not adequately protecting the class members’ interests”).
The value of the one-box refund program to unnamed class members is dubious on its face. The parties did not carry their burden to demonstrate otherwise, or indeed even try. The district court, for its part, did not even mention the refund program during the fairness hearing or in its order approving the settlement. Thus, for purposes of this case, the value of the refund program to unnamed class members is negligible.
That leaves the labeling and website changes. The issue, again, is whether the value of these changes is so great, for unnamed class members, as to render counsel’s $2.73 million fee reasonable rather than preferential in light of it. Here is the Pampers-box label change, in its entirety: “For more information on common diapering questions such as choosing the right Pampers product for your baby, preventing diaper leaks, diaper rash, and potty training, please consult Pampers, com or call 1-800-Pampers.” That is all. Greenberg argues that this language — to the extent it amounts to anything— amounts to little more than an advertisement for Pampers. We agree.
The parties do not offer any specific argument to the contrary, other than to say that the labeling change directs consumers to the Pampers.com website, where, it appears, they think the real value lies. Here is the website change in its entirety:
Diaper rash is usually easily treated and improves within a few days after starting home treatment. If your baby’s skin doesn’t improve after a few days of home treatment with over-the-counter ointment and more frequent diaper changes, then talk to your doctor. Sometimes, diaper rash leads to secondary infections that may require pre *720 scription medications. Have your child examined if the rash is severe or the rash worsens despite home treatment. See your child’s doctor if the rash occurs along with any of the following: (1) fever; (2) blisters or boils; (3) a rash that extends beyond the diaper area; (k) pus or weeping discharge.
Useful links: http://w ww.mayoclinic. com/health/diaperrash/DS00069 and http://www.patiented. aap. org/content. aspx?aid-5297
The first paragraph of this language provides only rudimentary information whose value to unnamed class members is negligible. Again the parties offer no specific argument to the contrary. The second paragraph instructs parents to “[s]ee your child’s doctor” if certain rather alarming symptoms develop. We suppose there is some modest value in that suggestion, although the district court said nothing about this point specifically. But we would denigrate the intelligence of ordinary consumers (and thus of the unnamed class members) if we concluded that — absent this suggestion from P & G — they would have little idea to “see [their] child’s doctor” if their child’s rash was accompanied by a fever or boils or “pus or weeping discharge.” And we would denigrate their intelligence still further if we concluded that the value of this suggestion was so great, to ordinary consumers, as to be commensurate with a fee award of $2.73 million. The information contained in this paragraph is neither unknown nor counter-intuitive to most people — the way that information about, say, toxic-shock syndrome would have been to consumers in 1980. Instead the information is common sense, within the ken of ordinary consumers, and thus of limited value to them.
The parties offer several arguments in response. The first is that, absent the new website language, unnamed class members might remain “unaware that diapers themselves can sometimes cause skin irritation requiring medical attention.” Plaintiffs’ Br. at 23. But the website language does not tell them that. The language says to see a doctor if certain symptoms develop, not that the cause of the rash or other symptoms might be the diapers themselves. Indeed it could hardly be otherwise: the settlement agreement expressly states that nothing therein shall be construed as an admission by P & G “of the truth of any fact alleged by Plaintiffs” — of which this alleged fact, above all, was one.
The second argument is that “every square centimeter” of a Pampers-box label is “extremely valuable” to P & G. That may well be true; but it is surely less true for unnamed class members, most of whose Pampers boxes, once.emptied, presumably end up by the curb. The third argument is of a piece: that “[n]o company with an $8-billion-per-year diaper brand wants to put the words ‘blisters,’ ‘boils,’ ‘pus’ or ‘weeping discharge’ on its website, but that is what the Settlement requires of P & G.” Plaintiffs’ Br. at 24. Again we have no reason to doubt that assertion, but it displays the same egocentrism as the last one. To be clear: “The fairness of the settlement must be evaluated primarily based on how it compensates class members ” — not on whether it provides relief to other people, much less on whether it interferes with the defendant’s marketing plans. Synfuel Techs., Inc. v. DHL Express (USA), Inc., 463 F.3d 646, 654 (7th Cir.2006) (emphasis added); see also, e.g., Katrina Canal Breaches Litig., 628 F.3d at 195; Gen. Motors Pick-Up Litig., 55 F.3d at 809-12. So these arguments too are meritless.
The parties’ remaining argument with regard to the website changes, in particular, is that the changes include links to two other sites that include more in-depth in*721formation about diaper rash. The implication, apparently, is that unnamed class members would not find those sites or others like them absent the hyperlinks on the Pampers site. The implication is risible; and the parties have not borne their burden to prove otherwise. Any unnamed class member with the means to access Pampers.com and then follow a link to a more informative website is almost certainly a class member who is familiar with Google. (Indeed, P & G itself cites the likelihood of Google searches by unnamed class members in arguing that the parties provided them with adequate notice of the settlement. P & G Br. at 58 n. 39.) And merely typing “diaper rash” into the Google search engine produces exponentially more links and information than a class member would find on Pampers.com.
