Maurer v. Sysco Albany, LLC

Opinion

UNITED STATES DISTRICT COURT NORTHERN DISTRICT OF NEW YORK

CHARLES MAURER, Plaintiff, V. No. 1:19-CV-821 SYSCO ALBANY, LLC, et al., (TJM/CFH) Defendants.

APPEARANCES: Law Offices of Wyatt & Associates PLLC BENJAMIN J. WYATT, ESQ. 17 Elm Street Suite C211 TIMOTHY J. BROCK, ESQ. Keene, New Hampshire 03431 Attorneys for Plaintiff

i Jackson Lewis P.C. KRISTI RICH WINTERS, ESQ. 677 Broadway 9th Floor Albany, New York 12207 Attorney for Defendants DECISION & ORDER Presently pending before the Court is Charles Maurer’s (“plaintiff”) motion to compel defendants Sysco Albany, LLC (“Sysco”), Thomas Teal (“Teal”), and Craig Wittenberg (“Wittenberg”) (collectively, where appropriate, “defendants”) to produce certain documents and electronically saved information (“ESI”). See Dkt. No. 23. Defendants filed a response in opposition to plaintiff's motion. See Dkt. No. 24. For the reasons that follow, plaintiff's motion is granted in part and denied in part.

1. Background A. Plaintiff's Complaint

For purposes of this motion, the undersigned will assume the parties’ familiarity with the facts and provide only a brief summation of the factual and procedural background as relevant to the present motion. Plaintiff is a New York State resident who was employed at Sysco from May 2003 until his employment was terminated on January 5, 2018. See Dkt. No. 1 (“Compl.”) at 191,298, 57 34, 69 38. Plaintiff o began his career at Sysco as a marketing associate, and was promoted to district sales manager. See id. at 312. Over the course of his employment at Sysco, plaintiff states that he “frequently received raises and bonuses because of his strong performance, including most recently in early December 2017.” Id. at 7.13. Plaintiff contends that he “suffers from diagnosed attention deficit/hyperactivity disorder (“ADHD”), an anxiety disorder, and a depression disorder[,]’ which “substantially limit[] m}one or more major life activities and/or bodily functions, including but not limited to his ability to be happy, focus, control his emotions, maintain a normal sleep schedule, and to socialize with others, as well as impacting his neurological functions.” Id. at J¥J 16, 17. Plaintiff posits that, “[t]lowards the latter half of 2017, [he] was experiencing heightened symptoms related to his disabilities and he was feeling extremely depressed m|and anxious.” Compl. at 3 17. Plaintiff alleges that, “[o]Jn or around . . . December 18, 2017, [he] met with .. . Wittenberg . . ., a regional manager for [Sysco], after a . . . visit

... With a customer.” Id. at 418. During this meeting, plaintiff states that he “disclosed that he had ADHD, anxiety, and depression and was suffering a severe flare up of these disabilities.” Id. at | 19. On December 20, 2017, plaintiff attended a meeting with Wittenberg and Teal, a vice president at Sysco, in which plaintiff “disclosed

to them that his medical provider had ordered blood tests to be done related to his disabilities and was evaluating his disability-related medication.” Id. at 5 29. Plaintiff also informed Wittenberg and Teal “that in the near future he planned to submit a formal request for medical leave (in the form of FMLA leave) related to his disabilities/serious medical conditions.” Id. at J 30. After a “pre-scheduled and approved vacation for the Christmas holidays[,]’ plaintiff “returned to work on or around January 2, 2018, at which time he had received the bloodwork back from his therapist and it had been determined that he was going to need to go out on medical leave.” Id. at ff] 31, 33. On January 5, 2018, plaintiff attended a meeting with Teal and non-party Lynn Harris (“Harris”), Sysco’s vice president of human resources. See id. at J 34. Plaintiff states that he “opened the meeting by reiterating to... Teal and... Harris that he was formally requesting a medical leave of absence pursuant to the Family and Medical Leave Act at the recommendation of his therapist (a nurse practitioner).” Id. at 6 J 35. Plaintiff contends that “[t]his constituted both a request for FMLA leave . . . and a requested [sic] for a disability-related accommodation.” Id. at 36. Plaintiff alleges that, following his request, Teal informed him that Sysco “would not allow [plaintiff] to go out on leave” and “then suddenly declared that [plaintiff] was being terminated, allegedly for performance.” at If] 37, 38. Plaintiff posits that he was not subject to Sysco’s “mandatory progressive discipline policy,” which he states the company typically implements for employees prior to termination for performance, and includes “a verbal warning, a written warning, and a 30/60/90 action plan (a performance improvement plan with goals that must be completed with 30 days, 60 days, and . . . 90 days), all prior to termination.” Id. at J] 39, 40. Plaintiff also contends that he was not demoted to a

