Class members have appealed, challenging the class-action settlement approved by the district judge. The judge ordered the challengers to post a $3.15 million appeal bond on the ground that if the settlement were delayed Fleet would lose the ability to pay the amounts that it had agreed to pay in the settlement. There was no basis for this concern, and we vacated the bond.

The suit was brought on behalf of approximately 1.6 million persons whose home mortgages were owned by Fleet Mortgage Corporation. It charges that without their permission Fleet transmitted information about their financial needs that it had obtained from their mortgage papers to telemarketing companies which then, in conjunction with Fleet, used that information and deceptive practices to sell those customers financial services they didn't want. The unauthorized transmission of the information to the marketers is alleged to have violated the federal Fair Credit Reporting Act along with state consumer protection laws plus state common law protections against invasion of privacy, while the use of the information to trick people into buying from the telemarketers is alleged to have violated both the federal Telemarketing and Consumer Fraud and Abuse Prevention Act and state consumer protection laws. There are thus two plaintiff classes, a "pure" information-sharing class of 1.4 million customers of Fleet whose financial information Fleet transmitted to the telemarketers but who did not buy anything from them, and a telemarketing class (technically a subclass, but nothing turns on that refinement) consisting of 190,000 class members who were victims of the telemarketers. As far as we can determine, no lawyer represents only members of the pure information-sharing class.

As is common, the suit was filed after a variety of smaller class action and individual suits, testing the legal waters as it were, had been filed against Fleet complaining of the practices that we have outlined. The hope doubtless shared by class counsel and Fleet alike was that a settlement approved by the judge in this comprehensive class action would lead the judge to enjoin the other suits, Williams v. General Electric Capital Auto Lease, Inc., 159 F.3d 266, 275 (7th Cir.1998); In re VMS Securities Litigation, 103 F.3d 1317, 1325-26 (7th Cir.1996); Flanagan v. Arnaiz, 143 F.3d 540, 545-46 (9th Cir.1998); In re Agent Orange Product Liability Litigation, 996 F.2d 1425, 1432 (2d Cir.1993), thus bringing the dispute between Fleet and its 1.6 million customers to a definitive end except for litigation instituted by class members who decided to opt out of the settlement. Even without an injunction, the issuance of a judgment based on the settlement would unless vacated extinguish further litigation by those bound by the judgment by operation of res judicata. But as the citations indicate, injunctions against other litigation are occasionally issued to terminate parallel litigation more quickly and securely than would be the case if the defendant had to interpose a res judicata defense in separate suits.

A settlement was negotiated that the judge approved and this appeal challenges. The challenge focuses on the fact that one of the classes, namely the pure information-sharing class, received absolutely nothing, while surrendering all its members' claims against Fleet. Of course, if their claims were worthless (more precisely, worth too little to justify a distribution — a qualification that we elaborate on below), they lost nothing. But the district judge did not find that their claims were worthless, and it would be surprising if they were. The allegedly unauthorized transmittal of information to the telemarketers may have violated state consumer protection statutes that authorize the victims of the violation to obtain damages; it may also have infringed state common law protections of privacy, including financial privacy. Fleet's "privacy policy" assures its consumers that Fleet "share[s] the minimum amount of information necessary for that company [i.e., the company with which it is sharing the information] to offer its product or service," and this statement, which if false as alleged may well be fraudulent, might support a claim under state consumer protection or privacy law. Such a claim would not be a sure bet, but colorable legal claims are not worthless merely because they may not prevail at trial. A colorable claim may have considerable settlement value (and not merely nuisance settlement value) because the defendant may no more want to assume a nontrivial risk of losing than the plaintiff does.