In sum, we reject the parties’ assertions regarding the value of this settlement to unnamed class members. Those assertions are premised upon a fictive world, where harried parents of young children clip and retain Pampers UPC codes for years on end, where parents lack the sense (absent intervention by P & G) to call a doctor when their infant displays symptoms like boils and weeping discharge, where those same parents care as acutely as P & G does about every square centimeter of a Pampers box, and where parents regard Pampers.com, rather than Google, as their portal for important information about their children’s health. The relief that this settlement provides to unnamed class members is illusory. But one fact about this settlement is concrete and indisputable: $2.73 million is $2.73 million.
“Cases are better decided on reality than on fiction.” United States v. Priester, 646 F.3d 950, 953 (6th Cir.2011). The reality is that this settlement benefits class counsel vastly more than it does the consumers who comprise the class. The conclusion is unavoidable: this settlement gives “preferential treatment” to class counsel “while only perfunctory relief to unnamed class members.” Vassalle, 708 F.3d at 755 (internal quotation marks omitted). The settlement in this case is not fair within the meaning of Rule 23, and the district court abused its discretion in finding the contrary.
B.
We briefly address Greenberg’s argument that the named plaintiffs are inadequate representatives of the class under Rule 23(a)(4). Under that Rule, we measure the adequacy of the class members’ representation based upon two factors: “1) the representatives must have common interests with unnamed members of the class, and 2) it must appear that the representatives will vigorously prosecute the interests of the class through qualified counsel.” Vassalle, 708 F.3d at 757 (internal quotation marks and alterations omitted). The Rule requires that “the class members have interests that are not antagonistic to one another.” Id. (internal quotation marks omitted). Thus, “the linchpin of the adequacy requirement is the alignment of interests and incentives between the representative plaintiffs and the rest of the class.” Dewey v. Volkswagen Aktiengesellschaft, 681 F.3d 170, 183 (3d Cir.2012).
These requirements are scrutinized more closely, not less, in cases involving a settlement class. For the reasons already explained, see supra at 717-18, “the need for the adequacy of representation finding is particularly acute in settlement class situations[.]” Gen. Motors Pick-Up Litig., 55 F.3d at 795. Thus, the Supreme Court itself has emphasized that the courts must give “undiluted, even heightened, attention in the settlement context[ ]” to the certification requirements of Rule 23. Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 620, *722117 S.Ct. 2231, 138 L.Ed.2d 689 (1997); see also UAW, 497 F.3d at 625 (same).
So we consider the alignment of interests and incentives here. They can be summarized as follows: The named plaintiffs (ie., the class representatives) exercise their Rule 23 rights and receive an award of $1000 per child in return; the unnamed members are barred from exercising those same rights and receive nothing but illusory injunctive relief. Therein lies the conflict. There is no overlap between these deals: they are two separate settlement agreements folded into one. Moreover, there is every reason to think— and again the parties have not attempted to show otherwise — that an award of $1000 per child more than compensates the class representatives for any actual damages they might have incurred as a result of buying Dry Max diapers. And thus, having been, promised the award, the class representatives had “no interest in vigorously prosecuting the [interests of] unnamed class members[.]’ ” Vassalle, 708 F.3d at 757.
Class counsel responds that the $1000 per child payments are merely “incentive” awards, and that incentive awards are common in class litigation. But neither point provides much comfort. Our court has never approved the practice of incentive payments to class representatives, though in fairness we have not disapproved the practice either. See Vassalle, 708 F.3d at 756. Thus, to the extent that incentive awards are common, they are like dandelions on an unmowed lawn— present more by inattention than by design. And we have expressed a “sensible] fear that incentive awards may lead named plaintiffs to expect a bounty for bringing suit or to compromise the interest of the class for personal gain.” Hadix v. Johnson, 322 F.3d 895, 897 (6th Cir.2003).
We have no occasion in this case to lay down a categorical rule one way or the other as to whether incentive payments are permissible. But we do have occasion to make some observations relevant to our decision here. The propriety of incentive payments is arguably at its height when the award represents a fraction of a class representative’s likely damages; for in that case the class representative is left to recover the remainder of his damages by mieans of the same mechanisms that unnamed class members must recover theirs. The members’ incentives are thus aligned. But we should be most dubious of incentive payments when they make the class representatives whole, or (as here) even more than whole; for in that case the class representatives have no reason to care whether the mechanisms available to unnamed class members can provide adequate relief. Accord Radcliffe v. Experian Info. Solutions, 715 F.3d 1157, 1161 (9th Cir.2013) (holding that the “incentive awards significantly exceeded in amount what absent class members could expect upon settlement approval” and thus “creatéd a patent divergence of interests between the named representatives and the class”).
This case falls into the latter scenario. The $1000-per-child payments provided a disincentive for the class members to care about the adequacy of relief afforded unnamed class members, and instead encouraged the class representatives “to compromise the interest of the class for personal gain.” Hadix, 322 F.3d at 897. The result is the settlement agreement in this case. The named plaintiffs are inadequate representatives under Rule 23(a)(4), and the district court abused its discretion in finding the contrary.
We express no opinion regarding Green-berg’s remaining objections to the settlement.
*723The judgment of the district court is reversed, and the case remanded for further proceedings consistent with this opinion.