USCA4 Appeal: 24-1507 Doc: 42 Filed: 08/25/2025 Pg: 1 of 13
UNPUBLISHED
UNITED STATES COURT OF APPEALS FOR THE FOURTH CIRCUIT
No. 24-1507
CLARENCE DAVIS,
Plaintiff - Appellant,
v.
CAPITAL ONE N.A.,
Defendant - Appellee.
Appeal from the United States District Court for the Eastern District of Virginia, at Alexandria. Anthony John Trenga, Senior District Judge. (1:22-cv-00903-AJT-IDD)
Argued: March 19, 2025 Decided: August 25, 2024
Before BENJAMIN and BERNER, Circuit Judges, and FLOYD, Senior Circuit Judge.
Affirmed by unpublished opinion. Judge Berner wrote the opinion, in which Judge Benjamin and Judge Floyd joined.
ARGUED: James S. Wertheim, HQ FIRM, P.C., West Jordan, Utah, for Appellant. Jonathan S. Hubbard, TROUTMAN PEPPER LOCKE LLP, Richmond, Virginia, for Appellee. ON BRIEF: Brittany N. Clark, HQ FIRM, P.C., Salt Lake City, Utah, for Appellant. Robert A. Angle, Brooke M. Conkle, Jonathan M. DeMars, TROUTMAN PEPPER LOCK LLP, Richmond, Virginia, for Appellee.
Unpublished opinions are not binding precedent in this circuit. USCA4 Appeal: 24-1507 Doc: 42 Filed: 08/25/2025 Pg: 2 of 13
BERNER, Circuit Judge:
Clarence Davis received multiple prerecorded messages on his cell phone from
Capital One seeking payment for a debt. Davis had never been a Capital One customer and
had never consented to receive calls from Capital One. Davis’s cell phone number
previously belonged to someone else, a Capital One customer who had consented to receive
calls from Capital One before falling delinquent on his account. Davis contacted Capital
One and informed its representative that he was not a customer and that Capital One was
calling the wrong person. Although Davis asked Capital One to stop calling, the
prerecorded messages continued.
Davis then filed this class action suit against Capital One. He alleges that Capital
One violated the Telephone Consumer Protection Act by leaving prerecorded messages on
his cell phone without his consent. Davis moved for certification of a class of individuals
who, like him, were not current Capital One customers but had nonetheless received
prerecorded calls from Capital One. In his motion for class certification, Davis relied
heavily on the testimony of an expert witness who had proposed a methodology to identify
class members.
Capital One opposed class certification, principally on the grounds that members of
the class could not be sufficiently ascertained and that individual questions predominate
over common ones. Capital One also moved to exclude Davis’s expert. The district court,
applying Rule 702 of the Federal Rules of Evidence, granted Capital One’s motion to
exclude Davis’s expert. The district court concluded that the expert’s testimony was not
based on reliable principles or methods. The district court then denied Davis’s motion for 2 USCA4 Appeal: 24-1507 Doc: 42 Filed: 08/25/2025 Pg: 3 of 13
class certification after concluding that the proposed class failed to satisfy Federal Rule of
Civil Procedure 23(b)(3)’s predominance requirement and this court’s ascertainability
requirement.
On appeal, Davis challenges both rulings. Because the district court did not abuse
its discretion either by excluding Davis’s expert or by denying his motion for class
certification, we affirm.
I. Background
A. The Telephone Consumer Protection Act
In 1991, Congress enacted the Telephone Consumer Protection Act (TCPA) “to
prevent abusive telephone marketing practices.” Krakauer v. Dish Network, L.L.C., 925
F.3d 643, 648 (4th Cir. 2019). The TCPA “prohibited almost all robocalls to cell phones.”
Barr v. Am. Ass’n of Pol. Consultants, Inc., 591 U.S. 610, 615 (2020). When the law was
passed, over 18 million Americans received unsolicited calls each day. Id. at 614. The
TCPA is a strict liability statute, because of the recognition that “few individuals would
have an incentive to bring suit, no matter how frustrated they were with the intrusion on
their privacy, the TCPA opted for a model that allows for resolution of issues without
extensive individual complications.” Krakauer, 925 F.3d at 656.