lower position, unlike other, non-disabled, district sales managers who “fail[ed] to meet the goals of a 30/60/90 plan.” Id. at 7 J 50. On July 10, 2019, plaintiff filed the present lawsuit, alleging claims of disability discrimination, failure to provide reasonable accommodations, and retaliation pursuant to Section 296 of the New York State Human Rights Law (“NYSHRL’); disability ° discrimination, failure to provide reasonable accommodations, and retaliation pursuant to the Americans with Disabilities Act (“ADA”), 42 U.S.C. §§ 21101, ef seq.; interference and retaliation for exercising rights under the Family Medical Leave Act (“FMLA”), 29 U.S.C. § 2615. See Compl. at 9-19. Defendants filed an answer in which they denied the allegations in the complaint and asserted various affirmative defenses. See DKt. No. 6. I B. Plaintiff’s Initial Discovery Demand On November 4, 2019, plaintiff served defendants with his first set of interrogatories, document requests, and proposed email custodian accounts to be searched with proposed search terms and protocols for ESI production and format. Se Dkt. No. 23-1 at 30, 31-89 (interrogatories), 89-130 (document requests), 132-39 (proposed search terms and protocols for ESI production/format). As relevant here, plaintiff's document request included: “[a]ll documents relating to the work performance of [plaintiff], including, but not limited to, evaluations or feedback by any other employee relating to [plainitff's] performance, conduct, or duties[,]’ id. at 100; “[a]ll documents relating to any weekly, monthly, annual or other performance goal of each district in the Company from January 1, 2013 to the present, including but not limited to sales growth”

id. at 101; “[a]ll documents related to whether the performance goals of each district in the Company were met from January 1, 2013 to present, including but not limited to sales growth[,]” id.; “[a]ll documents related to the goals and expectations of each and every employee in a District Sales Manager position (including but not limited to [plaintiff], Doug Sheraton, and Scott Nessis) at the Company from January 1, 2013 to present.” Id. at 102. Concerning the production of ESI, plaintiffs November 4, 2019 discovery demands requested that the “[c]ustodian [email a]Jccounts” of Charles Mauer, Thomas Teal, Craig Wittenberg, Lynn Harris, and any other individual who communicated with the foregoing persons concerning the termination of plaintiff's employment from Sysco, ' be searched for “all documents ever shown within inbox, outbox, drafts, or sent items, including where such custodian is listed in the to, from, cc, and bcc fields” generated between “January 1, 2013 to Present.” Dkt. No. 23-1 at 132. Plaintiff provided a proposed list of 22 separate “[s]earch [t]erms to be utilized for all custodians except” him. Id. These terms included numerous variations of plaintiffs name and initials, including “!Chares!”; “!Charlie!”; “!Mauer!”; “!CM”; “ICM’s”; and “disabl!”; “ADHD!”; “depress!”; “anxiety!”; “leave!”; “therapist!”; “racing!”; “racing thought!”; “EEOC!”; “Equal Employment Opportunity Commission!”; “NYSDHR!”; “NYDHR!”; “New York State Division of Human Rights!” “litig!”; “law!”; “illegal!’; and “unfair!” Id. Further, plaintiff proposed the following “[s]earch [t]erms to be utilized for [his] account:” “tThomas!’;

1 Plaintiffs proposal read “[a]ny individual(s) identified in Defendant Sysco Albany, LLC’s answer to Plaintiff's Interrogatory #2.” Dkt. No. 23-1 at 132. Plaintiff's reference to Interrogatory #2 appears to be a typographical error, as Interrogatory #2 reads, “Please describe each and every reason for your decision to implement, or your support of, the termination of [plaintiff's] employment in or around January 2018.” Id. at 39. Rather, plaintiff appears to have intended to reference Interrogatory #3, which states “Please identify each and every individual who you communicated with (whether verbally or in writing) related to the actual or contemplated decision to terminate [plaintiff's] employment.” Id.

“‘'Tom!”; “tTeall!”; “!Wittenberg!”; “Lynn!”; “tHarris!”; “disable!”; “ADHD!”; “anxiety!”; “leave!” “therapist!”; !”; “therapist!”; “counsel!”; “racing!”; “racing thought!”; “Human Resource!”; “HR!”; “complain!”; “concern!”; “EEOC!”; “Equal Employment Opportunity Commission!”; “NYSDHR!”; “NYDHR!”; “New York State Division of Human Rights!” “litig!”; “law!”; “illegal!’; “warn!”; “fir!’; “termin!’; bias!”; “improper!”; “discriminat!”; and “retaliate!” Id. Plaintiff further proposed that “[s]earch terms should be run in the following fields/areas: to, from, CC, BCC, subject line, text of all emails (including chain data), and text of all attachments.” Id. Moreover, plaintiff proposed various terms relating to production format of ESI. See id. at 133. As relevant, paragraph 9 of plaintiff's proposed production format, titled “Deduplication,” stated: A party is only required to produce a single copy of a responsive document and a party may globally across [sic] custodians. However, attachments to emails shall not be eliminated from the parent email, even if the attachment is an exact duplicate of a file elsewhere in the document set. Prior to producing documents, the parties shall disclose whether they intend to deduplicate their production and, if so, which de-duplication software and methodologies they intend to use. Id. at 136-37 | 9.

