*1 By Order dated September 30, 2020 (Doc. 108), Chief Judge Phillips appointed the undersigned as Special Master and authorized her, inter alia, to evaluate “the parties' arguments and objections with respect to the various discovery disputes and submit a report and recommendation on those objections.” (Doc. 108). This Report and Recommendation addresses the current disputes between Plaintiff Kemps LLC (“Kemps”) and Defendant IPL USA, Inc. (“IPL”) over the date range of electronically stored information (ESI) to be searched and the search terms each must use as part of their ESI discovery protocol. For reasons set forth below, the undersigned recommends that the Court:

The undersigned recommends that the Court deny IPL's other search term requests not enumerated above.

Pursuant to the Court's September 30, 2020 Order, the parties have fourteen days from the date of filing of this Report and Recommendation in which to file objections.

Kemps makes frozen dairy food products. IPL manufactures injection-molded plastic products, including frozen dairy food containers. In September 2019, Kemps filed this lawsuit against IPL. (Doc. 1). Kemps' operative complaint asserts claims for breach of express warranty (UCC Art. 2-313), breach of implied warranty of merchantability (UCC Art. 2-314), breach of implied warranty of fitness for a particular purpose (UCC Art. 2-315), breach of indemnity agreement, and breach of written terms and conditions for sale. (Doc. 55). All of Kemps' claims arise out of Kemps' allegations that IPL provided defective specialized injection-molded clarified polypropylene containers that would break at unacceptably high rates and cause some consumers to report sharp plastic pieces in their Kemps dairy food containers. (Doc. 55, p. 2). The allegedly defective containers were clear, one-pint ice cream/frozen yogurt containers used for Kemps' SweetMe and Yo2 ice cream and frozen yogurt products. Kemps voluntarily recalled its products packaged in the containers. On April 18, 2017, it sent IPL written notice of a claim. Kemps calculates its damages to be around $3 million and increasing. (Kemps' Submission No. 2 to Special Master, p. 3, 10).

*2 IPL denies the claims and counterclaims for breach of contract for nonpayment of some of the containers. (Doc. 60). According to IPL, in 2016, Kemps requested sample containers that IPL had successfully developed for Safeway's frozen dairy products. Kemps tested and validated the containers without IPL's involvement. Between August 2016 and March 2017, IPL supplied Kemps with some 2.7 million containers. After the voluntary recall, Kemps switched to a different IPL model container designed to withstand rougher handling. Kemps disputes IPL's characterizations.[2]

Initially, the Court entered a scheduling order that imposed a July 1, 2020 deadline for discovery and set a jury trial for February 1, 2021. (Doc. 29). On February 18, 2020, the Court granted the parties' joint motion to extend discovery and certain other deadlines in the scheduling order. (Doc. 59). On April 22, 2020, the Court granted the parties' joint request to further extend expert disclosure deadlines. (Doc. 73). On June 16, 2020, the Court held a telephone conference on a discovery dispute and ordered both parties to respond to certain topics raised in the preconference letters. (Doc. 92). The Court issued an amended scheduling order on July 8, 2020. (Doc. 100).

On July 31, 2020 and September 16, 2020, the Court held additional discovery-dispute teleconferences. (Doc. 102 & 105). During the latter teleconference, the Court advised the parties of its intent to appoint a Special Master. (Doc. 105).

After being appointed, the undersigned held Zoom meetings with, and obtained written information from, the parties. Early interactions revealed the parties to be at an impasse over several ESI disputes, yet in agreement about a general ESI protocol. Under the agreed protocol, ESI would be collected by each side from certain custodians. Then, each party would run search terms against the cache of electronic documents that party had collected. Next, the set of “hits” generated by the search terms would be reviewed and produced based on relevance to the claims and defenses in the case.[3]

Both parties appear to have believed in good faith that the ESI protocol would be less burdensome than other approaches. But agreement on search terms and date ranges eluded them. Discussions continued over several months, leading to a breakdown in productive communications and a “stalemate” on issues. Both parties began running some, but not all search terms against their own ESI.

*3 In early October 2020, the parties appeared unable to succinctly identify which search terms were currently disputed and which were not. On October 6, 2020, the undersigned requested each party to specifically list (1) each of the search terms that the party was currently running for its ESI production and identify each custodian whose ESI is being searched with those terms and date range of the search, and (2) each of the search terms that party contended the other party should run as part of its ESI production, each custodian of the other party whose ESI the party contends should be searched with those terms and the date range of the search. The parties complied and submitted the requested information by October 7, 2020.