Davis sued under Section 227(b) of the TCPA which makes it unlawful “for any
person within the United States . . . to make any call . . . using any automatic telephone
dialing system or an artificial or prerecorded voice . . . to any telephone number assigned
to a . . . cellular telephone service . . . unless such call is made solely to collect a debt owed
3 USCA4 Appeal: 24-1507 Doc: 42 Filed: 08/25/2025 Pg: 4 of 13
to or guaranteed by the United States” without the prior consent of the called party.
47 U.S.C. § 227(b)(1). Calls made with the prior express consent of the called party are
statutorily exempt from liability. Id. § 227(b)(1)(A).
B. Factual Background
In April 2021, a customer opened a credit card account with Capital One and gave
Capital One consent to call his cell phone. That customer later relinquished his cell phone
number, and the number was reassigned to Davis in March 2022. 1 The customer fell into
delinquency on his Capital One credit card, and Capital One began calling the cell phone
number that had been provided by the customer and later reassigned to Davis in an attempt
to collect the debt. Davis never had a Capital One account and had never consented to
receive calls from the company to his cell phone.
Capital One admits that it initiated debt collection calls to Davis on May 9, 10, 12,
13, 14, 15, and 18 of 2022 and left prerecorded messages on his voicemail on at least four
occasions between May 9 and 14. On May 13, Davis called Capital One to notify the
company that he was not a customer and that it was calling the wrong person. Davis asked
the Capital One representative for Capital One to stop calling him. The Capital One
representative told Davis that Capital One would stop calling but the representative wrote
down Davis’s number incorrectly. Thus, the prerecorded calls continued. Davis called
Capital One again on May 18. As before, he informed the representative that Capital One
1 For privacy reasons, Davis registered his cell phone under a pseudonym and he did not share his cell phone number with any businesses. 4 USCA4 Appeal: 24-1507 Doc: 42 Filed: 08/25/2025 Pg: 5 of 13
was calling the wrong person, and he asked not to be called. After the May 18 call, Davis
received no more prerecorded calls from Capital One.
C. Procedural Background
In August 2022, Davis filed a class action lawsuit against Capital One in the United
States District Court of the Eastern District of Virginia on behalf of himself and similarly
situated persons who were not current customers of Capital One yet received unsolicited
prerecorded messages from the company. Davis alleges that these calls violated the TCPA
because they used an artificial or prerecorded voice and the recipients never consented to
receive calls. Davis claims that the putative class members have received “tens of
thousands” of robocalls from Capital One. J.A. 106. The proposed class is defined as:
All persons or entities throughout the United States (1) to whom Capital One initiated a call (2) directed to a number assigned to a cellular telephone service, but not assigned to a current account holder of Capital One (3) in connection with which Capital One used an artificial or prerecorded voice (4) from four years before the filing of certification.
J.A. 657. Because Davis filed suit in August 2022, the class included individuals that
received robocalls between August 2018 and 2022, in accordance with the TCPA’s
four-year statute of limitations. 47 U.S.C. § 227.
Davis and Capital One engaged in extensive discovery and each retained experts to
analyze the feasibility of identifying class members. Davis’s expert Anya Verkhovskaya
asserted that her proposed methodology made it possible to identify individuals to whom
Capital One initiated prerecorded calls but who were not Capital One customers. Capital
One’s expert David Kalat attempted to discredit Verkhovskaya’s methodology. Kalat’s
testimony criticized the methodology as infeasible and inaccurate. Kalat conducted an 5 USCA4 Appeal: 24-1507 Doc: 42 Filed: 08/25/2025 Pg: 6 of 13
analysis on a sample of 5,000 unique phone numbers provided by Capital One from which
Verkhovskaya had identified 666 potential class members. Kalat found that of those 666
potential class members, over 75 percent were in fact Capital One customers.
