Justice Guzman delivered the opinion of the Court.

Electronic discovery plays an increasingly significant role in litigation and, often, at significant expense. Given the prevalence of discoverable electronic data, discovery disputes involving electronically stored information (ESI) are a growing litigation concern. With few occasions to enter the fray,1 we have an opportunity in these consolidated mandamus proceedings to provide further clarity regarding ESI discovery.

Though increasingly common, electronic discovery concerns manifest in variable shades and phases. In this dispute, the parties are at odds over the form in which ESI must be produced, presenting conflicting views regarding the proper interpretation and application of our discovery rules concerning such matters. The requesting party seeks ESI in native form while the responding party has offered to produce in searchable static form, which the responding party asserts is more convenient and accessible given its routine business practices. Agreeing with the requesting party, the trial court ordered production in native form, subject to a showing of infeasibility. The court of appeals denied mandamus relief.2

Under our discovery rules, neither party may dictate the form of electronic discovery.3 The requesting party must specify the desired form of production,4 but all discovery is subject to the proportionality overlay embedded in our discovery rules and inherent in the reasonableness standard to which our electronic-discovery rule is tethered.5 The taproot of this discovery dispute is whether production in native format is reasonable given the circumstances of this case. Reasonableness and its bedfellow, proportionality, require a case-by-case balancing of jurisprudential considerations, which is informed by factors the discovery rules identify as limiting the scope of discovery6 and geared toward the ultimate objective of "obtain[ing] a just, fair, equitable and impartial adjudication" for the litigants "with as great expedition and dispatch at the least expense ... as may be practicable."7

Delay and expense strain not only the resources of the parties, but also the judicial system.8 Consequently, the discovery rules imbue trial courts with the authority to limit discovery based on the needs and circumstances of the case, including electronic discovery. Thus, when a party asserts that unreasonable efforts are required to produce ESI in the requested form and a "reasonably usable" alternative form is readily available, the trial court must balance any burden or expense of producing in the requested form against the relative benefits of doing so, the needs of the case, the amount in controversy, the parties' resources, the importance of the issues at stake in the litigation, and the importance of the requested format in resolving the issues.9 Even without quantifying differences in time and expense, evidence that a "reasonably usable" alternative form is readily available gives rise to the need for balancing, and if these factors preponderate against production in the requested form, the trial court may order production as requested only if the requesting party shows a particularized need for data in that form and "the requesting party pay[s] the reasonable expenses of any extraordinary steps required to retrieve and produce the information."10 Unless ordered otherwise, however, "the responding party need only produce the data reasonably available in the ordinary course of business in reasonably usable form."11

Because neither the trial court nor the parties had the benefit of the guidance we seek to provide today, we deny the petitions for writ of mandamus without prejudice, affording the relator an opportunity to reurge its discovery objections to the trial court in light of this opinion.

I. Factual and Procedural Background

In these mandamus proceedings, residential homeowners sued their insurer and others alleging underpayment of insured hail-damage claims. The lawsuits assert statutory, contractual, and extra-contractual claims against the same insurer, State Farm Lloyds, in separate proceedings. We consolidated the mandamus petitions for argument because they present the same legal issues and substantially similar procedural underpinnings.

At issue are trial-court orders adopting a proposed protocol for the exchange of electronic discovery. As requested by the homeowners, the trial court ordered all ESI to be produced in its native or near-native forms rather than in the alternative, "reasonably usable" format State Farm proposed in a competing discovery protocol.12 The court-ordered protocol does not require State Farm to convert data stored in another form back to native form or to produce the same information in multiple forms. But it does require State Farm to produce ESI in native form regardless of whether a more convenient, less expensive, and "reasonably usable" format is readily available. If native form is "infeasible" to produce, however, a near-native form may be substituted if the parties agree on the substituted form.

Native format "retains the file structure associated with and defined by the original creating application."13 For example, the native format is XLS for Microsoft Excel spreadsheets and DOC for older versions of Microsoft Word documents. The homeowners insist production in native form is vital for several reasons related to the visibility, utility, and searchability of metadata. Metadata, "colloquially known as `data about data,' encompasses the structural information of a file that contains data about it as opposed to describing its actual substantive content. Often hidden and embedded within the original file, metadata does not normally appear on a printed page."14

State Farm has offered to produce ESI in searchable, but "static" form. PDF, TIFF, and JPEG files are common examples of static electronic formats. Static forms of ESI are created by converting native formats into static images, which removes metadata from the native files. Static form may be searchable — to a more limited extent than native form — using optical character recognition (OCR) software.15