Thereafter, the undersigned compiled and provided the parties charts that sought to summarize the apparent areas of ESI dispute and agreement. The summary for Kemps' ESI is attached as Exhibit 1. The summary for IPL's ESI is attached as Exhibit 2. After both sides confirmed agreement with the summaries, the undersigned permitted the parties to fully brief their respective positions for ruling.

A few disputes were eliminated by briefing concessions.[4] A dispute remains over whether IPL should search ESI dated after Kemps sent IPL its written notice of claim. IPL objects to searching past April 18, 2017. Kemps contends IPL should search and produce ESI dated April 19, 2017 through September 18, 2019. Kemps narrowed from 16 to 10 the number of IPL custodians whose ESI should be searched for this period. IPL suggests that if a search is necessary, it should be limited to two IPL custodians.

Remaining issues stem from search term disputes. For IPL's ESI, the parties agree to more than 60 search terms. Kemps contends that IPL should use three more terms: (1) Yo2 or Y02, (2) SweetMe or “Sweet Me”, and (3) Clear AND (Result OR Test). IPL objects on proportionality grounds.

For Kemps' ESI, the parties agree to more than 60 search terms. IPL contends that Kemps should run multiple additional search terms. These disputed terms can be broken into six categories for purposes of analysis. They are (1) the “pint proximity terms”[5], (2) the “FDA terms”[6], (3) the “product terms”[7], (4) the “result/test terms”[8], (5) the “resin terms”[9], and (6) IML.[10] Kemps objects on proportionality grounds. Both parties make accusations of delay and gamesmanship against each other in connection with their long-running ESI search-term discussions.

A party “may obtain discovery regarding any non-privileged 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); Haukereid v. National R.R. Passenger Corp., 816 F.3d 527, 534 (8th Cir. 2016). “Proportionality is weighed by considering ‘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.’ ” Shannon v. Honeywell Fed. Manufacturing & Technologies, LLC, 2018 WL 8244890 at *1 (W.D. Mo. July 20, 2018) (quoting Fed. R. Civ. P. 26(b)(1)). Proportionality determinations are made on a case-by-case basis, and no single proportionality factor is designed to outweigh the others. Oxbow Carbon & Minerals LLC v. Union Pacific Railroad Company, 322 F.R.D.1, 6 (D.D.C. 2017).

*4 When parties dispute the propriety of discovery, the party seeking the discovery must make a threshold showing of relevance. Fair v. Communications Unlimited, Inc., 2018 WL 2238177 at *2 (E.D. Mo. May 16, 2018); Hofer v. Mack Trucks, Inc., 981 F.2d 377, 380 (8th Cir. 1992). After this threshold showing, “the burden is on the party resisting discovery to explain why discovery should be limited.” Nwinee v. St. Louis Developmental Disabilities Treatment Ctrs, 2019 WL 2569927 at *1 (E.D. Mo June 21, 2019); Eramo v. Rolling Stone LLC, 314 F.R.D. 205, 209 (W.D. Vir. 2016). “The parties and the court have a collective responsibility to consider the proportionality of all discovery and consider it in resolving discovery disputes.” Fed. R. Civ. P. 26(b)(1) advisory committee's note to the 2015 amendment.

A. Time Frame Dispute

Kemps argues that IPL should search ESI dated April 19, 2017 through September 18, 2019 for 10 custodians. Relevant ESI exists during that time period, Kemps contends, because Kemps' withdrawal efforts were still underway, the parties continued to work together on the disputed packaging for a period of time, and circumstances were conducive to IPL having rearward-looking internal discussions and analysis that may contain probative admissions valuable to Kemps' claims or be revealing of product failures. Kemps states that the 10 requested custodians are management or staff in the areas of customer or sales contacts, quality control, or engineering, all with direct involvement with Kemps or other purchasers of the same containers. In Kemps' view, these persons are most likely to have had discussions analyzing the product failures, shortcomings, or necessary changes after the fact.

IPL points to Kemps' operative complaint and suggests that all conduct alleged to give rise to Kemps' causes of action occurred before April 18, 2017. IPL states that if a search after April 18, 2017 is required, the custodians should be limited to two persons who were primarily involved in preparing materials for IPL's insurer in response to Kemps' notice of claim letter. IPL states that running the agreed search terms across 10 custodians will require review of 13,807 additional documents that are unlikely to be irrelevant. IPL has not offered any evidence to quantify relevant versus irrelevant documents. IPL generally requests that the Court order a narrower set of search terms but offers no specific list for the Court's consideration.

