Court filing
Exhibit 3 — In re Bank of America California Unemployment Benefits Litigation (Dkt. 614-5, S.D. Cal. No. 3:21-md-02992)
Filed January 8, 2026 in In re Bank of America California Unemployment Benefits Litigation; one of 1415 filings from this case.
Record facts
| Court | U.S. District Court for the Southern District of California |
|---|---|
| Filed | 2026-01-08 |
U.S. District Court for the Southern District of California · No. 3:21-md-02992-GPC-MSB · Doc. 614-5 · 2026-01-08 · Docket on CourtListener
Full text
Exhibit 3
Case 3:21-md-02992-GPC-MSB Document 614-5 Filed 01/08/26 PageID.43746
Page 1 of 24
Confidential
UNITED STATES DISTRICT COURTY
SOUTHERN DISTRICT OF CALIFORNIA
IN RE BANK OF AMERICA CALIFORNIA
UNEMPLOYMENT BENEFITS
LITIGATION
Case No. 3:21-md-02992-GPC-MSB
EXPERT REBUTTAL REPORT OF JAY MINNUCCI
November 21, 2024
REDACTED PUBLIC VERSION
Case 3:21-md-02992-GPC-MSB Document 614-5 Filed 01/08/26 PageID.43747
Page 2 of 24
i
TABLE OF CONTENTS
I.
INTRODUCTION ...........................................................................................................1
II.
SUMMARY OF OPINIONS ............................................................................................2
III.
MR. HINDLE IS NOT QUALIFIED TO OPINE ON CALL CENTER STAFFING,
FORECASTING, OR PERFORMANCE MANAGEMENT ...........................................3
IV.
SUBSTANTIALLY ALL CUSTOMER SERVICE CLASS MEMBERS
EXPERIENCED LENGTHY WAIT TIMES FAR EXCEEDING ANY INDUSTRY
STANDARDS ..................................................................................................................4
V.
THE BANK’S SUBSTANDARD CLAIMS CALL CENTER PERFORMANCE IN
FALL 2020 CANNOT BE BLAMED ON UNANTICIPATED PANDEMIC-
RELATED EVENTS ........................................................................................................5
VI.
WAIT TIMES FOR EDD CARDHOLDERS WERE THE SAME AS—OR WORSE
THAN—WAIT TIMES FOR OTHER PREPAID CARDHOLDERS .............................9
VII.
IT IS STANDARD PRACTICE TO USE MULTI-INDUSTRY BENCHMARKS
AND STRATEGIES TO EVALUATE CALL CENTER PERFORMANCE ................11
VIII. THE SYSTEMS TO MAINTAIN AND ACCESS INDIVIDUAL CALL
WAIT TIME ARE IN NEAR-UNIVERSAL USE AMONG LARGE
CALL CENTERS ...........................................................................................................16
Appendix A: Supplemental Materials List ................................................................................21
Case 3:21-md-02992-GPC-MSB Document 614-5 Filed 01/08/26 PageID.43748
Page 3 of 24
Confidential
1
I.
INTRODUCTION
1.
I have been retained as an expert witness in this matter by Cotchett, Pitre &
McCarthy LLP and Altshuler Berzon LLP, co-lead counsel for the Class Plaintiffs. On August
29, 2024, I provided my initial expert report in this matter (“Minnucci Report”).
2.
Based on more than forty years of experience, I described in my Minnucci Report
the call forecasting, staffing, data management, and security practices that are standard across the
call center industry. I also offered the opinions that from September 13, 2020 through November
21, 2020,
.
3.
I have reviewed the October 24, 2024 Declaration of Kelley Lorenzen (“Lorenzen
Decl.”) and the October 24, 2024 Expert Declaration of Steven Hindle (“Hindle Report”), and
documents cited therein. I stand by all of the opinions expressed in my August 29, 2024
Case 3:21-md-02992-GPC-MSB Document 614-5 Filed 01/08/26 PageID.43749
Page 4 of 24
2
Minnucci Report, none of which have been altered by my review of the Hindle Report or
Lorenzen Declaration.1
4.
In this report, I evaluate the portions of the Hindle Report that refer or relate to
my previous report, noting my agreement with some parts of some of his opinions, while noting
my disagreement with many others.
II.
SUMMARY OF OPINIONS
5.
Evaluating the reasonableness of the Bank’s call center staffing and procedures
and assessing the degree of the Bank’s deviation from industry norms requires substantial
experience with and knowledge of call center performance metrics, forecasting processes, and
staffing practices. To the extent that Mr. Hindle’s CV is an accurate description of his
background, he lacks such experience and is not qualified to opine on whether the Bank’s call
center staffing, forecasting, and performance management decisions were reasonable or in
conformity with industry standards.2
6.