To support ESI production in searchable static form, State Farm offered evidence that it processes more than 35,000 new claims each day and, in the ordinary course of business, information related to those claims is routinely converted into static format. When claims are being processed, claims-related information is necessarily created in native form. With regard to some types of claims-processing information, the native form is static, for example, handwritten notes and photographs. But to facilitate efficient business operations, State Farm employs a central repository — the Enterprise Claims System (ECS) — that is "the system of record" for claims handling at State Farm.16 Claims-related information originally created in disparate systemic locales must be uploaded to the ECS, where it is converted and stored in secure, read-only formats for data integrity and access (e.g., PDF, TIFF, or JPEG). By consolidating information from different sources into the ECS, the claims-file information becomes readily accessible for processing claims on behalf of policyholders and enables effective management of claims processes. Some ESI information exists solely in the ECS platform, but other information may also exist in native forms elsewhere within thousands of State Farm servers.

According to State Farm's discovery expert, static format is beneficial because the information can be searched, reviewed, and handled without inadvertent or intentional alteration by ECS users. With regard to litigation impacts, the expert reported that ESI in static format is easier to Bates number for discovery; allows efficient management of documents as exhibits at depositions, hearings, and trials; enables redaction, which is not possible with most native forms of ESI; and avoids intentional or unintentional alteration of the information, which may be difficult to detect or propagate further disputes about data integrity. According to a State Farm business analyst, "ECS is the most reasonably available source of claim file information in the ordinary course of business. It is the most convenient, least burdensome and least expensive means of producing the information plaintiff requested."

State Farm's expert averred that production in the native form of files "would require State Farm to engineer a new process that includes determining upstream sources of the data, validating the upstream sources, determining whether native files of the information still exist, and developing an extraction method for the native versions."17 Without quantifying the time or expense involved, the expert opined that "[t]hese additional steps would be an extraordinary and burdensome undertaking for State Farm" and are unnecessary because State Farm's proffered production form is "reasonably usable."

State Farm's business analyst further elaborated on the burden of native-form production relative to the convenience and cost-effectiveness of producing the information as it is maintained in the ECS, explaining that:

The homeowners supported their proposed electronic discovery protocol with expert testimony that static images have less utility compared to native format, which would allow them to see formulas in Excel spreadsheets, search and sort the information by data fields, analyze the relationship of data, and see information in color that may not translate as accurately to stored or printed static images. Referring to static-form production as "the electronic equivalent of a print out," the homeowner's expert explained that useful metadata would not be viewable in static form, including tracked changes and commenting in Word documents; animations, other dynamic information, and speaker notes in static printouts of PowerPoint documents; and threading information in emails that would allow construction of a reasonable timeline related to State Farm's processing of the homeowners' claims. The expert also opined that production of ESI in static form is significantly more expensive for the requesting party, due to the fact that storage costs rise with the size of the file and conversion to static form drastically increases the size of ESI files. The homeowners thus assert searchable static format is not a "reasonably usable form," as State Farm contends.

The homeowners' expert also refuted State Farm's claim of burden, testifying production would be as simple as handing over native ESI on a "thumb drive or on an external hard drive." Noting that ESI in native form has to be gathered to create static form in the first instance, the expert disclaimed the existence of any added burden on State Farm, stating: "Not only would [producing in native format not] require extraordinary steps, it would require fewer steps than those that they are employing right now. When you take the native data, you are dealing with it as it lies." Summarizing the homeowners' position, the expert explained, "[W]e're not imposing any additional duties, we're only asking that they not be allowed to dumb down, to downgrade the data for production."

State Farm sought mandamus relief from the court of appeals, arguing in both cases that Texas Rule of Civil Procedure 196.4 allows for production of ESI in reasonably usable forms and, considering the proportionality concerns delineated in discovery Rule 192.4, the trial court abused its discretion in requiring native production in lieu of the reasonably usable form State Farm offered.

The court of appeals denied mandamus relief in both cases, holding:

The court rejected State Farm's proportionality concerns, deeming State Farm's evidence of undue burden conclusory and lacking estimates of the time, expense, and "extraordinary steps" required to retrieve and produce ESI in the requested form.19 The court found a complete absence of "data with which to conclude that the burden or expense of the proposed discovery outweighs its likely benefit."

State Farm, supported by several amici, characterize the lower court rulings as granting requesting parties "essentially unlimited power" to dictate how the responding party must conduct electronic discovery under the Texas Rules of Civil Procedure.20 We set the matter for oral argument and write to (1) clarify that neither the requesting nor the producing party has a unilateral right to specify the format of discovery under Rule 196.4 and (2) provide guidance regarding the application of Rule 192.4's proportionality factors in the electronic-discovery context.