ll. Present Motion A. Plaintiff's Memorandum of Law in Support of Motion to Compel 1. Plaintiff's Recitation of Relevant Procedural History and Negotiations Concerning Disclosure of ESI and Sales Report Data Plaintiff states that defendants responded to his initial discovery demands on January 23, 2020 responses, and “included limited documents and virtually no emails (approximately two emails were produced, in fact it was clear that no ESI search had

been run at all)” and “did not contain any information related to [p]laintiff's sales numbers or the sales numbers of the other members of the sales department for which [plaintiff's] sales were compared.” Dkt. No. 23-1 at 8. On March 3, 2020, plaintiff requested a meet and confer with defendants, which took place on March 19, 2020, but did not resolve the discovery disputes. See id. According to plaintiff, on April 28, 2020, defendants represented that they were “unwilling to search within the accounts of key decisionmakers for [p]laintiff's name standing alone.” Id. at 9. Rather, plaintiff states that defendants indicated “that they would consider searching only for emails in the decisionmakers [sic] accounts that contained both [p]laintiff's name and a small group o other search terms.” Id. Plaintiff posits that limiting the search in this manner would “obviously would have cut out numerous relevant documents.” Id. On May 7, 2020, plaintiff states that he “agreed to modify the search request for the custodian accounts of the key decision makers . . . to comply with [dJefendants’ request for the time period of January 1, 2013 to November 30, 2017, so that emails during this period would be reviewed only if they contained both [p]laintiff's name and another relevant search term.” Dkt. No. 23-1 at 10. However, plaintiff “asked that the search terms be run individually in his account and in the narrow two[-]month time- period of December 1, 2017 to January 31, 2018 in the key decision makers’ accounts,” which he states “was the key two[-]month time period leading up to [his] termination,” but that “[djefendants objected and refused to run the search.” Id. at 9-10. Thereafter, at a May 13, 2020 meet and confer, “[djefendants offered to provide the quarterly reports from January 1, 2013 to December 31, 2018 but not the weekly report.” Id. at 10. However, plaintiff states that “[djefendants evaluated the sales department monthly,

not quarterly, so at a minimum[,] monthly reports should be produced.” Id. Moreover, plaintiff posits, “to date not even the quarterly reports, which were promised, have been produced.” Id. In a June 5, 2020 teleconference, plaintiff states that defendants informed him that he “would [either] need to agree to eliminating from the search all names, including [p]laintiffs own name,” which plaintiff contends “would certainly exclude large numbers of relevant emails,” or, in the alternative, use “predictive coding.” Id. at 11. Later that month, plaintiff posits that “[djJefendants unilaterally limited the search set by suggesting that they pull a highly narrow sub-set of emails and claiming to pursue predictive coding only within this overly narrow set,” rather than using predictive coding to search the entire custodian accounts of the individuals who made the decision to terminate his m| employment, “as is standard.” Id. (parenthesis omitted). Defendants further proposed to limit the predictive coding search to “only email sent from one of four custodians to another of these same four custodians.” Id. at 11-12. Plaintiff argues that “[t]his sub-se appeared to have been designed to purposefully exclude the vast majority of potentially responsive emails” and “would almost certainly eliminate very relevant emails such as emails to other members of human resources (“HR”) or other managers apprising them Of key details related to this case.” Id. at 11, 12. Moreover, plaintiff states that, during a June 30, 2020 meet and confer, the parties reached an agreement concerning the use of “predictive coding to search the full custodian accounts of the four decisionmakers and [p]laintiff[,] as is industry standard.” Id. at 12-13 (parenthesis omitted). However, plaintiff avers, “[djefendants suddenly revoked their agreement on July 7", 2020,” and

have since “refused to conduct meaningful ESI searches and have produced no further ESI.” Id. at 13. On August 26, 2020, a conference was held before the undersigned, wherein, plaintiff contends, “[djefendants proposed the use of predictive coding on the 27,000 hit from the latest search proposal rather than the entire custodian accounts of the decisionmakers as is standard.” Dkt. No. 23-1 at 13 (parenthesis omitted). Plaintiff posits that the “proposed search is inadequate as the 27,000 hits are already narrowed to the most relevant time periods and in order for predictive coding to be accurate it must be applied to a larger set of documents.” Id. Plaintiff states that, in order to reach a compromise, he “willing to remove auto-generated emails from the search to reduce the number of hits” and states that he “has agreed to every narrowing of the search m| terms [djefendants have proposed except the elimination of names run on their own during a narrow two-month period (December 2017 through January 2018).” Id.

2. Plaintiff's Argument in Support of Motion to Compel Plaintiff states that, through his present motion, he “seeks to compel [djefendants to run limited additional searches for highly relevant documents.” Dkt. No. 23-1 at 15. particular, plaintiff contends that, contrary to defendants’ proposal which seeks to eliminate from the list of search terms plaintiff's and the individual defendants’ names, “[s]earching the email accounts of the officials who decided to terminate [p]laintiff for [his] own name is highly relevant to this case.” Id. at 16. Plaintiff also urges that the search of his “name in a narrow two month period (December 1, 2017 to January 31, 2018) and searching [p]laintiff's email account of the names of the key decision makers

for the same two month period (and the name of HR for a longer period as [p]laintiff communicated less frequently with HR)” will not, as defendants contend, “produce[] results that are too voluminous to review.” Id. at 17. Further, plaintiff contends that the use of predictive coding “greatly decreases the burdensomeness of searching large numbers of documents”; therefore, plaintiff avers, defendants attempt to limit their search to emails sent between decisionmakers “appears to be specifically intended to exclude highly relevant documents sent or received by one of the decisionmakers that are directly relevant to [p]laintiff's performance or firing, as well as the potential biases and discriminatory motives of that decisionmaker.” Id. Moreover, plaintiff avers, the use of predictive coding, the elimination of auto-generated emails, and de-duplication will allow defendants to search m| the entire email accounts of the decisionmakers without limiting the search to the 27,000-email set identified by defendants. See id. at 18.