The undersigned concludes that Kemps has made a threshold showing of relevance and proportionality that has not been sufficiently refuted by IPL. It is probable that, in the wake of consumer complaints, Kemps' product recall, and Kemps' request for a different IPL container, there were communications within IPL that bear on the central issues in the case of whether the disputed containers were defective or suitable for their intended purpose. Kemps has sufficiently articulated why the 10 custodians are reasonably likely to have been involved in such communications. IPL's counter proposal to limit a search to the two custodians most involved in an insurance response would not capture the types of communications Kemps reasonably seeks.

Other proportionality factors favor Kemps or are neutral. The issues involved in this case are important, involving commercial standards and consumer safety. The amount in controversy— damages exceeding $3 million—is significant. IPL has sole access to its own internal communications, with Kemps' access limited to what it may obtain through discovery. The types of communications Kemps seeks in this time period may be significant in the ultimate resolution of some claims. The parties are companies that are relatively balanced in available resources. According to IPL, running the currently agreed terms across the more recent time period will require review of 13,807 additional documents for relevancy. This volume is not unreasonable in a commercial case of this magnitude. Most of the search terms Kemps requests are those to which IPL has already agreed to use for the earlier time period. IPL advances no persuasive argument that the same terms it agrees will be effective in finding relevant documents during the former time period will not be effective during the latter time period. Moreover, IPL is not required to individually examine each document that produces a search term “hit” but may use other acceptable methods to identify the subset of documents that are relevant and should ultimately be produced to Kemps, such as computer-assisted review or predictive coding. See, e.g., Lawson v. Spirit AeroSystems, Inc., 2020 WL 18113395, *6-*7 (D. Kan. April 9, 2020).

*5 Because Kemps has made a threshold showing and IPL has failed show that the burden or expense outweighs the benefits, the undersigned recommends that the Court order IPL to search ESI for the 10 custodians identified by Kemps for the period April 19, 2017 through September 19, 2019, using the same search terms as those required to be used for January 1, 2015 through April 18, 2017.

B. Additional Search Terms Requested by Kemps

The parties dispute whether IPL must employ three additional search terms that Kemps requests for IPL's ESI. The terms at issue are (1) Yo2 or Y02, (2) SweetMe or “Sweet Me”, and (3) Clear AND (Result OR Test).

Kemps contends that the first two terms are likely to lead to discoverable information because they are the common spellings of Kemps' unique trade names for the two product lines it placed in the disputed containers. Because the names are unique, Kemps argues, every document in IPL's ESI collection with those terms is likely related to the claims or defenses. But Kemps does not detail how this may be so. IPL counters that the terms are overbroad and not tailored to the issues in the case. IPL states that running the first two terms across ESI dating January 1, 2015 through September 18, 2019 generates a total of 21,859 hits, which IPL estimates will cost $152,958.48 to review. The additional cost of running these terms, according to IPL, will exceed its costs for running all agreed search terms during the January 1, 2015 to April 18, 2017 period.

The undersigned concludes that Kemps' initial relevance showing for these terms is weak and that proportionality considerations outweigh the need. Key issues in this case relate to the design and performance of the containers Kemps purchased from IPL. The trade names of the products Kemps placed in the containers have no special import. Moreover, parties have already agreed that IPL will run 60 other search terms that appear better tailored than simple trade name searches to net documents related to the disputed issues. Some 16 of those agreed search terms already include the party name Kemps with a proximity limiter, such as “Kemps” w/12 recall, “Kemps” w/12 break*, etc. Some of the terms include the trade name with a proximity limiter, such “polypropylene” w/6 (Y02 or “Y02”).” Any relevant document that includes Yo2 or Y02, or SweetMe or “Sweet Me” is likely to include a word or phrase already covered by other search terms that are more specific to the issues than simply Yo2 or Y02, or SweetMe or “Sweet Me” alone. In other words, to the extent that these terms are likely to lead to the production of relevant information, that information would merely duplicate information that will be produced by searching other, more specific terms.