The members of the telemarketing class received something in the settlement. Fleet agreed to disgorge the profits of its allegedly unlawful conduct. These profits, it appears, had actually come from the members of the information class, but because the total profits were only $243,000, so that the per capita recovery for the 1.4 million members of the class would amount to less than 20 cents, the settlement transferred the profits to the telemarketing class. Fleet further agreed to set aside $2.4 million (roughly 10 times the profits) for payments ranging from $10 to a maximum of $135 per transaction with a telemarketing class member, depending on the character of his transaction with the telemarketers. The part of the $2.4 million that is not claimed will revert to Fleet, and it is likely to be a large part because many people won't bother to do the paperwork necessary to obtain $10, or even a somewhat larger amount. In re Mexico Money Transfer Litigation, 267 F.3d 743, 748 (7th Cir.2001); Martin H. Redish, "Class Actions and the Democratic Difficulty: Rethinking the Intersection of Private Litigation and Public Goals," 2003 U. Chi. Legal Forum 71, 103-04; Gail Hillebrand & Daniel Torrence, "Claims Procedures in Large Consumer Class Actions and Equitable Distribution of Benefits," 28 Santa Clara L.Rev. 747, 751-53 (1988). The district judge has approved a handsome fee for the class lawyers, $750,000, despite the meagerness of the relief agreed to in the settlement.

Fleet, joined by the class counsel, argues that the members of the pure information-sharing class didn't really receive nothing in exchange for giving up their claims; they received the emotional satisfaction of knowing that Fleet had been forced to give up its profits. That is a preposterous argument. Supposing that each of the 1.4 million members of the information-sharing class could expect a damages award of, say, $25, the total damages of the class would be $35 million. The idea that a rational fiduciary would surrender a claim worth $35 million in exchange for the satisfaction of knowing that his wrongdoer had been forced to pay $243,000 to members of another class staggers the imagination.

The lawyer for plaintiff Mirfasihi, a representative of the information-sharing class who for his service as representative will receive $250 if the settlement is approved even though the settlement is worth nothing to the people he supposedly is representing, tells us that although the information-sharing class did not obtain a "formal" injunction against Fleet's unlawful sharing of personal financial information, it obtained injunctive relief "in effect" because several months after this suit was filed Fleet sold its mortgage business. But at argument the lawyer repeatedly disclaimed any suggestion that the suit had induced the sale. According to press reports, Fleet sold because "it wasn't thrilled with the mortgage business, which operates in a cutthroat market and offers narrow profit margins." John Hechinger & Nikhil Deogun, "Washington Mutual Nears a Deal to Buy Loan Unit of FleetBoston," Wall St. J., p. B8, Apr. 2, 2001. Nor would the sale as such provide any prospective relief to the information-sharing class, since the buyer is free to continue the same practices that the seller engaged in. No retrospective relief, no prospective relief, and no "emotional balm" relief.

Fleet and the class counsel contend that the information-sharing class obtained a "cy pres" remedy. The reference is to the trust doctrine that if the funds in a charitable trust can no longer be devoted to the purpose for which the trust was created, they may be diverted to a related purpose; and so the March of Dimes Foundation was permitted to reorient its activities from combating polio to combating other childhood diseases when the polio vaccine was developed. The doctrine, or rather something parading under its name, has been applied in class action cases, In re Mexico Money Transfer Litigation, supra, 267 F.3d at 748-49; Six (6) Mexican Workers v. Arizona Citrus Growers, 904 F.2d 1301, 1305 (9th Cir.1990); 4 Alba Conte & Herbert B. Newberg, Newberg on Class Actions § 11:20 (4th ed.2002), but for a reason unrelated to the reason for the trust doctrine. That doctrine is based on the idea that the settlor would have preferred a modest alteration in the terms of the trust to having the corpus revert to his residuary legatees. So there is an indirect benefit to the settlor. In the class action context the reason for appealing to cy pres is to prevent the defendant from walking away from the litigation scot-free because of the infeasibility of distributing the proceeds of the settlement (or the judgment, in the rare case in which a class action goes to trial) to the class members. There is no indirect benefit to the class from the defendant's giving the money to someone else. In such a case the "cy pres" remedy (badly misnamed, but the alternative term — "fluid recovery" — is no less misleading) is purely punitive.