3. Plaintiff's Proposed Methods for Producing the Requested ESI Based on the foregoing, plaintiff proposes using the following methods to produce the requested ESI: [C]onduct a search for responsive documents encompassing all emails that contain Mr. Maurer’s name (specifically the search terms “!Charlie!”; “!Charles!’; and “IMaurer!”) that were sent or received from December 1, 2017 to January 31, 2018 within the custodian email accounts of the key decision makers in Mr. Maurer’s termination: Thomas Teal, Craig Wittenberg, Lynn Harris, and Mike Pilkington; [Conduct a] search [of] Mr. Maurer’s email account of “Craig!”; “Wittenberg!”; “Tom!”; “Thomas!”; and “Teal!” from December 1, 2017 to January 31, 2018;

[Conduct a] search [of] Mr. Maurer’s email account for or “Lynn!”; “tHarris!”; “disable!”; “ADHD!”; “depress!’; “anxiety!”; “anxious!”; “therapist!”; “counsel!”; “racing!”; “racing thought!”; “HumanResource!”; “complain!”; “concern!”; “EEOC!”; “Equal Employment Opportunity Commission!”; “NYSDHR!”; “NYDHR!’; “New York State Division of Human Rights!”; “litig!’; “illegal!”;“warn!’; “teminat!”; “bias!”; “improper!”; “discriminat!”; and “retaliat!” 5 from January 1, 2013 to present. Dkt. No. 23 at 1-2; Dkt. No. 23-1 at 3-4. Alternatively, plaintiff request that defendants “[c]onduct industry standard predictive coding searches of the full custodian email accounts of the key decision makers in Mr. Maurer’s termination (Thomas Teal, Craig Wittenberg, Lynn Harris, and Mike Pilkington) and Mr. Maurer’s own email account from January 1, 2013 to present.” Dkt. No. 23 at 2; Dkt. No. 23-1 at 4.

I B. Defendants’ Memorandum of Law in Opposition to Motion to Compel 1. Defendants’ Recitation of Relevant Procedural History and Negotiations Concerning Disclosure of ESI and Sales Reports Data Defendants state that, in response to plaintiff's November 4, 2019 discovery demands, they provided their initial responses and objections and “endeavored to work with [p]laintiff to narrow the scope of many of the objectionable [rJequests and to establish an ESI protocol that would be workable for both parties.” Dkt. No. 24 at 7. On April 7, 2020, defendants state that they informed plaintiff that the proposed search terms “had been run and had resulted in a wildly disproportionate number of ‘hits.” Id. In particular, defendants noted that the proposed search terms yielded 269,205 hits, and stated that a review of that many emails “would far exceed the scope of proportional discovery but offered to work with [p]laintiff to narrow them.” Id. After a

meet and confer in which plaintiff agreed to “limited reductions,” defendants re-ran the search with the newly modified search terms, which still yielded over 50,000 hits. See id. Defendants urged that the parties would need to further “limit the search much further [because they were]n’t going to agree to an overly broad, burdensome and costly ESI document production in a single plaintiff case.” Id. (internal quotation marks and citation omitted). For example, defendants posit that a search of the “‘Charles’ alone returned approximately 15,490 hits” and a search of the name “Craig” yielded “approximately 35,266 hits,” even when limited to the period covering December 1, 2017 through January 31, 2018—the ‘limited] period that [p]laintiff agreed to have searched for names alone.” Id. at 8. Defendants expressed to plaintiff that the foregoing search results “was the reason for proposing that name-only ‘hits’ be eliminated or that only ‘name plus’ hits (i.e. hits that contained both a name and a relevant search term) be included.” Id. Defendants posit that plaintiff responded by proposing “largely redundant” search terms that “contained only trivial term and date range limitations.” Dkt. No. 24 at 8. Defendants responded by demonstrating to plaintiff, through a spreadsheet that his proposed search would require the review of approximately 263,769 emails, which defendants urged “would take an enormous amount of time and money and [wa]s . . . not proportional to the needs of the case.” Id. (internal quotation marks and citation omitted). Defendants also informed plaintiff that “a slightly narrowed version of the search terms was also attempted but that it too returned a disproportionate number of responses.” Id.