II. Discussion

A. Standard of Review

The scope of discovery is generally within the trial court's discretion, but the court "must make an effort to impose reasonable discovery limits."21 A writ of mandamus will issue only if the trial court reaches a decision "so arbitrary and unreasonable as to amount to a clear and prejudicial error of law" and the relator has no adequate remedy by appeal.22 In determining whether the trial court clearly abused its discretion, an appellate court may not substitute its judgment for the trial court's determination of factual or other matters committed to the trial court's discretion, even if the mandamus court would have decided the issue differently.23 Mandamus relief is only appropriate in such cases when the relator establishes that the trial court could have reached only one conclusion and that a contrary finding is thus arbitrary and unreasonable.24 But with regard to questions of law and mixed questions of law and fact, "`a trial court has no "discretion" in determining what the law is or applying the law to the facts,' even when the law is unsettled."25

B. Form of Electronic Discovery

The rules of civil procedure generally extend the scope of discovery to "any matter that is not privileged and is relevant to the subject matter of the pending action, whether it relates to the claim or defense of the party seeking discovery or the claim or defense of any other party."26 As a counterbalance rested in concerns about "unwarranted delay and expense,"27 Rule 192.4 expressly constrains the scope of discovery as to otherwise discoverable matters:

To put it succinctly, "the simple fact that requested information is discoverable ... does not mean that discovery must be had."29 So while metadata may generally be discoverable30 if relevant31 and unprivileged,32 that does not mean production in a metadata-friendly format is necessarily required. Indeed, as a federal district court recently observed, "a weak presumption against the production of metadata has taken hold," which may be due to "metadata's status as `the new black,' with parties increasingly seeking its production in every case, regardless of size or complexity."33

Whether production of metadata-accessible forms is required on demand engages the interplay between the discovery limits in Rule 192.4 and production of electronic discovery under Rule 196.4, which provides:

In In re Weekley Homes, we summarized the "proper procedure" under Rule 196.4, including the directive that the parties "make reasonable efforts to resolve the dispute without court intervention."35 Meeting and conferring to resolve e-discovery disputes without court intervention is essential because discovery of electronic data involves case-specific considerations and each side possesses unique access to information concerning reasonable and viable production methods, resources (technological or monetary, for instance), and needs.36 In re Weekley Homes did not consider the precise issues presented here, however — namely, whether the form requested controls and how proportionality factors into the analysis.

At its core, the homeowners' claim that native-form production is required presumes the requesting party can unilaterally determine the form of production. But Rule 196.4 cannot be construed so narrowly given its focus on "reasonable" efforts and "reasonabl[e]" availability.37 Though the term "reasonable" cannot be comprehensively defined, it naturally invokes the jurisprudential considerations articulated in Rule 192.4.

Thus, if the responding party objects that electronic data cannot be retrieved in the form requested through "reasonable efforts" and asserts that the information is readily "obtainable from some other source that is more convenient, less burdensome, or less expensive," the trial court is obliged to consider whether production in the form requested should be denied in favor of a "reasonably usable" alternative form.38 In line with Rule 192.4, the court must consider whether differences in utility and usability of the form requested are significant enough — in the context of the particular case — to override any enhanced burden, cost, or convenience. If the burden or cost is unreasonable compared to the countervailing factors, the trial court may order production in (1) the form the responding party proffers, (2) another form that is proportionally appropriate, or (3) the form requested if (i) there is a particularized need for otherwise unreasonable production efforts39 and (ii) the court orders the requesting party to "pay the reasonable expenses of any extraordinary steps required to retrieve and produce the information."40

Here, State Farm contends searchable static form is not only adequate, but more cost-effective and convenient. In other cases, a party may resist static-format production, if requested, because producing in native format is easier and less expensive. When a reasonably usable form is readily available in the ordinary course of business, the trial court must assess whether any enhanced burden or expense associated with a requested form is justified when weighed against the proportional needs of the case.41 The proportionality inquiry requires case-by-case balancing42 in light of the following factors:

1. Likely benefit of the requested discovery: If the benefits of the requested form are negligible, nonexistent, or merely speculative, any enhanced efforts or expense attending the requested form of production is undue and sufficient to deny the requested discovery. In such cases, quantifying or estimating time and expenses would not be critical, as it may be when benefits clearly exist. At the opposite end of the spectrum, a particularized need43 for the proposed discovery will weigh heavily in favor of allowing discovery as requested but, depending on the force of other prudential concerns, may warrant cost-shifting for any "extraordinary steps" required.44

Courts should consider cumulative effects rather than viewing benefits and burdens in a vacuum. Here, for example, many similar cases arising from the same extreme weather event are currently pending against State Farm. The identification and retrieval process State Farm would have to develop for native-form production may be a ticket for one train only — exponentially increasing the burden when considered in the context of repeated litigation — or have broader utility, which could have a cumulatively reductive effect. The record does not tell us, but if there are likely uses for the identification and retrieval process beyond the instant mandamus cases, initial burden and expense may be substantially ameliorated, and if not, the burden and expense may be significantly enhanced.