The class counsel and Fleet point out that if the $243,000 in restitution due the members of the information-sharing class were spread evenly across the 1.4 million members of that class, each member would receive an amount smaller than the cost of postage; therefore the telemarketing class should receive it (as part of the $2.4 million) because it has fewer members. The argument founders on the fact that restitution is merely an alternative remedy to damages. The aggregate damages of the information-sharing class might, as we said, be as high as $35 million (or even higher), and then there would be no reason for thinking distribution to the class members infeasible. See Mace v. Van Ru Credit Corp., 109 F.3d 338, 345 (7th Cir. 1997); Molski v. Gleich, 318 F.3d 937, 945-55 (9th Cir.2003). Remember that the $10 minimum entitlement of the telemarketing class members will be distributed to them. Of course the per capita damages of the information class might turn out to be too low to cover the costs of distribution (not 20 cents, but not $10 either), or might cover them so thinly as to be disproportionate to the net distribution. And then the question would arise whether a punitive remedy is appropriate in a suit in which the plaintiffs might never have been able to claim punitive damages. We needn't try to answer that question.

Would it be too cynical to speculate that what may be going on here is that class counsel wanted a settlement that would give them a generous fee and Fleet wanted a settlement that would extinguish 1.4 million claims against it at no cost to itself? The settlement that the district judge approved sold these 1.4 million claimants down the river. Only if they had no claim — more precisely no claim large enough to justify a distribution to them — did they lose nothing by the settlement, and the judge made no finding that they had no such claim.

Because class actions are rife with potential conflicts of interest between class counsel and class members, Reynolds v. Beneficial National Bank, 288 F.3d 277, 279-83 (7th Cir.2002); In re General Motors Corp. Pick-Up Truck Fuel Tank Products Liability Litigation, 55 F.3d 768, 801-05 (3d Cir.1995); Weinberger v. Great Northern Nekoosa Corp., 925 F.2d 518, 524 (1st Cir.1991); John C. Coffee, Jr., "Class Action Accountability: Reconciling Exit, Voice, and Loyalty in Representative Litigation," 100 Colum. L.Rev. 370, 385-93 (2000), district judges presiding over such actions are expected to give careful scrutiny to the terms of proposed settlements in order to make sure that class counsel are behaving as honest fiduciaries for the class as a whole. Uhl v. Thoroughbred Technology & Telecommunications, Inc., 309 F.3d 978, 985 (7th Cir.2002); Culver v. City of Milwaukee, 277 F.3d 908, 915 (7th Cir. 2002); In re Cendant Corp. Litigation, 264 F.3d 201, 231 (3d Cir.2001); Grant v. Bethlehem Steel Corp., 823 F.2d 20, 23 (2d Cir.1987). Unfortunately the district judge's decision approving the settlement does not discuss the settlement's questionable features — not only the one we've stressed, namely the denial of any relief to an entire class, the kind of thing that led to rejection of the settlements in Crawford v. Equifax Payment Services, Inc., 201 F.3d 877, 880-82 (7th Cir.2000), and Molski v. Gleich, supra, 318 F.3d at 953-54; cf. In re General Motors Corp. Engine Interchange Litigation, 594 F.2d 1106, 1133-34 (7th Cir.1979), but also the reversion of unclaimed refunds to the putative wrongdoer and the fact that the class that was denied relief did not have separate counsel from the counsel for the favored class. Ortiz v. Fibreboard Corp., 527 U.S. 815, 852-53, 119 S.Ct. 2295, 144 L.Ed.2d 715 (1999); Culver v. City of Milwaukee, supra, 277 F.3d at 912-13; In re General Motors Corp. Pick-Up Truck Fuel Tank Products Liability Litigation, supra, 55 F.3d at 800-01; cf. Amchem Products, Inc. v. Windsor, 521 U.S. 591, 625-28, 117 S.Ct. 2231, 138 L.Ed.2d 689 (1997). Also not discussed is the failure of the notice of settlement to inform members of the information-sharing class (who, remember, were being told that their share of the settlement was a big fat zero) of a pending case in a Massachusetts state court seeking monetary relief on behalf of residents of that state who were members of the information-sharing class and whose rights would be extinguished if the settlement in the present case was approved, unless they opted out of it.