As an alternative, defendants proposed running a revised search for the period between May 10, 2016, through May 10, 2018, excluding individual names, and using the following search terms for plaintiff's email: “disabl!”; “ADHD!”; “depress!”; “anxiety!’; “anxious!”; “leave!”; “therapist!”; “counsel!”;“racing!”; “racing thought!”; “complain”; “concern!”; “EEOC!”; “Equal Employment Opportunity Commission!”; “NYSDHR"; ° “NYDHR!”; “New York State Division of Human Rights!”; “litig!”; “illegal!’; “warn!”; “teminat!”; “bias!”; “improper!”; “discriminat!”; “retaliat!”; “fail!!”, “below target”. Dkt. No. 24 at 8. Defendants also ran a search of the following terms for the period between May 10, 2016, through May 10, 2018, for “the remaining custodian’s accounts”: “disabl!”; “ADHD!”; “depress!”; “anxiety!”; “leave!”; “therapist!”; “racing!”; “racing thought!”; “EEOC!”; “Equal Employment Opportunity Commission!”; “NYSDHR’; m| “NYDHR!”; “New York State Division of Human Rights!”; “litig!”; “illegal!”; “unfair!” Id. Defendants then ran this search, which “returned a universe of approximately 27,000 hits.” Dkt. No. 24 at 8. Defendants informed plaintiff that “this was still too many emails to review and proposed reviewing a random sampling of 500 hits and then circling back to determine how best to proceed from there.” Id. (internal quotation marks and citation omitted). Defendants posit that, “[o]f the 500 hits reviewed, 11 were marked relevant.” Id. at 9 (internal quotation marks and citation omitted). Defendants also informed plaintiff that the results demonstrated that “there [wa]s not going to be a great deal of ESI in this case” and proposed narrowing the search further. Id. In response, plaintiff advanced the position that review of the approximately 27,000 emails was “necessary and proportional to th[e] case.” Id. Defendants state that, at this point, it was clear an impasse had been reached, so they proposed the alternative method of

using predictive coding. See id. Defendants then set forth proposed parameters for the use of predictive coding in a June 11, 2020 communication in which they made clear that they were “not conceding that it is proportional to the needs of the case” and that they were working with Sysco and “the vendor to determine the cost of hosting the full mailbox[es] of all five custodians,” which, if “prohibitively high,” would require further o narrowing. Id. Defendants state that, after working with Sysco and an ESI vendor, they determined that “the cost of hosting all five inboxes” from January 1, 2013, to the present would “be prohibitively high.” Dkt. No. 24 at 9. Defendants indicate that “hosting (i.e. storage) off] the enormous amount of data contained in the five custodians’ mailboxes would have cost an estimated $23,755 per month—before review costs [We]re even factored in.” Id. Defendants calculate that the cost of review using predictive coding with contract attorneys would yield a sum of “well over $200,000.” Id. Defendants communicated to plaintiff that the proposed method was prohibitively expensive and, therefore, further modification would be necessary. See id. Defendants proposed that “review of the approximately 27,000 [hits] that resulted from the agreed upon search terms [wa]s a good place to pick up” discussions for finding a more cost- permissible method of searching. Id. Contrary to plaintiff's suggestion, defendants contend, this was not a “sudden reversal” of any agreement between the parties as to how to produce the requested ESI or done for the purpose of delaying discovery; rather, defendants state, they were merely following up and communicating to plaintiff, as they had told plaintiff they would, to explain that, after testing the proposed ES! search

method, it was prohibitively costly and that a different, more cost-effective manner of searching was necessary. Id.

2. Defendants’ Argument in Opposition of Motion to Compel Defendants contend that plaintiff misapprehends the standard applicable to motions to compel insofar as plaintiff's argument ignores the need that discovery of the ESI at issue be proportional to the needs of the case. See Dkt. No. 24 at 11. First, defendants posit, plaintiff's proposed time period for searches of the ESI from January 1, 2013 is overbroad and not relevant or proportional to the claims and defenses involved. See id. Defendants point out that plaintiff allegedly disclosed his disability in December 2017, and, state that, although “e-mails demonstrating [p]laintiff's poor sales m| performance or mismanagement of his team may have originated” prior to December 2017, “[djefendants have not alleged that they relied on any actions [p]laintiff took as long ago as 2013 weighed into the determination to terminate.” Id. Next, defendants contend that the elimination of name-only searches is necessary to eliminate large numbers of “false positives.” Dkt. No. 24 at 12. Defendants explain that, because “names are often included in signature lines and are always included in Sysco email addresses,” a search of “someone’s name effectively pulls every single e-mail they sent or received within the timeframe searched[]” and “[t]here is nothing tailored about that process.” Id. Therefore, defendants posit, searches implementing plaintiff's or one of the decisionmakers’ names must also include another relevant search term. See id. Thus, defendants suggest that the two- year window of January 5, 2016, to January 5, 2018 (the date of plaintiff's termination)