2. The needs of the case: In these mandamus cases, the homeowners seek native production both for optimal search capability and to access metadata. Recognizing that metadata serves no genuinely useful purpose in many cases, "many parties, local rules and courts have [in current practice] endorsed the use of [static] image production formats, principally TIFF and [PDF] formats."45 But metadata may be important, even dispositive in some cases.46

Relevance of metadata and the relative significance to the case must be determined on a case-by-case basis. But metadata's relevance must be obvious or at least linked, more or less concretely, to a claim or defense. Hypothetical needs, surmise, and suspicion should be afforded no weight. As a general proposition, metadata may be necessary to the litigation when the who, what, where, when, and why ESI was generated is an actual issue in the case, not merely a helpful or theoretical issue.47 Take, for instance, a wrongful termination case where timing of the events leading up to and following termination or authorship of case-critical documents might be a central issue in the case.48

Here, the homeowners have argued production in native format is necessary to ensure disclosure of all potentially relevant information. By way of example, the homeowners provided evidence that captions annotating some photographs of hail damage to a house — such as "north elevation with hail damage missed by [claims adjuster]" — were not captured when the photographs were converted to static format in the ECS. State Farm insists, however, that PDF production of photos from the ECS does not support the necessity for native-form production, as the homeowners claim, because the omitted photo captions were provided to the homeowners through an ECS "caption log" as well as via production from another database State Farm had identified as storing discoverable ESI.49 In evaluating whether a particular form of production is required, the court should consider not only the relative importance of the information to the central issues in the case, but also availability of that information from some other source that is more convenient, less burdensome, or less expensive.50

3. The amount in controversy: Accessibility — or relative inaccessibility — of electronic data contributes to increased costs and burdens associated with electronic discovery.51 "While large companies are still learning to cope with e-discovery costs, e-discovery remains costly and complex for the small company, small case, and unrepresented litigant. Because e-discovery is very expensive and quite complicated, the advent of e-discovery is forcing settlements, and thus, denying litigants an opportunity to litigate the merits of the case."52

When the discovery rules were adopted, an explanatory guide explained an initial impetus for discovery constraints that rings just as true in today's electronic discovery frontier:

For these reasons, the amount in controversy plays a pivotal role in determining whether production in a specified form is justified given the burden or expense required to meet the demand.

4. The parties' resources: Whether the producing party has the means to fairly and realistically produce in the requested format is a significant proportionality consideration. An expense that is a drop in the bucket to one party, may be insurmountable to another.54 While this factor is important to the balancing inquiry "considerations of the parties' resources does not foreclose discovery requests addressed to an impecunious party, nor justify unlimited discovery requests addressed to a wealthy party."55 Rather, "`the court must apply the standards in an even-handed manner that will prevent use of discovery to wage a war of attrition or as a device to coerce a party, whether financially weak or affluent.'"56

But beyond financial resources, one must also consider whether the requesting party has the technological resources to make proper use of ESI in the form requested.57 A high-powered luxury sports car is useless to someone who lacks a license to drive it. To that point, the homeowners' discovery expert testified it would be more convenient for lawyers lacking advanced technology to use image formats, but the homeowners' counsel in this case has "invested considerably to have the tools necessary to be able to deal with advanced forms of information." This may be another way of looking at the benefit versus the burden; if the potential benefits could not be realized, any associated burden would be unwarranted.

5. Importance of the issues at stake in the litigation: Legal disputes are always important to those who are litigating them. For one side, however, the precedential value may be more significant than the other, justifying an outlay of time and expenses that would otherwise be unwarranted. Likewise, "`many cases in public policy spheres, such as employment practices, free speech, and other matters, may have importance far beyond the monetary amount involved.' Many other substantive areas also may involve litigation that seeks relatively small amounts of money, or no money at all, but that seeks to vindicate vitally important personal or public values."58

6. The importance of the proposed discovery in resolving the litigation: Discovery must bear at least a reasonable expectation of obtaining information that will aid the dispute's resolution. Reasonable discovery does not countenance a "fishing expedition."59

7. Any other articulable factor bearing on proportionality: The foregoing factors are derived directly from the discovery rules, but are certainly not exclusive. As history tells, technology is constantly evolving at rapidly increasing rates. The legal system is not nearly as agile, leaving us in a perpetually responsive posture. Trial courts have flexibility to consider any articulable factor that informs this jurisprudential inquiry.60

C. Parity with the Federal Rules of Civil Procedure

Our application of proportionality principles in this context aligns electronic-discovery practice under the Texas Rules of Civil Procedure with electronic-discovery practice under the Federal Rules of Civil Procedure.

