In Wilson v. Babcock Home Furniture, No. 8:17-C-02739-T-02AAS, 2018 WL 6660029 (M.D. Fla. December 19, 2018), Judge Jung denied class certification in a TCPA “wrong-number” class due to lack of ascertainability or common questions of law/fact.

At the most fundamental level, the parties’ experts dispute the precise amount of “wrong numbers.” Based on Plaintiff’s analysis, Defendant made 8,253 calls marked as wrong number to 7,705 cellphone numbers. Dkt. 56-1 at 16 According to the defense expert, Jan Kostyun, the number of relevant calls is significantly lower than this. Dkt. 59-7 at 36-40. The dispute extends to whether it is even possible to identify the cell phone subscribers of the alleged wrong numbers. According to Defendant’s expert, there is no public database of cell phone subscribers and private services are often inaccurate and incomplete, and even a carrier-by-carrier investigation would not ensure accuracy. Dkt. 59-7 at 13-19. Plaintiff’s expert claims this is possible, though one of her chief sources is the unobservable, proprietary “black box” techniques of LexisNexis. Dkt. 59-7 at 264. Be that as it may, the processing of the thousands of “wrong numbers” in Defendant’s database is where the matter truly falls apart. Most glaringly, Plaintiff’s method would not even have discovered Plaintiff as a class member because she is not the subscriber of the number and there is no identification of her in the provided Metro PCS/T-Mobile records. Dkt. 59-6 at 1870-74. In fact, this expert encountered a similar problem in another TCPA case, Goins v. Palmer Recovery Attorneys, PLLC, 6:17-cv-654-Orl-GAP-KRS. Dkt. 59-7 at 18. There, despite her stated ability to ascertain the plaintiff as a class member, an individualized inquiry based on Ms. Verkhovskaya’s own knowledge about the case was necessary to identify the plaintiff as the called party. Dkt. 59-7 at 17-19. While this is by itself not fatal, it does highlight a weakness in the proposal: Short of relying on a claimant’s assertions, there is no way definitively to determine who actually answered the call from Defendant and stated “wrong number.” See Stalley v. ADS Alliance Data Sys., Inc., 296 F.R.D. 670, 679 (M.D. Fla. 2013) (denying certification where the court was “not been presented with reasonable methods for ascertaining the identity of the individuals who answered [the defendant’s] collection calls”). Making matters worse, there are documented instances of more than one customer providing the same phone number. Dkt. 59-1 at 6; Dkt. 59-2; Dkt. 59-3. This “multiple hit” scenario means that a call to an otherwise consenting customer might be designated as “wrong number” simply because Defendant had intended to call—and asked for—the other customer who provided the number. Additionally, Defendant was able to contact by phone some of its customers at their provided numbers after those numbers were initially designated as “wrong.” Dkt. 59-1 at 6; Dkt. 59-2 at 2. Plaintiff points out that the names obtained through subpoenas or reverse directory databases can be cross-referenced against Defendant’s customer records and only names that appear on both lists will proceed to the next stage of the certification process.3 Dkt. 56 at 13-14. This additional step would provide some indicia of reliability, here concerning the subscriber/user distinction, notably absent in other cases. See, e.g., Marcus v. BMW of N. Am., LLC, 687 F.3d 583, 593-94 (3d Cir. 2012). Yet this precaution risks excluding otherwise appropriate members from the class. Plaintiff also assures the Court that any discrepancies will be worked out by a subsequent self-attestation process of mailing notices and claim forms to the matched names and addresses, but the Court is skeptical. First, such a step ignores the very purpose of ascertainment and will, in any event, require an individualized inquiry. Secondly, while an affidavit certifying inclusion in a class might be appropriate in some cases where damages for an individual claimant are negligible, see e.g., Mullins v. Direct Digital, LLC, 795 F.3d 654, 667 (7th Cir. 2015), here Defendant could face up to $1500 per call. This amount is relevant both as an incentive for individuals to improperly enter the class and, as discussed more fully below, a danger that impacts due process protections for Defendant. Karhu, 621 F. App’x at 948-49 (“[A]llowing class members to self-identify without affording defendants the opportunity to challenge class membership provide[s] inadequate procedural protection to … [defendant[s] and implicate[s their] due process rights[, but] protecting defendants’ due-process rights by allowing them to challenge each claimant’s class membership is administratively infeasible, because it requires a series of minitrials just to evaluate the threshold issue of which [persons] are class members.”) (alterations in original) (internal quotation marks and citations omitted). It is worth mentioning that, as the class is most recently defined, the Court would not need to evaluate consent as a gatekeeping issue. By contrast, the original version limited the class to users from whom Defendant “did not have express consent to call said cellular telephone number,” and a subclass of users who “had revoked their consent for [Defendant] to place such calls.” Dkt. 28 at ¶ 54.5 Perhaps realizing this would require the Court to make individualized inquiries at the ascertainment stage, Plaintiff restyled the class to its current form of “wrong number.” But this only delays the necessary, fact-intensive inquiry into consent, as discussed below.   Notwithstanding the above points, a general reluctance by many courts to deny class certification because of administrative difficulties, even in ensuring that all members of a class are accounted for, gives the Court pause. Because the Court finds that predominance is not satisfied, however, this Court need not base its ruling upon, nor definitively resolve, the issue of ascertainment. . . . Simply considering what Plaintiff wants to do shows the problem: Plaintiff says first identify the “wrong numbers” listed in Defendant’s calling records. That can be done. Plaintiff then claims her expert can divine the subscribers and affidavits can be mailed to those addresses to determine those with standing. For example, Plaintiff, an authorized user, is shown on no phone records but does have standing. But Plaintiff’s grandmother, the subscriber and payer of the unlimited calling plan, has no standing. This determination is highly doubtful and the self-attestation process has been rejected by other courts. See Karhu, 621 F. App’x at 948-49. Worse yet, it ignores the real hole in this case: deriving consent. A mail-in affidavit for a $45,000 claim is not going to work in this class setting. Consent requires an individualized inquiry, especially when the source list, by definition, is consented as with credit applications.