All these were warning signs, no more; we do not suggest that deletion of the reversion provision, or notice of the Massachusetts suit, or even the award of some relief for the information class, were per se requirements of an acceptable settlement. A reversion provision might encourage a more generous settlement offer. Notice of a pending suit that might offer only remote prospects of success might confuse class members and precipitate imprudent opting out. Even the denial of all relief to the information class might be justified if careful scrutiny indicated that the class had no realistic prospect of sufficient success to enable an actual distribution to the class members. But these were matters to be considered, not assumed. The last is the most important, especially in view of the fact that the information class did not have separate counsel. In Reynolds v. Beneficial National Bank, supra, 288 F.3d at 284-85, we emphasized the district judge's duty in a class action settlement situation to estimate the litigation value of the claims of the class and determine whether the settlement is a reasonable approximation of that value. Id.; see also Mars Steel Corp. v. Continental Illinois National Bank & Trust Co., 834 F.2d 677, 682 (7th Cir.1987); In re General Motors Corp. Engine Interchange Litigation, supra, 594 F.2d at 1132 n. 44; In re General Motors Corp. Pick-Up Truck Fuel Tank Products Liability Litigation, supra, 55 F.3d at 806; In re Traffic Executive Association-Eastern Railroads, 627 F.2d 631, 633 (2d Cir.1980). The district judge in this case made no estimate of the value of the legal claims of the information-sharing class.

So the settlement cannot stand. But we disagree with the objectors that the members of the information-sharing class are entitled to individual notice. Such notice is preferable to newspaper or other collective notice but it was impossible here because Fleet has no record of those customers whose financial information it gave the telemarketers but who did not buy anything from the latter. (Those who did buy, the members of the telemarketing class, received individual notice.) Maybe Fleet should have such a record, but it doesn't and so individual notice is impossible. Nor could it have enclosed a notice with the monthly statements that it sends all its mortgage customers (necessarily including the members of the information class, or at least members current as of the date the notice would have been mailed), because it sold its mortgage business before it was ordered to notify the class.

When individual notice is infeasible, notice by publication in a newspaper of national circulation (here USA Weekend, a magazine that is included in hundreds of Sunday newspapers) is an acceptable substitute. Fed.R.Civ.P. 23(c)(2); Mullane v. Central Hanover Bank & Trust Co., 339 U.S. 306, 317, 70 S.Ct. 652, 94 L.Ed. 865 (1950); In re Agent Orange Product Liability Litigation, 818 F.2d 145, 167-69 (2d Cir.1987); Montelongo v. Meese, 803 F.2d 1341, 1351-52 (5th Cir.1986). Something is better than nothing. But in this age of electronic communications, newspaper notice alone is not always an adequate alternative to individual notice. (See Brian Walters, "`Best Notice Practicable' in the Twenty-First Century," 2003 UCLA J.L. & Tech. 4, discussing N.D. Cal. Civ. L.R. 23-2, which requires that notice of securities class actions be posted to an online clearinghouse maintained by Stanford Law School.) The World Wide Web is an increasingly important method of communication, and, of particular pertinence here, an increasingly important substitute for newspapers. Although Fleet did not post a notice on its own website, a firm that was hired to administer the settlement maintained a website with details of the case, and so far as appears that was an acceptable substitute.

The question of the adequacy of the notice is separate from whether the settlement should have been approved. It should not have been, given what was before the district judge. The judgment is therefore reversed and the case is remanded for further proceedings consistent with this opinion. Circuit Rule 36 shall apply on remand.

REVERSED AND REMANDED.