Although
for EDD debit cardholders who called the Bank’s Claims call center between September 13,
2020 and November 21, 2020,
.3
7.
1 I have no changes to my CV or compensation, nor additional publications or testimony, to
disclose.
2 See infra ¶¶11-12.
3 See infra ¶¶13-14.
Case 3:21-md-02992-GPC-MSB Document 614-5 Filed 01/08/26 PageID.43750
Page 5 of 24
3
.4
8.
.5
9.
Because callers compare their experiences—and form their reasonable
expectations—based on interactions with call centers in different industries, it is standard
practice for call center leaders to compare across industries when establishing performance
targets. It is also true that callers will contact whichever of a company’s call centers they
anticipate will have the shortest wait; as such, it is standard practice—
—to establish similar, if not identical, wait time targets across generalist and specialty
call centers. Accordingly, multi-industry call center performance data across call center types,
like the data contained in the 2021 U.S. Contact Center Decision-Maker’s Guide, is the best
source for identifying industry-standard performance benchmarks that apply to even specialty
call centers like the Bank’s Claims call center.6
10.
As long as the Bank did not destroy the relevant data,
.7
III.
MR. HINDLE IS NOT QUALIFIED TO OPINE ON CALL CENTER STAFFING,
FORECASTING, OR PERFORMANCE MANAGEMENT.
11.
Mr. Hindle’s CV mentions a position he held in which he was “supporting”
160,000 contact center employees, which is an appropriate and commonly used term for
someone who oversees the Information Technology used by contact center staff. In the body of
4 See infra ¶¶ 15-22.
5 See infra ¶¶ 23-27.
6 See infra ¶¶ 28-35.
7 See infra ¶¶ 36-43.
Case 3:21-md-02992-GPC-MSB Document 614-5 Filed 01/08/26 PageID.43751
Page 6 of 24
4
his report, he also states that he “oversaw” 160,000 employees,8 which suggests that he was in
the direct line of supervision of call center managers, supervisors, and agents. However, nothing
in his CV or Report suggests that he supervised call center managers, supervisors, or agents.
During the time he worked at the various outsourcing organizations he identifies, the
accomplishments he lists are almost exclusively in the realm of cybersecurity, business
continuity, and compliance assurance.
12.
Providing an opinion on call center speed of answer performance requires a
thorough understanding of the data maintained by and metrics used in call centers, typical
industry results, the most reliable forecasting processes used to predict future call volume, and
the staffing practices and timing required to prepare and maintain staff levels appropriate for a
predicted workload. Mr. Hindle’s CV does not demonstrate the requisite level of expertise
regarding any of these topics.
IV.
SUBSTANTIALLY ALL CUSTOMER SERVICE CLASS MEMBERS
EXPERIENCED LENGTHY WAIT TIMES FAR EXCEEDING ANY INDUSTRY
STANDARDS.
13.
The Hindle Report stated: “[A]verage ASAs cannot be used to represent the wait
time experienced by individual class members who called during that week either. In my
experience, each individual caller’s experience varies. There will be individual data points both
above and below the average, and those individual wait times can be very different than the
‘average.’”9 That Report also states: “Neither Mr. Regan nor Mr. Minnucci consider how many
proposed class members experienced wait times either above or below the ASA.”10 Although it
is mathematically true that an average, by definition, does not reflect each individual’s personal
8 Hindle Rep. ¶ 4.
9 Hindle Rep. ¶ 16.
10 Hindle Rep. ¶ 17.
Case 3:21-md-02992-GPC-MSB Document 614-5 Filed 01/08/26 PageID.43752
Page 7 of 24
5
experience, the Hindle Report overstates the amount of meaningful variation from the average
during the Customer Service class period.
14.
As I concluded in my report, “[v]irtually all EDD debit cardholders who
telephoned the Bank between September 13, 2020 and November 21, 2020 and had their calls
routed to the Bank’s Claims call center . . .
.”11 This conclusion is fully supported by the Bank’s performance data,
which shows that between September 13, 2020 and November 21, 2020,
.12 This means that substantially all class members
.
V.
THE BANK’S SUBSTANDARD CLAIMS CALL CENTER PERFORMANCE IN
FALL 2020 CANNOT BE BLAMED ON UNANTICIPATED PANDEMIC-
RELATED EVENTS.
15.