Because Kemps has made only a limited showing of relevance and the proportionality factors suggest the burden or expense outweighs the benefits, the undersigned recommends that the Court order deny Kemps' request that IPL use the search terms (1) Yo2 or Y02 or (2) SweetMe or “Sweet Me.”

Kemps contends IPL should run the search term Clear AND (Result OR Test). Kemps' rationale for using this term is a belief that it may capture “evidence of IPL's testing.” (Kemps Submission No. 2, p. 16). Kemps argues that a key theory of its case is that the disputed containers may have been structurally unfit because they are made of “clear” plastic, and the process of clarifying the plastic resin in the containers could have contributed to its brittleness or quality control issues. Kemps notes that IPL originally developed the containers as clarified propylene for a container to be sold to Safeway/Albertson for clear, one-pint ice cream.

*6 IPL argues that the proposed search term is unconnected to any actual test or IPL product at issue in the case. IPL states that it is involved in many different markets—food, consumer, agricultural, logistics, and environmental—and the term results in hits on an overwhelming number of emails not relevant to the parties' claims or defenses. IPL states that running this term across ESI dating January 1, 2015 through September 18, 2019 generates a total of 24,889 hits, which IPL estimates will cost $174,160.93 to review. The additional cost of running this term, according to IPL, will exceed its costs for running all agreed search terms during the January 1, 2015 to April 18, 2017 period.

The undersigned concludes that requiring IPL to search ESI with Clear AND (Result OR Test) and to perform further relevancy review of the hits of that search is not proportional to the needs of this case. While design and performance of the disputed containers is a key issue in the case, Kemps has failed to show the proposed search term is tailored to identifying ESI that is revealing of this issue and not already captured by other, undisputed search terms. The words in the disputed search term—clear, result, test—are common words used in many contexts. There is nothing in the briefing to suggest that IPL conducted a particular test of the containers Kemps purchased or that, if it did, it would have discussed the test or its results solely in the type of non-specific language captured by Clear AND (Result OR Test). Kemps itself describes the material or qualities of the containers using more specific terms, such as “polypropylene,” “clarified resin,” “crystalline structure,” “clarity”, “brittle (less ductile).” Among the 60 search terms IPL is already running, there are several search terms that include the word “propylene” or “resin” or the words “clear or clarified” within proximity to Kemps, YO2, SweetMe, Safeway or Albertson, or other searches that include names of key persons. To the limited extent that Clear AND (Result OR Test) is likely to lead to the production of relevant information, the undersign concludes it will most likely duplicate information that will be produced by searching other, more specific terms.

Because Kemps' showing of relevance is limited and the proportionality factors indicate the burden or expense outweighs the benefits, the undersigned recommends that the Court deny Kemps' request that IPL use the search term Clear AND (Result OR Test).

C. Additional Search Terms Requested by IPL

IPL contends that Kemps should run multiple additional search terms. As noted above, the disputed terms can be divided into six categories for purposes of analysis: (1) the pint proximity terms, (2) the FDA terms, (3) the product terms, (4), the result/test terms, (5) the resin terms, and (6) IML. Kemps argues that IPL's late suggestion of some terms is improper, and none are proportional to the needs of this case because Kemps has already agreed to run many search terms that will capture relevant information.

IPL argues Kemps should utilize the pint proximity search terms because they seek documents related to Kemps' handling and filling and palletization processes used during manufacturing, which IPL believes may have damaged the disputed containers. Kemps contends the terms are overbroad and mostly return documents already within other searches. The pint /12 pall* term, Kemps states, will add 808 new documents to Kemps' review and otherwise returns documents captured by other search terms. The pint /12 fill* term will add 2,519 new documents to Kemps' search and otherwise returns documents captured by other search terms.

*7 The undersigned concludes that requiring Kemps to include the new/unique documents returned by the pint proximity terms as part of its relevancy review is proportional to the needs of the case. IPL has made a threshold showing that these terms seek relevant information. The other search terms Kemps is running do not appear directed at the specific aspects of Kemps' manufacturing process targeted by these terms. The burden on Kemps of reviewing approximately 3,300 new documents returned by the pint proximity terms does not outweigh the benefit of the discovery, which is linked to an important issue in the case.