be searched, excluding the use of name-only searches or the use of the search term “Human Resources!” because that term is overbroad, “but do not take issue with any of [p]laintiffs other proposed search terms.” Id. Defendants further assert that the benefits of using plaintiff's proposed methods for searching the ESI are far outweighed by the high costs involved in the storage and review of such large amounts of data. See Dkt. No. 24 at 13-14. Defendants note that “[p]laintiff's disclosure and alleged request for leave were made verbally,” and aver that “ESI is not especially critical to this case.” Id. at 14. In any event, defendants contend, “any ... relevant ESI relating to [p]laintiff's discrimination and retaliation claims is almost certain to have been generated during a three-week window spanning from mid- December 2017 to Early January 2018” between the time he disclosed his disability to m| Wittenberg and his January 5, 2018 termination. Id. Indeed, defendants suggest, “[a]n e-mails discussing that request or alleged retaliation stemming from it would inevitably have been exchanged in the time between the alleged disclosure (December 18, 2017) and [pllaintiff's termination (January 5, 2018).” Id. Therefore, defendants urge, plaintiff's request for ESI dating back to 2013—a period of more than seven years—is overbroad and seeks irrelevant information. See id. In addition, defendants posit, their review of the random 500-email sampling taken from the approximately 27,000 email that resulted from the previously agreed upon search terms, indicates that only a minimal percentage—2.2% (11 of 500 emails)—were “relevant.” Id. With respect to costs, defendants posit that plaintiff's proposed methods of “running searches and conducting a linear review” or searching “each custodians’ full mailbox” using predictive coding would result in storage and contract attorneys’ fees in

excess of $150,000 plus an additional $1,000/month for the former; and $150,000 in storage fees, $74,050 in contract attorneys’ fees, and additional legal fees for review by defense counsel, resulting in costs in excess of $200,000 for the latter. See id. at 13- 14. Thus, defendants contend, given the likely minimal importance of ESI to the claims and defenses involved in the present matter, the costs are disproportional the potential benefits. See id. at 14. In addition, defendants aver that plaintiff's demand for monthly or weekly sales reports is cumulative and disproportional to the needs of the case. Dkt. No. 24 at 16. Defendants note that they “have already produced annual sales reports for fiscal years 2015, 2016, 2017 and 2018 and performance evaluations for [p]laintiff and comparators,” and “agreed to produce quarterly sales reports from January 2015 m| through January 2018, but have been delayed in doing so because of furloughs resulting from the COVID-19 pandemic.” Id. at 17. In particular, defendants argue that information contained in the weekly and monthly sales reports “is necessarily reflected by the slightly less frequent reports that [dJefendants have already produced or agreed to produce” and, therefore, “[t]he production of additional reports would be cumulative and is therefore unnecessary.” Id. Moreover, defendants aver that “[p]laintiff's sales m| from 2013 did not factor into Sysco’s decision to terminate him in 2018. Nor did his sales in any particular month.” Id. Rather, defendants state, “Sysco, like any successful company, takes a broader view of sales performance.” Id. Thus, defendant contend, “[p]laintiff's insistance that monthly reports be produced in addition to quarterly and annual reports is also overbroad for this reason.” Id. Based on the foregoing, defendants propose producing for plaintiff “quarterly sales reports for the period of

January 5, 2016 through January 5, 2018,” which when “paired with the annual sales data [djefendants have already produced, should be deemed sufficient to meet [djefendants’ obligations with respect to performance and comparator data.” Id. at 18.

3. Defendants’ Proposed ESI Protocol o Defendants propose the following method for searching and producing relevant ESI: 1) Narrow the existing universe of approximately 27,000 documents by eliminating any “hits” that were triggered only because they contained a custodian’s name and by imposing a date range of January 5, 2016 through January 5, 2018; 2) Undersigned counsel reviews a statistically significant sample of the remaining e-mails at issue and marks them relevant/irrelevant to create a “training set;” 3) That training set is then used to “train” the eDiscovery vendor's artificial intelligence/predictive coding tool, which “reviews” the remaining e-mails and assigns each a percentage-based score that measures likelihood to be responsive; 4) Contract attorneys are retained through the eDiscovery vendor, briefed on the facts of the case by undersigned counsel, and begin their first level review of the remaining e- mails—starting with those most likely to be responsive; tr! 5) Contract attorneys provide daily updates to the undersigned counsel regarding the percentage of e-mails being marke[d] relevant. If there comes a time when almost no relevant e-mails are being found after a certain percentage of “likelihood to be responsive” has been reached or disproportionate costs have been incurred, the review will be ended. 1. If this is done, undersigned counsel will review a statistically significant sample of any remaining e-

mail to ensure that they are indeed irrelevant/non- responsive. 6) In the alternative, contract attorneys will complete their first-level linear review of all “hits”; 7) Undersigned counsel conducts second-level review of all documents marked relevant;

5 8) Any documents still marked relevant are produced. Dkt. No. 24 at 15-16.

lil. Legal Standard Under Federal Rule of Civil Procedure 26, “[pJarties may obtain discovery regarding any nonprivileged matter that is relevant to any party’s claim or defense and proportional to the needs of the case.” FED. R. Civ. P. 26(b)(1). Rule 26 tasks the Court to consider “the importance of the issues at stake in the action, the amount in controversy, the parties’ relative access to relevant information, the parties’ resources, the importance of the discovery in resolving the issues, and whether the burden or expense of the proposed discovery outweighs its likely benefit.” Id. “A district court has broad latitude to determine the scope of discovery and to manage the discovery process.” EM Ltd. v. Republic of Argentina, 695 F.3d 201, 207 (2d Cir. 2012) (citing In re Agent Orange Prod. Liab. Litig., 517 F.3d 76, 103 (2d Cir. 2008)).