The Hindle Report stated: “Further, it is unreasonable to assume that any financial
institution’s call center, particularly one responsible for servicing complex requests like
transactional fraud claims, could immediately respond to such a significant spike in call volume”
from the COVID-19 pandemic.13 My Report made no such assumption.
16.
Although I observed that the Bank
, my Report did not focus on the immediate aftermath of the
pandemic and the related spike in call volume. Thus, the Hindle Report attempts to refute a claim
that my Report did not make. Instead, my Report observed that, after the Bank had
11 Minnucci Expert Rpt. ¶ 13; see also id. ¶ 44 (
).
12 Ex. 124 at -719115. “Ex.” refers to exhibits to the Chan Declaration and Supplemental Chan
Declaration in support of Plaintiffs’ class certification (“Mot.”).
13 Hindle Rep. ¶ 24.
Case 3:21-md-02992-GPC-MSB Document 614-5 Filed 01/08/26 PageID.43753
Page 8 of 24
6
the Bank then
.14 My report focused on the September to November 2020
time frame, and concluded that the Bank’s decision to
.
17.
As early as April 2020, the Bank
.15
. Although the Bank
because it was
and decided to
16—
Indeed, during a brief five-week period in July and early August,
.
18.
Nonetheless, from the week of July 5, 2020 to the week of September 27, 2020
14 Minnucci Rep. ¶¶ 59-65.
15 Minnucci Rep. ¶¶ 56-57; see also Ex. 133 at -60339.
16 See Ex. 18 (Rule 30(b)(6) Depo. of William Golden (“Golden Tr.”)) 67:8-68:6.
Case 3:21-md-02992-GPC-MSB Document 614-5 Filed 01/08/26 PageID.43754
Page 9 of 24
7
. As stated in my Report,
.17
19.
The Hindle Report emphasizes the extensive nature of the training that must be
provided to call center agents before they can begin answering phones.18 But the
required to
is a common
, and was fully accounted for in my Report.19 This lead time underscores
the importance of the industry-standard practice of staffing based on a forecast, rather than
waiting for performance to deteriorate before activating the hiring and training process.
20.
The Hindle Report also suggested that the Bank’s poor performance could be
attributable to high turnover associated with the pandemic.20 Industry turnover actually dropped
in 2020,21 but in any event, normal turnover
. As shown in Table 1 below, at a time when Finance call centers were seeing a
staffing attrition rate of 28% over 52 weeks (0.54% per week, on average),22
17 Minnucci Rep. ¶¶ 15, 58-63; see also Ex. 77 at -118438 (describing
); Ex. 134 at -106094 (strategy to
by
, after which
).
18 Hindle Rep. ¶ 24.
19 Ex. 18 (Golden Tr.) 69:5-20; Minnucci Rep. ¶¶ 53-58;
20 Hindle Rep. ¶ 24.
21 See Ex. 169 (2024 U.S. Contact Center Verticals: Finance, ContactBabel) at 27 fig. 13
(showing lower annual agent attrition rates for all call centers and for Finance call centers in
2020 compared to 2019).
22 Id.
Case 3:21-md-02992-GPC-MSB Document 614-5 Filed 01/08/26 PageID.43755
Page 10 of 24
8
.23 The
.
Table 1: Attrition Rates Comparison, July 5 to November 21, 2020
21.
Regardless of the reasons behind
, the Bank was clearly aware of
trying to reach the call center. 24 The Bank
23 See Minnucci Rep. ¶ 64, tbl. 5.
24 See Ex. 18 (Golden Tr.) 60:14-61:14 (testifying
).
>""#ABCD
EF**+AICC"-".
LMFN.BNO"NPA
4FP"
LRLASR"T+8
VAWA;<N
LTP=F*A>?@ >""#*AALPP-<P<BNA_
CaWabcbc
Cagbabcbc
Caglabcbc
CabMabcbc
eababcbc
ealabcbc
eagMabcbc
eabdabcbc
eadcabcbc
laMabcbc
lagdabcbc
labcabcbc
labCabcbc
gcaiabcbc
gcaggabcbc
gcageabcbc
gcabWabcbc
ggagabcbc
ggaeabcbc
ggagWabcbc
chWi_
E*F<O+AEF**AE"NP"-An"-CB-OFNT"A;"P-<T+
bcbcA><NFNT"AoN.=+P-AA
p"-P<TF*AR=-T"AALT"-Fr"A
>""#*AALPP-<P<BN
RB=-T"DAAsLkLt@uut;uvmccCglggW
uFPFAkBP"+DAAE*F<O+AEF**A;"P-<T+A+wB[NA<NT*=."AF**Ay-"yF<.Ay-Br-FOAPAy"+
Case 3:21-md-02992-GPC-MSB Document 614-5 Filed 01/08/26 PageID.43756
Page 11 of 24
9
was also aware of
.25 Yet it took the Bank
.