IPL contends Kemps should run eight search terms that are varying combinations of FDA or “Food /3 Drug” and the terms IPL, pint, recall or break. IPL contends the terms reach relevant documents because they address Kemps' withdrawal and recall of the Sweet Me and Yo2 products and related decision making and communications. Kemps responds that IPL originally proposed FDA and “Food /3 Drug” as standalone searches, which Kemps contends are facially overbroad and would unreasonably increase Kemps' ESI search burden by some 140,000 documents, as all of its products are FDA regulated. In July, Kemps offered to search FDA with the 12-word proximity limiters Yo2 or “Sweet Me” and is running those searches now. Kemps argues that IPL's most recent permutations of the FDA terms are duplicative, cumulative, or reach irrelevant information.

The undersigned finds that two of the disputed FDA terms are proportional to the needs of the case and the remainder are not. The search of FDA or “Food /3 Drug” combined with IPL together adds 173 new documents to Kemps' review. Each of these documents is likely related to the facts at issue because they reference both FDA and IPL. Kemps, as far as the record suggests here, has issued only one FDA recall for a product with an IPL container. Kemps has offered no factual showing that the proximity limiter search it is currently running nets most of the relevant documents associated with Kemps' recall of the products packaged in IPL's containers. The limited burden on Kemps of reviewing an additional 173 documents for relevancy does not outweigh other proportionality factors.

The remaining terms that combine FDA or Food /3 Drug and the words pint, recall or break stand on different footing. They would require Kemps to undergo the burden of reviewing some 30,000 additional documents. Yet IPL has not advanced a credible argument that these searches would net relevant, non-cumulative information central to the issues in the case, particularly since all of Kemps' products are FDA-regulated and Kemps has already agreed to many other search terms that appear better tailored to the issues in this case. Accordingly, the undersigned recommends that the Court order Kemps to use include the terms FDA AND IPL and “Food /3 Drug” AND IPL, but deny IPL's request that Kemps utilize other FDA terms.

IPL contends that Kemps should use eight terms that are variations of Kemps' product name (Sweet /2 Me) or Yo2 and the terms IPL, pint, recall, or break*. IPL argues the terms are likely to net relevant information related to Kemps' initial testing and purchase of IPL's containers, breakage of the containers, and the subsequent FDA recall. Kemps responds that IPL originally proposed SweetMe (or Sweet Me) and Yo2 (or Y02) as standalone searches, which were overbroad. Kemps argues that as a result of a previous discovery dispute teleconference with Judge Phillips, the parties effectively resolved any issues about product name searches by agreeing that Kemps' searches would include a search of each product name within 12 words of the term IPL. Kemps has provided information regarding the number of new documents each of the currently disputed product terms would capture as well as the number of documents that would be duplicative of other searches.

*8 The undersigned has reviewed the transcript of the July 31, 2020 teleconference with the Court. It reflects that IPL's counsel stated he was willing to compromise and not press Kemps to run standalone product name searches by adding IPL as a 12-word proximity limiter to the product name searches, but he needed more information from Kemps about the number of hits resulting from searches to assist his analysis in determining whether this narrower search term should be used. (July 31, 2020 Tr. p. 10). IPL's counsel expressly stated he was requesting hit volume information, not that the 12-word proximity limiter search be run. (Id.) The transcript further reflects that Kemps counsel agreed to provide information about the number of hits that would result from the product name combined with a 12-word proximity limiter. (July 31, 2020 Tr. p. 15). Based on submissions to the undersigned, it appears that after the July 31, 2020 teleconference, Kemps unilaterally began to run the 12-word proximity limiter searches after reviewing the requested hit count numbers. Kemps own submission states that it proceeded with this search “to avoid further delay and dispel any perception that Kemps was not acting reasonably.” (Kemps Submission No. 2, p. 7). Neither party's submission shows that IPL acquiesced in Kemps' unilateral decision to run the 12-word proximity limiter searches.

Kemps argues that requiring it to run further permutation of search terms involving the product names is not proportionate to the needs of the case. The undersigned finds that six of the disputed product terms are proportional to the needs of the case and two are not. According to Kemps, the search term (“Sweet /2 Me”) AND IPL returns 110 unique documents not covered by other searches it has already agreed to run, while the search of Yo2 AND IPL returns 496 unique documents. Kemps' submission does not state whether there is overlap between the two sets of search hits, but at most the two terms return 606 additional documents, all of which are likely related to the issues in this case. Kemps has offered no factual showing that the proximity limiter search it is currently running has already netted most of the relevant documents associated with a search that combines the two product names and IPL.