IV. Analysis A. ESI As an initial matter, with respect to the ESI search procedure, the Court finds reasonable the defendants’ proposed use of a general search-term-based search of the

custodian accounts, similar to that previously performed which yielded the 27,000-email universe, to be followed by a second level of review using predictive coding based on the use of a “training set” generated from the review of a “statistically significant sample” set of emails. See Dkt. No. 24 at 15. Concerning the time period for ESI discovery, the Court agrees with plaintiff that defendants have not provided clarity as to the time period used for their purported decision to terminate him based on failure to meet sales goals or adequately manage his sales team. See Dkt. No. 23 at 16; Dkt. No. 24 at 11. However, plaintiff has not established that ESI dating back to January 2013—over seven years before plaintiffs employment at Sysco was terminated—is relevant to the claims and defenses presented. Defendants have stated that “[p]laintiff's sales from 2013 did not factor into Sysco’s decision to terminate him in 2018. Nor did his sales in m}any particular month.” Dkt. No. 24 at 17. Despite plaintiff's indication that he began working at Sysco sometime in 2003, plaintiff has not set forth any facts or evidence to establish that ESI dating back to 2013 “would provide unique, relevant and noncumulative evidence[.]” Coventry Cap. US LLC v. EEA Life Settlements Inc., No. 17-CV-7417 (VM/SLC), 2020 WL 7383940, at *7 (S.D.N.Y. Dec. 16, 2020), objections overruled, No. 17-CV-7417 (VM), 2021 WL 961750 (S.D.N.Y. Mar. 15, 2021). In fact, the alleged basis of plaintiff's claims—his disclosing his purported disabilities on December 18, 2017 to Wittenberg—occurred over four years after the January 1, 2013 date plaintiff seeks to use as the look-back date for ESI. See Compl. at 4 Jf 18, 20-23. Thus, the timeframe for the initial search term search of ES! will be May 10, 2016, through May 10, 2018—the time period originally used by defendants that yielded the approximately 27,000 email universe. See, e.g., Obiajulu v. City of Rochester, 166

F.R.D. 293, 296 (W.D.N.Y. 1996) (“Courts have exercised discretion in discrimination cases by placing reasonable limits on the time periods for which discovery must be produced.”). The Court notes that defendants have already conducted a search of ESI between May 10, 2016, through May 10, 2018, without complaining of any undue burden or cost. See Dkt. No. 24 at 8. ° Further, as defendants aver, and as the analysis of the random 500-email sample within the 27,000 emails generated by their previous search demonstrates, while potentially important, the ESI relevant to the claims and defenses in this case is likely not voluminous. See Dkt. No. 24 at 14. Indeed, of the 500-email sample defendants reviewed, only 11 were deemed relevant. See id. By contrast, defendants have established that the review procedures proposed by plaintiff would result in Storage and review costs that would be disproportionate to the likely limited amount of relevant ESI that will be discovered. See id. at 13-14. Therefore, the Court concludes that the cost of conducting a linear review of every hit resulting from a search term- based search that includes all custodians’ names and name derivatives or reviewing the full custodian accounts using predictive coding dating back to 2013 is not proportional t the benefit and importance of ESI in resolving the issues presented in this case. See m| Fed. R. Civ. P. 26(b)(1). Concerning the specific search terms, although nothing before the Court supports plaintiffs conclusory assertion that defendants manipulated their prior search for the purpose of excluding relevant ESI, as defendants acknowledge, their search did not include plaintiffs name or the names of the four decisionmakers. See id. at 8. The Court agrees with plaintiff that the omission from his name in the search of the

decisionmakers’ email accounts could omit relevant documents; therefore, defendants’ initial search-term-based search of the ESI shall include a name-only search of the decisionmakers’ email accounts for plaintiff's name, including such derivatives thereof as the parties agree upon, but excluding hits resulting from auto-generated emails, email addresses, or signature lines, for the limited time period proposed by plaintiff of December 1, 2017, through January 31, 2018. See Dkt. No. 23 at 1; Dkt. No. 23-1 at 3. Indeed, as defendants acknowledge, “any . . . relevant ESI relating to [p]laintiff's discrimination and retaliation claims is almost certain to have been generated during a three-week window spanning from mid-December 2017 to Early January 2018” and “[a]ny e-mails discussing [his FMLA] request or alleged retaliation stemming from it would inevitably have been exchanged in the time between the alleged disclosure m| (December 18, 2017) and [p]laintiff's termination (January 5, 2018).” Dkt. No. 24 at 14. The Court concludes that a name-only search, as described above, for this short period of time will not result in an unduly burdensome number of emails for defendants to review, especially since defendants will be able to further filter out irrelevant hits through the use of predictive coding. Thus, insofar as plaintiff's motion seeks to compel discovery of ESI via a protocol that includes a name-only search of plaintiff's name, and agreed-upon derivatives thereof, in the decisionmakers’ email accounts for the period of December 1, 2017, through January 31, 2018, but excluding hits resulting from auto-generated emails, including email addresses and/or signature lines alone, plaintiffs motion is granted. See Dkt. No. 23 at 1; Dkt. No. 23-1 at 3-4. However, to the extent that plaintiff's motion seeks to compel discovery of ESI via a name-only search of plaintiff's Sysco email