Capital One moved to exclude Verkhovskaya’s testimony on the basis that she was
not qualified to serve as an expert. The district court noted that although Verkhovskaya
described her proposed methodology to identify class members, she had never fully
implemented it. Davis v. Cap. One, N.A., Civil Action No. 1:22-CV-00903 (AJT/IDD),
2023 WL 6964051, at *8 (E.D. Va. Oct. 20, 2023). The district court was persuaded by
Kalat’s testimony and concluded that Verkhovskaya’s methodology was unreliable. Id. at
*9. The district court expressed concern that Verkhovskaya was not able to explain how
she would eliminate non-class members, leaving only unproven methodology that “could
potentially be employed at some point in the future, but with no objective measure of its
accuracy in identifying class members.” Id. at *8. The district court also expressed concern
that a database upon which Verkhovskaya proposed to rely on in her analysis was unreliable
for this purpose. Id. at *10–11.
After carefully reviewing the reports of both experts, the district court determined
that Verkhovskaya’s proposed methodology failed to identify proposed class members with
“a sufficient—or any—degree of reliability or feasibility.” Id. at *11. On the basis of this
review, the district court deemed Verkhovskaya’s expert testimony inadmissible under
Federal Rule of Evidence 702. Id.
The district court next analyzed whether Davis’s proposed class satisfied the
requirements for class certification under Rule 23 of the Federal Rules of Civil Procedure: 6 USCA4 Appeal: 24-1507 Doc: 42 Filed: 08/25/2025 Pg: 7 of 13
numerosity, commonality, typicality, and adequacy. Id.; Fed. R. Civ. P. 23(a). Because
Davis sought to certify a class under Rule 23(b)(3), “the questions of law or fact common
to class members [must] predominate over any questions affecting only individual
members.” Fed. R. Civ. P. 23(b)(3). The district court also considered this circuit’s
requirement for class certification, ascertainability. See EQT Prod. Co. v. Adair, 764 F.3d
347, 358 (4th Cir. 2014). The district court ruled that the proposed class satisfied the
numerosity, commonality, typicality, and adequacy requirements, but did not satisfy the
requirements of predominance and ascertainability. Davis, 2023 WL 6964051, at *14–15.
In ruling on ascertainability, the district court concluded that Davis failed to meet
his burden to show that class members could be “‘readily identified’ according to objective
criteria in an ‘administratively feasible’ manner.” Id. at *13.; see Adair, 764 F.3d at 358.
For the same reasons the district court had found Verkhovskaya’s proposed methodology
infeasible and unreliable, the district court found her methodology unhelpful on the issue
of ascertainability. The district court expressed concern that “the resulting group of
individuals identified. . . [might contain] a large number of Capital One customers.” Davis,
2023 WL 6964051 at *13. The district court thus denied Davis’s motion to certify the class.
Id. at *15.
Davis sought leave to file an interlocutory appeal of the district court’s denial of
class certification, which the district court denied. In lieu of trial, Capital One agreed to
entry of judgment on the merits of Davis’s case as an individual plaintiff and agreed to pay
him $2,000 for violating the TCPA. After the district court entered final judgment, Davis
7 USCA4 Appeal: 24-1507 Doc: 42 Filed: 08/25/2025 Pg: 8 of 13
timely appealed the exclusion of his expert witness and the denial of his motion for class
certification.
II. Analysis This appeal presents two issues. First, whether the district court erred in excluding
Davis’s expert witness Anya Verkhovskaya. Second, whether the district court erred in
denying Davis’s motion for class certification. Because Davis relied almost exclusively on
Verkhovskaya’s testimony and methodology in seeking class certification, the two issues
are closely linked. Yet we review each separately and review both under the deferential
abuse of discretion standard. Sardis v. Overhead Door Corp., 10 F.4th 268, 280 (4th Cir.
2021); Adair, 764 F.3d at 357.