22.
,26
. This did not happen.
.27
VI.
WAIT TIMES FOR EDD CARDHOLDERS WERE THE SAME AS—OR WORSE
THAN—WAIT TIMES FOR OTHER PREPAID CARDHOLDERS.
23.
The Hindle Report stated: “[U]sing data from the Bank’s Claim call center relies
on an unsupported assumption that wait times for CA EDD customers were the same as all
prepaid card programs.”28 With EDD Claims call center volume
.
24.
The Bank did not
.29 It is possible with modern call center phone equipment to
route different groups of cardholders differently;
25 Id.; see also Ex. 126 at -90046
); Ex. 128 at -12883 (
).
26 See Ex. 18 (“Golden Tr.”) 67:8-68:6.
27 See Minnucci Rep. ¶ 64, tbl. 5.
28 Hindle Rep. ¶ 15.
29 See Ex. 18 (Golden Tr.) 57:19-58:4.
Case 3:21-md-02992-GPC-MSB Document 614-5 Filed 01/08/26 PageID.43757
Page 12 of 24
10
. The Bank’s
, it logically follows that any substantially large sub-group of these
callers will experience the same average wait times. As I stated in my Report—and as the Hindle
Report does not dispute30—
.31 As a result, during any given period, EDD cardholders
.
25.
In light of the data provided by the Bank,
. The Bank’s records show that, between September 13 and November 21, 2020,
.32 Because calls
from EDD debit cardholders
.
26.
If there were any variations in service, in my opinion it is likely that EDD
cardholders would have received worse service than other prepaid cardholders received. In my
experience, the only callers who experience meaningfully shorter wait times when contacting a
severely understaffed center are those who place their calls as soon as the call center begins
operations each day. This opinion is consistent with the source cited in the Hindle Report, which
states that “7 AM is the best time to call customer service” and that call center wait times
30 See Hindle Rep. ¶ 15.
31 Minnucci Rep. ¶¶28-29, tbl. 1.
32 Ex. 124 at -719115.
Case 3:21-md-02992-GPC-MSB Document 614-5 Filed 01/08/26 PageID.43758
Page 13 of 24
11
significantly increase after noon.33 Because the Bank’s Claims call center was open from 8:00
am to 10:00 pm Eastern Time Monday through Friday and 9:30 am to 8:00 pm Eastern Time on
Saturday,34 callers living in the Pacific Time Zone, which would include most California EDD
cardholders, would have had to call at 5:00 am Pacific Time to receive the faster early morning
service.35 And while EDD callers were likely under-represented in these early morning hours,
they were likely over-represented at the 7:00 pm Pacific close of the business day. Unfortunately,
any callers in queue at the end of the day would have been dropped from the queue when the last
CSR signed out, regardless of how long they had waited, adding to those EDD callers’
frustration.36
27.
Further, to the extent that the Bank’s call center leaders could influence whether
EDD cardholders received better or worse service than other prepaid cardholders,
37
VII.
IT IS STANDARD PRACTICE TO USE MULTI-INDUSTRY BENCHMARKS
AND STRATEGIES TO EVALUATE CALL CENTER PERFORMANCE.
28.
The Hindle Report stated: “[T]he expectations of target ASAs and realized ASAs
vary depending on the industry, contracts, and the nature and complexity of the requests handled.
For example, in my experience, wait times can be very different for call centers that handle
general inquiries about account balance and certain transactions and those that handle more
complex requests such as unauthorized transactions and account security. In my view, there is no
33 See Hindle Rep. ¶ 16 n.16 (citing TalkDesk, 7 Tips for Getting Better Customer Service,
https://www.talkdesk.com/resources/infographics/tips-for-getting-better-customer-service).
34 Ex. 170 at -1362 ¶ 3.
35 See, e.g., Ex. 12 (Aug. 19, 2024 Willrich Decl.) ¶ 7.
36 See, e.g., id. ¶ 6; Ex. 10 (July 27, 2024 Oosthuizen Decl.) ¶ 5.
37 Ex. 77 at -118438.
Case 3:21-md-02992-GPC-MSB Document 614-5 Filed 01/08/26 PageID.43759
Page 14 of 24
12
standard ASA applicable to all call centers.”38 Hindle’s opinion that there are no relevant
standards is incorrect and reflects a lack of understanding of performance management in call
centers. In my experience, competent call center leaders adopt similar ASA standards, and
deliver similar ASA results, across all lines of business within a company. Doing anything
different not only exposes call centers to complaints from certain segments of a business, but
also encourages callers to “number shop”—that is, calling different numbers or choosing the
wrong options from a menu just to get into a queue that might have a shorter wait.