account using the decisionmakers’ names, the motion is denied. See id. Based on the volume of hits generated by name-only searches using the decisionmakers’ names, see Dkt. No. 23 at 8, the Court concludes that defendants’ have demonstrated that their proposal to use the “name plus” search criteria for the initial search-term-based search of the decisionmakers’ names in plaintiff's Sysco email account, excluding auto- o generated emails and hits resulting solely as a result of an email address or signature line, is more proportional to the discovery needs of this case. Id. In addition, although defendants posit that the inclusion of the search term “Human Resources!” is overly broad, see Dkt. No. 24 at 12, defendants have not set forth arguments or evidence to support this position; therefore, that search term will be included in the list of search terms. The Court concludes that plaintiff has established that this search term is relevant to the claims and defenses presented, as Harris— Sysco’s vice president of human resources, was allegedly involved in Sysco’s decision to terminate plaintiff's employment. See Compl. at 6 4 35. Any irrelevant hits resulting from the search of the term “Human Resources!” may be filtered out during the second level of review using predictive coding. Thus, to the extent that plaintiff's motion seeks to require that the search term “Human Resources!” be incorporated into defendants’ initial search-term-based search, the motion is granted. As courts in this circuit have stated, “the best solution in the entire area of electronic discovery is cooperation among counsel,” Bagley v. Yale Univ., 307 F.R.D. 59, 63 (D. Conn. 2015) (quoting William A. Gross Constr. Assocs., Inc. v. Am. Mfrs. Ins. Co., 256 F.R.D. 134, 136 (S.D.N.Y.2009)). “Ideally, the parties should agree on the search methods, including search terms or concepts.” Id. (internal quotation marks and

citation omitted). Thus, the parties are directed to submit a final ESI protocol consistent with the conclusions set forth in this Decision & Order for the Court's approval within 30 days of the date thereof.

B. Sales Reports o Defendants have already produced annual sales reports for fiscal years 2015- 2018, as well as performance evaluations for plaintiff and comparators. See Dkt. No. 2 at 16-17. Further, defendants have agreed to produce quarterly sales reports from January 2015, through January 2018, but posit that they have been hindered in their efforts to do so because of the COVID-19 pandemic. See id. Although, as defendants state, the sales data included in the annual report would necessarily also be included in quarterly or monthly reports, defendants do not submit any evidence or advance any compelling arguments as to why they are amenable to providing quarterly sales data while simultaneously contending that the provision of monthly data “would be cumulative and therefore unnecessary.” Id. at 5. Further, defendants do not controvert plaintiff's statement that Sysco evaluates district managers’ sales performance on a monthly basis. See Dkt. No. 23-1 at 10; Dkt. No. 24 at 5, 16-17. Thus, as the parties not dispute that Sysco evaluates sales performance on a monthly basis, not quarterly or otherwise, the Court concludes that the monthly—rather than quarterly— sales reports for plaintiff and comparators are relevant. Based on the parties’ representations in this regard, the monthly sales reports for fiscal years 2015-2018 will be readily available and easy to produce given that Sysco uses the monthly sales data to evaluate sales department, and defendants have advanced no arguments or offered

any evidence to establish that such production would be unduly burdensome or disproportional. Accordingly, plaintiff's motion is granted to the extent that he seeks to compel defendants to produce the monthly sales reports of plaintiff and comparators, fo fiscal years 2015-2018 within thirty (80) days of the date of this Decision & Order.

V. Conclusion WHEREFORE, for the reasons stated above, it is hereby: ORDERED, that plaintiff's motion (Dkt. No. 23) is GRANTED IN PART insofar as plaintiff seeks to: a) Compel the inclusion of a name-only search of his name, and derivatives thereof that the parties can agree upon, into defendants’ initial search term-based search, excluding any hits that result solely from auto- generated emails, plaintiff's email address, and/or plaintiff's signature line; b) Compel the inclusion of the search term “Human Resources!” into defendants’ initial search term-based search; and c) Compel defendants to produce the monthly sales reports for plaintiff and comparators for the fiscal years 2015, 2016, 2017, and 2018; and it is further ORDERED, that plaintiff's motion is otherwise DENIED; and it is ORDERED, that the parties submit for the Court’s approval a final ESI protocol consistent with the conclusions set forth in this Decision & Order within thirty (30) days of the date thereof; and it is further

ORDERED, that defendants produce the monthly sales reports for plaintiff and comparators for the fiscal years 2015, 2016, 2017, and 2018, within thirty (80) days of the date of this Decision & Order; and it is further ORDERED, that copies of this Decision and Order be served on the parties in accordance with the Local Rules. ° IT IS SO ORDERED. Dated: May 27, 2021 Albany, New York

Christian F. Hummel U.S. Magistrate Judge

I