A similar analysis applies for search terms (“Sweet /2 Me”) AND recall, (“Sweet /2 Me”) AND break*, Yo2 AND recall, and Yo2 AND break*. According to Kemps, each of these terms returns between 1,076 and 2,810 documents not covered by other agreed searches. Again, Kemps' submission does not state whether there is overlap between the four sets of unique documents hits it reports as being returned by these search terms individually. At most, the four terms return a combined set of 7,184 unique documents for Kemps' relevancy review under the parties' ESI protocol. It is likely that these additional documents relate to significant issues in this case because they will pull in documents that reference one of the singular product names and breakage or recall, and Kemps has not shown its other searches have already netted most of the relevant information. In sum, the undersigned concludes that IPL has made a threshold relevancy showing, and Kemps has not demonstrated that review of the unique documents returned by these terms is disproportionate to the needs of this case.

The proportionality analysis differs for the proposed terms (“Sweet /2 Me”) AND pint and Yo2 AND pint. These two search combinations are not tailored to any important issues in the case. They would return documents that merely identify the products and the fact that the products are sold in pint containers. These are not contested issues. To the limited extent that the search terms (“Sweet /2 Me”) AND pint and Yo2 AND pint are likely to lead to the production of relevant information, the undersign concludes they are most likely to duplicate information that will be produced by searching other, more specific terms.

Accordingly, the undersigned recommends that the Court order Kemps include in its relevancy review any new/unique documents produced by the terms (“Sweet /2 Me”) AND IPL, Yo2 AND IPL, (“Sweet /2 Me”) AND recall, (“Sweet /2 Me”) AND break*, Yo2 AND recall, and Yo2 AND break*, and to deny IPL's request that Kemps use the search terms (“Sweet /2 Me”) AND pint and Yo2 AND pint.

*9 IPL contends Kemps should run four search terms that are variations of (Clear AND (Result OR Test)) and the terms IPL, pint, recall, or break* and the term Pint AND (Result OR Test). IPL argues the search terms are important because they relate to Kemps' decision making and testing before purchasing IPL's disputed containers and additional testing Kemps performed after receiving the reports of consumer complaints. Kemps objects on proportionality grounds. In its reply brief, IPL agrees to further narrow the term (Clear AND Result or Test) AND break* to include the word (sweet /2 me) or (Yo2).

The undersigned finds that the terms (Clear AND (Result OR Test)) AND IPL is proportionate to the needs of the case and the remainder are not. Kemps reports that running the term (Clear AND (Result OR Test)) AND IPL adds 231 new documents to Kemps' review. Each of these documents could be related to significant facts at issue because they reference both IPL and a test or result that IPL alleges Kemps performed. The limited burden on Kemps of reviewing 231 additional documents for relevancy does not outweigh other proportionality factors.

Kemps reports that the term (Clear AND (Result OR Test) AND pint does not return any documents that were not covered by other agreed searches. Accordingly, there is no need for Kemps to perform this wholly duplicative search. Kemps reports that the term (Clear AND (Result OR Test) AND recall returns 1,307 new documents not covered by agreed searches. However, the undersigned has already concluded that Kemps should run a variety of other “recall” related search terms that also include a reference to the FDA or the product names. To the limited extent that (Clear AND (Result OR Test) AND recall is likely to lead to the production of relevant information, the undersign concludes it will most likely duplicate information that will be produced by searching other, more specific terms that also include FDA or the product names in connection with the term recall.

Kemps reports that the term (Clear AND (Result OR Test) AND break* will return 18,470 documents not covered by other searches. Perhaps in recognition that this search term is non-specific to the issues in the case and appears to cast a net across a large number of immaterial documents, IPL offers in its reply brief to compromise by adding a requirement that the term be adjusted to also include the word (sweet /2 me) or (Yo2). However, in subsection 3, supra, the undersigned has already recommended that Kemps be required to perform searches that combine the words (“Sweet /2 me”) AND break* and Yo2 AND break*. Accordingly, running IPL's compromise version offered in its reply brief will be duplicative and unnecessary considering the undersigned's recommendations above.

Finally, the undersigned concludes that IPL's proposed search term of “pint” AND (Result or Test) is not proportional to the needs of this case. IPL argues that the term would assist in finding documents related to Kemps internal pre-purchase testing of IPL's containers or other testing after the reports of consumer complaints. Kemps points out that the term would add 14,896 documents to its review, many of which are likely irrelevant due to the likelihood it will draw in hits based on Kemps' unrelated product testing of flavors and marketing combinations for “pint” products. The undersigned agrees. In this instance, proportionality factors do not support requiring Kemps to run this non-specific, overinclusive term. There are a variety of other search terms that Kemps either has agreed to employ or which the undersigned has recommended that the Court order Kemps to employ that are better tailored to finding information within the scope of discovery in this case.