29.
Although there may be expected differences in handle time, call centers do not
generally expect to see meaningful differences in ASA or abandonment rate between general and
specialty call centers. It should therefore not be surprising that EDD did not distinguish between
Main, Claims, and/or Fraud call centers with respect to the performance benchmarks established
in its contract with the Bank.39 Nor did the Bank
.40 Hindle’s opinion
that average speed to answer varies based on “nature and complexity of the requests handled” is
simply untrue.
30.
The Hindle Report’s assertion that call center performance cannot properly be
compared across industries is also incorrect. In my experience working across industries, there
are not substantial differences between the ASA or abandonment rate targets set in different
38 Hindle Rep. ¶ 22 (footnote omitted).
39 Ex. 171 (Bank-EDD Contract) at 187, Req. 226 (“The Contractor shall, limit the average wait
time to speak to a live CSR to no more than 30 seconds for 70 percent of the calls, and no more
than two (2) minutes for all calls.”).
40 See, e.g., Ex. 126 at -90046 (
;
Ex. 127 at -13092 (
); Ex. 128 at -12883
(
).
Case 3:21-md-02992-GPC-MSB Document 614-5 Filed 01/08/26 PageID.43760
Page 15 of 24
13
industries. In my consulting practice, I have found the most aggressive ASA targets set at 10
seconds and the least aggressive at 300 seconds, while the most aggressive abandoned targets are
2% with the least aggressive at 10%. These targets represent the extremes for individual call
centers; industry averages have much less variation between the most and least aggressive.
31.
Callers develop their expectations regarding how long it should take for a call to
be answered based on their collective experience with all call centers over time, not just call
centers in the same industry or providing the same service. In other words, when an individual
evaluates how quickly Comcast picked up their call, they would not compare Comcast only to
Verizon and AT&T. They would also compare it to AllState, Wells Fargo, Aetna, Wayfair, and
any other company that they may have called recently. Because all call center experiences are
used by customers in evaluating performance, it is standard practice for call centers to consider
cross-industry benchmarks when setting performance targets. Therefore, the other vertical
markets mentioned in Table 1 of the Hindle Report are relevant in the trier of fact’s
consideration of what constitutes an appropriate benchmark for adequate call center performance
in this case. Accordingly, although there may be some differences in performance across
industries, it is appropriate to use a multi-industry composite to identify adequate performance
targets.
32.
The Hindle Report significantly overstates the importance of the travel industry in
the 2021 U.S. Contact Decision-Makers Guide. First, Transport & Travel respondents made up
only 5% of the 2020 survey, while the Finance industry had the greatest representation among
respondents.41 Indeed, 29 (14%) of the respondent call centers were in the Finance field—the
largest group of respondents—while another 28 (13%) were call center outsourcers, similar to
41 Ex. 172 (2021 Decision Makers Guide) at 15.
Case 3:21-md-02992-GPC-MSB Document 614-5 Filed 01/08/26 PageID.43761
Page 16 of 24
14
TTEC, Sykes, and ACT, which operated the Bank’s prepaid call centers.42 Even if one assumes
that the other industries are not relevant, there remain 57 respondents—27%—that are directly
similar to the Bank. The 2021 Decision-Makers Guide is therefore far more heavily weighted
toward call centers similar to the Bank than toward Transport & Travel, or any other industry.
The Hindle Report also overstates the extent to which Transport & Travel respondents were
likely to have meaningfully different speed of answer performance than the Bank. Even if it were
true that the Transport & Travel respondents saw unusually low call volume during some parts of
202043—and it is not clear from the Hindle Report that any supporting evidence exists for such
an assumption—any decrease in call volume was likely matched with a decrease in staffing, as it
was widely reported that airlines and other travel business inflicted prolonged wait times on their
customers during the early months of the pandemic.44 That should not be surprising, as it is
standard practice for businesses to reduce call center staffing commensurate with substantial
reductions in call volumes because by doing so, the businesses avoid the expense of over-
staffing. Accordingly, there is no basis for the Hindle Report’s suggestion that the Travel &