A. Exclusion of Davis’s Expert Witness In reviewing the district court’s decision to exclude Davis’s proposed expert witness
Anya Verkhovskaya, we are guided by Rule 702 of the Federal Rules of Evidence, which
governs the admissibility of expert witness testimony, and by the decision of the Supreme
Court in Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993). The “main
purpose of Daubert exclusion is to protect juries from being swayed by dubious scientific
testimony.” Nease v. Ford Motor Co., 848 F.3d 219, 231 (4th Cir. 2017). Daubert
recognized that the district court has a special obligation to act as a gatekeeper to “ensur[e]
that an expert’s testimony both rests on a reliable foundation and is relevant to the task at
hand.” Sardis, 10 F.4th at 281 (alteration and emphasis in original) (quoting Daubert, 509
U.S. at 596). Where the admissibility of expert testimony is specifically challenged, Rule
8 USCA4 Appeal: 24-1507 Doc: 42 Filed: 08/25/2025 Pg: 9 of 13
702 “require[s] that the district court make explicit findings . . . as to the challenged
preconditions to admissibility.” Id. at 283.
District courts may consider a number of factors when determining the reliability of
an expert’s testimony, including whether the expert’s theory or technique can be (and has
been) tested, whether the theory or technique has been subjected to peer review and
publication, the rate of error, and whether the theory or technique is generally accepted
within the relevant community. Daubert, 509 U.S. at 593–94; see also Sardis, 10 F.4th at
283. This inquiry must focus on “principles and methodology, not on the conclusions that
they generate.” Daubert, 509 U.S. at 595; see also Kumho Tire Co. v. Carmichael, 526 U.S.
137, 141–42, 145 (1999) (applying Daubert to non-scientific expert witnesses). These
factors are “flexible,” and district courts have “broad latitude” to determine whether an
expert’s opinion is reliable. Kumho Tire Co., 526 U.S. at 141–42.
Here, the district court properly exercised its discretion to exclude Verkhovskaya’s
expert testimony. Verkhovskaya outlined a four-part methodology for identifying whether
Capital One had called a non-customer. First, she planned to analyze Capital One’s records
and records belonging to telecommunications carriers to determine which of Capital One’s
calls had been made to cell phones. Second, Verkhovskaya planned to use the Reassigned
Numbers Database (RND) to identify whether Capital One initiated a call after a particular
cell phone number had been reassigned. The “RND system securely houses US geographic
and toll-free numbers that have been permanently disconnected (and thus possibly
reassigned) and the date of the most recent permanent disconnection of each those
numbers.” J.A. 157. RND is relatively new, however, and its records only go back to 9 USCA4 Appeal: 24-1507 Doc: 42 Filed: 08/25/2025 Pg: 10 of 13
January 2021. Third, Verkhovskaya planned to use a data broker, PacificEast, to query the
RND to determine whether and when a given number was disconnected and reassigned.
Fourth, Verkhovskaya planned to implement a “historical reverse append” to identify
names and addresses of potential class members, and then to subpoena telecommunications
carriers to identify the owner of each number. 2
Verkhovskaya tested her methodology on a random sample of 5,000 phone numbers
identified by Capital One as having received prerecorded calls. Of these 5,000 numbers,
Verkhovskaya identified 666 as belonging to potential class members because the numbers
had been reassigned since they were initially provided to Capital One. Verkhovskaya never
fully implemented her proposed methodology, however. She forwent entirely the fourth
and final step, the reverse historical append. Verkhovskaya conceded that, in fact, she never
implemented “most of the methodology described in [her] reports in this case.” J.A. 274.
Capital One’s expert David Kalat tested Verkhovskaya’s methodology. His test
showed that 75 percent of the 666 numbers Verkhovskaya identified appeared to belong to
Capital One customers. Kalat’s test further demonstrated that 89 percent of the 666
numbers identified had received prerecorded calls from Capital One before RND records
ever became available. Significantly, Kalat pointed out that even Davis—the named
plaintiff—would not have been identified as a potential class member through
2 A reverse historical append involves inputting a phone number into a database to find out who it belongs to. This relies upon public records, utilities, customer service providers, and credit reporting agencies. When an individual shares his phone number with an entity, such as a business rewards program, that entity often sells the customer’s number to data brokers which then make that number searchable. 10 USCA4 Appeal: 24-1507 Doc: 42 Filed: 08/25/2025 Pg: 11 of 13
Verkhovskaya’s proposed methodology because he had not registered his cell phone
number under his own name but had rather used a pseudonym.