*10 IPL contends Kemps should run four search terms that are varying combinations of resin and the terms IPL, pint, recall or break*. IPL states that resin “is a common product used in manufacturing, including products manufactured by Kemps and IPL” and that “both parties have had discussions regarding resin used in the IPL containers.” (IPL Br. p. 6). IPL further explains that Kemps' knowledge and understanding of resin is significant to its misrepresentation and warranty claims, that resin has been a central issue of discovery, and Kemps' expert is conducting resin testing as part of the expert opinions in this case. Kemps responds that none of the resin searches are proportional to the needs of the case in part because Kemps has already agreed to search for resin within 12 words of IPL following the July 31, 2020 teleconference discussed in part 3 above. Kemps also argues that many of the resin search terms IPL suggests are overbroad and would catch any conversation any one of Kemps' 33 custodians would have had about resin in connection with many other unrelated products and vendors.

The undersigned finds that the term Resin AND IPL is proportionate to the needs of the case and the remainder of the resin terms are not. Kemps reports that a search of Resin and IPL will add 279 documents to its review. Each of these documents could be related to significant facts at issue because they reference both IPL and resin, which is a subject of Kemps' expert testimony. The limited burden on Kemps of reviewing 279 additional documents for relevancy does not outweigh other proportionality factors.

The remaining search terms involving Resin AND recall, pint or break* lack tailoring to the issues and appear unreasonably overinclusive or otherwise duplicative of the many other searches the undersigned recommends or which Kemps has agreed to run that include other variations of the term recall or break* combined with a party's name, the product trade names, or FDA.

Finally, IPL requests that Kemps search for the term “IML.” IPL notes that IML is an acronym for in-mold labeling and that Kemps requested IML for the IPL containers. IPL states that Kemps has served numerous discovery requests concerning IML and that Kemps has previously stated it would add 8,070 unique documents to Kemps' review. Kemps responds that use of the term IML is facially overbroad and not constrained to any of the products or issues in the case. Kemps also states that it is already running a search for IML within 12 words of IPL. In reply, IPL indicates that the search term is important because Kemps knowledge and understanding of IML is important. Based on the limited submissions provided by the parties on this particular term, the undersigned cannot conclude that requiring Kemps to run a standalone search for IML is proportionate to the needs of the case. While the term IML appears to involve a relevant subject of discovery, IPL has not provided a persuasive reason for extending the IML search as a single term across all ESI of Kemps' 33 custodians. The search Kemps is running that connects the term IML to IPL appears to be a reasonable and proportional effort to find relevant IML-related ESI.

D. Other findings

In its September 30, 2020 Order, the Court stated it “will allocate the Special Master's fee between the parties after considering the nature of the controversy, the parties' means, and the extent to which any party is more responsible than the other for the necessity to refer this matter to the master.” (Doc. 108, citing FED. R. CIV. P. 53(g)(3)). The Court directed the Special Master “to include in her report a summary on any findings to this effect.”

With respect to the discovery issues addressed in this Report and Recommendation, the undersigned finds that the parties have relatively equal means and neither party is significantly more responsible than the other for the necessity of referring these disputes to the Special Master. The undersigned finds that both parties contributed to delays and impasses by directing more focus on raw numbers of search hits than on a collaborative or practical approach to cost-effectively finding relevant ESI. The undersigned finds no evidence that either party acted in bad faith in any search term discussion and further finds that both parties acted within a range of practice norms she has seen in other similar matters.

*11 ESI discovery can be an iterative process that requires prompt and practical communication between parties. In the type of ESI protocol to which the parties agreed, the undersigned suggests that to the extent additional disputes arise, the parties should focus on ensuring that key sets of communications or documents that are relevant to the issues are prioritized for review and production. Both parties may be able to reduce their own cost burden by using methods they have investigated and reasonably selected as the best tailored methods for finding relevant information within their own files. Cross checking with search terms or methods suggested by the other party may help ensure relevant information is efficiently found. Ultimately, the emphasis must not be on the volume of ESI to review, but on finding relevant information that is proportional to the needs of the case.