42 Id.
43 Hindle Rep. ¶ 23.
44 See, e.g., Hannah Klein, Slate, How Long Does it Take to Ask Delta a Question (June 16,
2020), https://slate.com/business/2020/06/delta-airlines-customer-service.html (describing
Delta’s graceful disconnect practice and wait times exceeding four hours in June 2020); Matt
Hochberg, Royal Caribbean Blog, Royal Caribbean Hires Back over 100 Laid off Workers to
Help with Long phone Hold Times (May 27, 2020),
https://www.royalcaribbeanblog.com/2020/05/27/royal-caribbean-hires-back-over-100-laid-
workers-help-long-phone-hold-times (describing Royal Caribbean cruises efforts to address
“longer than normal wait times” driven by cruise cancellation announcements and customer
service lay offs); Zach Honig, Points Guy, Having Trouble Getting through to Delta? You’re Not
Alone (June 4, 2020), https://thepointsguy.com/news/how-to-reach-delta (describing the
difficulties of Delta customers and noting that “current call center issues are related to a staffing
shortage, after [Delta] asked agents to take a leave of absence in an effort to control costs”).
Case 3:21-md-02992-GPC-MSB Document 614-5 Filed 01/08/26 PageID.43762
Page 17 of 24
15
Transport respondents to the U.S. Decision-Makers report had meaningfully different
performance on the metrics at issue here: average speed to answer and abandonment rate.
33.
Although it is my opinion, for the reasons stated above, that multi-industry call
center performance data is the best source for identifying an ASA benchmark applicable to the
Bank’s Claims call center, performance data specific to the Finance and Outsourcing industries is
available.45 The average ASA reported in 2020 across the 29 Finance call centers surveyed by
ContactBabel was 145 seconds (2.42 minutes).46 The average ASA reported in 2020 across the
28 Outsourcers surveyed by ContactBabel was 39 seconds.47
34.
From September 13, 2020 to November 21, 2020, the average ASA at the Claims
call center was
. Even if one were to limit the comparison group to the 29
Finance call centers in the survey, the Bank
the 2 minute 25 second average of
the Finance call center peer group. Regardless of whether one compares the Bank’s Claims call
center to the Finance group, the Outsourcing group, or all call centers,
.
35.
The 1.25 minute benchmark identified in my Report is an appropriate, industry
standard ASA.48 Although some businesses may set a faster or slower ASA target, those
variations are minimal. I have never encountered a call center that set an ASA target of more
45 See Ex. 169 (2024 U.S. Contact Center Verticals: Finance, ContactBabel); Ex. 173 (2024 U.S.
Contact Center Verticals: Outsourcing, ContactBabel).
46 See Ex. 169 (Finance Vertical) at 31.
47 See Ex. 173 (Outsourcing Vertical) at 32.
48 Indeed, Mr. Hindle’s own source suggests a much lower benchmark. See Hindle ¶ 16 n.16;
TalkDesk, 7 Tips for Getting Better Customer Service,
https://www.talkdesk.com/resources/infographics/tips-for-getting-better-customer-service/
(indicating that the “average speed to answer is 8.2 seconds”).
Case 3:21-md-02992-GPC-MSB Document 614-5 Filed 01/08/26 PageID.43763
Page 18 of 24
16
than five minutes, and no reputable business would
.
VIII. THE SYSTEMS TO MAINTAIN AND ACCESS INDIVIDUAL CALL WAIT
TIME ARE IN NEAR-UNIVERSAL USE AMONG LARGE CALL CENTERS.
36.
The Hindle Report stated: “[T]he Claim call center systems of record between
September 13, 2020 and November 21, 2020 did not contain or retain data or information
showing the time a particular, individual EDD prepaid cardholder spent on hold when calling the
Claims call center during the Proposed Class Period. This is consistent with my experience that
while call centers may use [individual caller Automatic Number Identifications] ANI, they
typically do not retain . . . wait time[] associated with particular ANIs or account holders.”49 The
Hindle Report bases this opinion on a statement in the Lorenzen Declaration that stated: “None
of the systems of record during the relevant time period that store data related to these customer
service calls recorded data regarding individual call wait times, the speed to answer, or call
handle times. This information is not available on a per caller basis identifiable through Caller
ID, Automatic Number Identification, Alias ID, or any other individual identifier.”50
37.