The district court thus had ample bases to question the reliability of Verkhovskaya’s
testimony. First, she admitted that she never fully tested her own methodology. Moreover,
Capital One submitted evidence to show that two of the third-party entities upon which
Verkhovskaya proposed to rely doubted whether they could reliably produce the requisite
data. The RND administrator stated in a declaration that “[a]lthough the RND contains all
reported disconnect dates, the disconnect dates are confidential and not publicly accessible
through the RND public query interface.” J.A. 181. Verkhovskaya also proposed using
PacificEast to query the RND and conduct a reverse historical append. According to
PacificEast, results of its queries “are inherently ambiguous and indefinite with regard to
identifying a date of disconnection.” J.A. 375. With respect to the reverse historical append,
a PacificEast employee stated in a declaration that “determining definitively whether, and
the dates, a particular person subscribed and/or stopped subscribing to or using an input
telephone number, cannot reliably be achieved using the . . . service PacificEast provides.”
J.A. 376.
The district court’s conclusion that Verkhovskaya’s testimony was not the product
of reliable principles and methodology was thus well supported. Verkhovskaya’s failure to
fully test her methodology on the sample provided by Capital One calls into question its
reliability. We also agree with the district court that the high error rate Kalat identified also
“weighs heavily against the admissibility of Verkhovskaya’s expert opinion.” Davis, 2023
WL 6964051, at *11. Davis failed to rebut Capital One’s challenges to the reliability of 11 USCA4 Appeal: 24-1507 Doc: 42 Filed: 08/25/2025 Pg: 12 of 13
Verkhovskaya’s testimony. Id. at *10 Having carefully reviewed the experts’ reports, we
discern no abuse of discretion by the district court in excluding Verkhovskaya as an expert
witness.
B. Class Certification Davis also challenges the district court’s denial of class certification. A district court
abuses its discretion when it “materially misapplies the requirements of Rule 23” or clearly
errs in its factual findings or legal analysis. Stafford v. Bojangles’ Rests., Inc., 123 F.4th
671, 678 (4th Cir. 2024) (quoting Adair, 764 F.3d at 357). The district court found that the
proposed class failed the requirements of ascertainability and predominance. Davis, 2023
WL 6964051, at *12–15. Because we conclude that the district court did not abuse its
discretion in finding that the proposed class was not sufficiently ascertainable, we decline
to reach the question of whether the district court properly concluded the proposed class
also failed the predominance requirement.
Our circuit precedent requires, in addition to the requirements listed in Federal Rule
of Civil Procedure 23, that the members of a proposed class be “readily identifiable.” Adair,
764 F.3d at 358 (quoting Hammond v. Powell, 462 F.2d 1053, 1055 (4th Cir. 1972)). To
satisfy this ascertainability requirement, the district court must be able to “readily identify
the class members in reference to objective criteria” and it must be “administratively
feasible” for the court can do so. Id.
Beyond his reliance on Verkhovskaya’s expert reports and testimony, Davis put
forward little explanation of how the members of the proposed class could be ascertained.
Therefore, our conclusion that the district court did not abuse its discretion in excluding
12 USCA4 Appeal: 24-1507 Doc: 42 Filed: 08/25/2025 Pg: 13 of 13
Davis’s expert also supports our conclusion that the district court did not abuse its
discretion in denying Davis’s motion for class certification. Davis contends that the district
court improperly required the identification of all the members of the class at the time of
certification. Not so. The district court expressly rejected this approach. Davis, 2023 WL
6964051, at *12. The district court concluded that Davis failed to identify objective criteria
that would allow members of the proposed class to be readily identified “without extensive
[and] individualized fact-finding or ‘mini-trials.” Id. (quoting Adair, 764 F.3d at 358)
(internal citations omitted). Because this conclusion was supported by the evidence, we
hold the district court did not abuse its discretion in denying Davis’s motion for class
III. Conclusion For the reasons set forth above, we affirm the order of the district court in full.
AFFIRMED