Rule 34 of the federal procedural rules, which governs electronic discovery in federal-court proceedings, states:

Rule 34's plain language does not permit either party to unilaterally dictate the form of production for ESI. The default form is "a form or forms in which [ESI] is ordinarily maintained or in a reasonably usable form or forms," and the trial court retains discretion to order discovery in a format that is appropriate to the circumstances.

The Advisory Committee Notes to Rule 34 explain:

The federal rules thus place the power to decide the form of electronic discovery not with the parties, but within the trial court's sound discretion.

In Dizdar v. State Farm Lloyds, a similar hail-storm lawsuit against State Farm involving a similar ESI protocol, a federal district court exercised its discretion to deny the plaintiff's request for native and near-native production.68 The court explained that while the federal rules give "preference to documents produced as they are ordinarily maintained absent a specific request from the other party ... it does not follow that if the requesting party specified a form for production, the information must be produced in that form."69 The court determined that the plaintiffs could gain all the information necessary to litigate the claims against State Farm from the ECS claim file such that the existence of any additional burden was not justified. Finding State Farm's proffered form of production "a reasonably usable format" within the meaning of the federal discovery rules, the court granted State Farm's request to produce ESI in the searchable static image claims files archived in the ECS.70

To be sure, there are differences in language between the Texas rule and the federal rule. But as we affirmed in In re Weekley Homes, "our rules as written are not inconsistent with the federal rules or the case law interpreting them," even though they may not "mirror the federal language."71

In this regard, we observe the proportionality principles under the federal rules similarly limit discovery of otherwise discoverable information. Rule 26 delineates the scope of discovery as inherently limited by proportionality:

The factors expressly identified as limitations on the scope of discovery accord with the proportionality analysis we have explicated under our discovery rules, keeping our procedures in line with their federal counterparts.73

Though the proportionality factors were recently relocated within the federal rules, proportionality has long been a required constraint on the scope of discovery, enacted decades ago "`to deal with the problem of overdiscovery.'"74 Proportionality, it is said, acts as a governor "`to guard against redundant or disproportionate discovery by giving the court authority to reduce the amount of discovery that may be directed to matters that are otherwise proper subjects of inquiry.'"75 The 2015 amendments to the federal rules have not altered "the existing responsibilities of the court and the parties to consider proportionality," do not require the requesting party to address all proportionality considerations, and do not permit the opposing party to rely on boilerplate objections of disproportionality.76 Rather, the amendment is directed to changing the existing "mindset" that relevance is enough,77 restoring proportionality as the "collective responsibility" of the parties and the court.78

Discovery is necessarily a collaborative enterprise, and particularly so with regard to electronic discovery.79 The opposing party must object and support proportionality complaints with evidence if the parties cannot resolve a discovery dispute without court intervention,80 but the party seeking discovery must comply with proportionality limits on discovery requests and "may well need to ... make its own showing of many or all of the proportionality factors."81 Recent federal cases provide helpful examples of proportionality analyses in e-discovery cases.82 Consistent with an individualized, case-specific inquiry, all manner of outcomes are represented — denials of requested discovery on proportionality grounds,83 rejection of proportionality complaints,84 and cases taking a more graduated approach by allowing limited bell-weather discovery to inform the propriety of further discovery.85 Ultimately, the "court's responsibility, using all the information provided by the parties, is to consider [the proportionality] factors in reaching a case-specific determination of the appropriate scope of discovery."86 The same applies with regard to electronic-discovery practices under the Texas Rules of Civil Procedure.

III. Conclusion

Today, we elucidate the guiding principles informing the exercise of discretion over electronic-discovery disputes, emphasizing that proportionality is the polestar. In doing so, we further a guiding tenet of the Texas Rules of Civil Procedure: that litigants achieve a "just, fair, equitable and impartial adjudication ... with as great expedition and dispatch and at the least expense ... as may be practicable."87 Because the trial court and the parties lacked the benefit of our views on the matter, neither granting nor denying mandamus relief on the merits is appropriate. Accordingly, we deny the request for mandamus relief without prejudice to allow the relator to seek reconsideration by the trial court in light of this opinion.88