Neither the Hindle Report nor the Lorenzen Declaration actually responds to the
point I was making in my Report, which was that the “share of cardholders calling the Claims
call center who are members of Plaintiffs’ Customer Service class and assert an injury from the
Bank’s conduct
” because the Bank’s
phone systems
”51
49 Hindle Rep. ¶ 18 (citing Lorenzen Decl. ¶ 13).
50 Lorenzen Decl. ¶ 13.
51 Minnucci Report ¶ 94; see, e.g. Ex. 174 at -1044.
Case 3:21-md-02992-GPC-MSB Document 614-5 Filed 01/08/26 PageID.43764
Page 19 of 24
17
38.
Moreover, the Bank’s own records show—contrary to the declarations of Ms.
Lorenzen and Mr. Hindle52—that the Bank does have the database systems available through
Avaya and NICE that provide individual caller data, and the Bank has used these database
systems in the course of managing its business.53
39.
Ms. Lorenzen’s assertion regarding the Bank’s record keeping practices, if true,
would be substantially out of line with nearly universal call center data management practices,
and Mr. Hindle’s opinion that such a practice is common reflects his lack of experience with call
center performance management.54
40.
In my experience, all phone systems built for call centers create and maintain
automated logs of every incoming call. In the Avaya system
,55 one of the most
robust of these logs is the Call Record Database, which consists of approximately 75 database
items.56 One of these items is named CALLING_PTY, and it houses the phone number of the
52 See Lorenzen Decl. ¶ 13; Hindle Rep. ¶ 18.
53 See, e.g., Ex. 174 at -1044 (call logs showing date of call and start and stop time for certain
class members).
54 Mr. Hindle also states “the outsourced call centers for large financial institutions typically
do not have access to the systems that generate these [ASA] data, nor are they provided these
data by the Bank.” Hindle Rep. ¶ 25. This assertion further demonstrates Mr. Hindle’s lack of
knowledge of both call center operations generally and
. First, an outsourcer could not possibly run their operation without access to ASA
and a range of other performance data, because it is held accountable for results like wait time,
abandonment rate, and average handle time. See, e.g., Ex. 126 at -90046 (
). Second, it is often the outsourcer that
provides this data to the business. At Bank of America,
See Ex. 18 (Golden Tr.) 60:24-
61:14. Indeed, at one point, the Bank’s counsel asserted that such data was exclusively held by its
vendors. Ex. 175 (Feb. 9, 2024 Letter from Bank Counsel indicating that “a portion of the Fraud
Call Center [performance] data was housed on a vendor’s platform” and is not available to the
Bank).
55 Ex. 126 at -90042 (
).
56 See Ex. 176 at 48-50 (July 2016 Avaya Call Management System Database Items and
Calculations listing items in the Call Record Database).
Case 3:21-md-02992-GPC-MSB Document 614-5 Filed 01/08/26 PageID.43765
Page 20 of 24
18
person who made the call (known in the industry as the ANI), which is used to the link a call
record with a specific cardholder.57 The database also contains an item called QUEUETIME,
which provides the time the call segment spent in queue before being answered.58 To determine
the amount of time a specific caller waited in the Claims call center queue, one could query the
database with the caller’s ANI, and once the record is located the time the caller spent in queue
would be provided. Unless the Bank failed to preserve this data, Avaya’s Call Record Database
can be used to identify the precise wait time of each member of the Customer Service class.
41.
There is evidence that the Bank used
. For instance,
59 This
could only have been
, as the Bank
does not
.60
the Agent Trace database and the same Call Record Database that has the
amount of time the caller spent queueing.
42.
Although the Call Record Database from the phone system is the most accurate
way to determine individual caller wait times, individual caller wait times can also be
mechanically calculated by subtracting handle time, available in the Bank’s call recording
system, from total call time, derived from the date and time stamps in the Visa Prepaid
Administration system (Visa PAS). I’ve used records that the Bank produced for Class
57 Id. at 49, 181.
58 Id. at 50, 280.
59 Ex. 177 at -190443.
60 The specific database item accessed to make this determination would have been the
AGT_RELEASED database item, available in the Call Record Database. See Ex. 176 at 48, 161.
Case 3:21-md-02992-GPC-MSB Document 614-5 Filed 01/08/26 PageID.43766
Page 21 of 24
19
Representative Roland Oosthuizen to illustrate this methodology. Table 1 is an excerpt from
Oosthuizen’s Archived Account History on September 29, 2020.61 The first two entries in this
table show Oosthuizen authenticated his identity and left the “VRU” (or “voice response unit,”
which is another name for an IVR) at 6:40 pm. Below that are three actions taken by the CSR
followed by an entry for “CSR Call Completion,” which marks the end of Oosthuizen’s call at
9:21 pm.62 The time between leaving the IVR at 6:40 pm and ending the call with the CSR at
9:21 pm (2 hours and 41 minutes) is equal to Oosthuizen’s hold time (waiting) plus handle time
(talking to the CSR).
Table 2: Account History Transaction Log Excerpt
43.
The Bank also produced records for certain plaintiffs showing the duration of
their calls (handle time) with CSRs.63 An excerpt of a call record produced by the Bank for
61 Ex. 178 at -56916. The Lorenzen Declaration states that Visa PAS was used in the regular
course of business to record transactions, freezes, blocks, and CSR interactions with EDD
cardholders. See Lorenzen Decl. ¶¶ 14-15.
62 During this interaction with the CSR, Oosthuizen was able to file an unauthorized transaction
claim, suggesting that this was the Claims call center; that can be verified by cross-checking the
customer service agent ID (prc384rhph40959u).
63 The Lorenzen Declaration states that call information for EDD cardholders is routinely stored
by the Bank in a system from Neptune Intelligent Computer Engineering (“NICE”), see
Case 3:21-md-02992-GPC-MSB Document 614-5 Filed 01/08/26 PageID.43767
Page 22 of 24
20
Oosthuizen is shown below as Table 2.64 The first entry is the call Oosthuizen had on September
29, 2020, showing a duration of 9 minute and 18 seconds.65 Subtracting the call duration from
the total call time of 2 hours and 41 minutes documented in Oosthuizen’s Archived Account
History isolates his wait time: 2 hours and 32 minutes. The same calculation can be performed
based on the data in the Bank’s records for every Customer Service class member.
Table 3: Call Recording Log
Executed on November 21, 2024, at Collegeville, Pennsylvania.
______________________________
Jay Minnucci
Lorenzen Decl. ¶¶ 6-7, which appears to be the source of the call logs produced by the Bank.
Unless the Bank failed to preserve this data, this information is available for every call answered
by the Claims call center. This contradicts the assertion in paragraph 13 of the Lorenzen
Declaration that “[n]one of the systems of record during the relevant time period that store data
related to these customer service calls recorded data regarding individual call wait times, the
speed to answer, or call handle times.”
64 Ex. 174 at -1044.
65 This can be confirmed as the same call that appears in Oosthuizen’s Archived Account History
by comparing the agent ID, which is also prc384rhph40959u.
Case 3:21-md-02992-GPC-MSB Document 614-5 Filed 01/08/26 PageID.43768
Page 23 of 24
21
Appendix A: Supplemental Materials List
In addition to sources cited in the Report and Appendix D, I considered the following in
developing my opinions:
Declarations and Deposition Transcripts and Accompanying Exhibits
Declaration of Stephen Hindle, October 24, 2024
Declaration of Kelley Lorenzen, October 24, 2024
Call Center Industry Documents and Other External Sources
Avaya Call Management System Database Items and Calculations, July 2016 release
U.S. Contact Center Verticals: Finance, ContactBabel, 2024
U.S. Contact Center Verticals: Outsourcing, ContactBabel, 2024
https://slate.com/business/2020/06/delta-airlines-customer-service.html
https://www.royalcaribbeanblog.com/2020/05/27/royal-caribbean-hires-back-over-100-laid-
workers-help-long-phone-hold-times
https://www.talkdesk.com/resources/infographics/tips-for-getting-better-customer-service/
https://thepointsguy.com/news/how-to-reach-delta
Pleadings and Other Case Documents
Defendant’s Memorandum of Points and Authorities in Opposition to Plaintiffs’ Motion for
Class Certification, In re: Bank of America California Unemployment Benefits Litig., No. 3:21-
md-02992-GPC-MSB, United States District Court for the Southern District of California,
October 24, 2024
February 9, 2024 Letter from Bank Counsel to Plaintiffs Counsel re: Call Center Data
Documents Produced by Defendant
BANA_EDD_MDL-00001044
BANA_EDD_MDL-00001361
BANA_EDD_MDL-00056916
BANA_EDD_MDL-00190443
BANA_EDD_MDL-00291083
BANA_EDD_MDL-00809945
BANA_EDD_MDL-00812294
Case 3:21-md-02992-GPC-MSB Document 614-5 Filed 01/08/26 PageID.43769
Page 24 of 24File and source
- File
- gov.uscourts.casd.709615.614.5.pdf
- Size
- 490,868 bytes
- SHA-256
- 116abfe335ddec62ee6322a5269fd77d217665d93a7b05c5ea177219fb552e1d
- Original
- PACER (login required)