MEMORANDUM OPINION
BRUCE J. McGIVERIN, United States Magistrate Judge.This action involves numerous claims among the FDIC as receiver of Westernbank (“FDIC-R”), former directors and officers of Westernbank (collectively, “D & Os”), various insurers, and the FDIC in its corporate capacity (“FDIC-C”). This case was referred for an initial scheduling conference. (Docket No. 305). Prior to the conference, FDIC-R proposed an order establishing a “protocol” for the discovery of Western-bank’s electronically stored information (“ESI”), and to reduce the number of written interrogatories. (Docket No. 359). Certain D & Os offered a competing protocol. (Docket No. 363). Oral argument was heard at the conference, and the parties were granted leave to provide additional briefing. FDIC-R filed an initial brief (Docket No. 391), the D & Os opposed (Docket No. 395, 398), and FDIC-R replied (Docket No. 406). For the following reasons, FDIC-R’s request to alter the number of interrogatories is denied. A separate order will be issued to establish an ESI protocol.
BACKGROUND
The court summarized the general travel of the case and the nature of FDIC-R’s claims in its opinion and order denying the D & Os’ motions to dismiss; in short, it is a $176 million suit alleging D & 0 negligence in the course of making certain loans. (Docket No. 304). There have also been counterclaims, cross-claims, and third-party claims among the D & Os, FDIC-C, and the insurance companies. Prior to the initial scheduling conference, the parties noted their disagreement over the “protocol” for producing ESI held by FDIC-R, and filed competing proposals. (See Docket Nos. 359-1, 363). The court heard oral argument and granted the parties an opportunity for briefing. (Docket No. 372).
Ray Rivard, an FDIC-R employee, provided a written statement about the data FDIC-R is maintaining. (Docket No. 391-2, or “Rivard Deck”). When FDIC stepped in as receiver for Westernbank, it “immediately” possessed approximately 6.8 terabytes of ESI and 921,000 paper documents. “Most” of this original material was brought into a system called DMS iConnect (“DMS”), an internal database operated through a contractor. A subset of paper documents that FDIC “anticipated would be relevant to litigation against former directors and officers” were scanned into digital images and processed to generate searchable text. By this point, FDIC had spent $2.1 million. However, the contracts and sub-contracts for DMS are not litigation-specific, making it impractical to estimate the costs in this case with any greater detail.
FDIC plans to use a second contractor-maintained system called Relativity to give its litigation opponents searchable access to selected data. The FDIC will have to pay $450 per gigabyte to move data from DMS to Relativity. FDIC estimates its costs for producing documents to include: $0.185 per page for scanning paper documents and generating searchable text; $0.025 per page for Bates and confidentiality stamping; $325 per gigabyte for imaging native-format ESI into TIFF files; and $35 to $300 per labor-hour for technicians, quality control, and management staff. (Id. ¶¶ 4, 9, 12, 13, 15, 17). The FDIC-R has also provided lists naming specific ESI sources.
FDIC-R describes five steps of production. First, it would turn over “roughly 86,-000 pages” of “certain scanned documents,” which are “targeted sets of documents not amenable to search terms,” a stage it dubs “Phase I.” (Docket No. 391-1 ¶ 6; Rivard Deck ¶ 11) (emphasis added). Second, beginning “Phase II,” FDIC-R would confer with defendants, “collectively,” to “identify a reasonable set of search terms” across the DMS data set, including feedback statistics on positive matches. Third, once a set of search *71terms is fixed, FDIC-R will transfer matching data from DMS to Relativity. Fourth, D & Os will be allowed to access Relativity and select documents for final production. Fifth, FDIC-R will Bates-stamp, confidentiality-stamp, and export those documents from Relativity, completing “Phase II.” The recipients of Phase II production would pay FDIC-R $0.06 per page.
The D & Os’ process would begin with an initial production of documents FDIC-R identified in its initial disclosures. Next, the parties would confer and develop search terms to run against the searchable data, though FDIC-R would have to propose the first search terms. FDIC-R would provide both feedback statistics as well as sampled test productions. FDIC-R would then “perform a responsiveness review” on any documents subject to production. Should the parties agree on search terms, the universe of this review would be restricted to documents matching those search terms. In the absence of agreement, the FDIC would be required to review “all documents amenable to search terms.” FDIC-R would also manually review and produce the non-searchable documents, either “in the exact order in which they are kept in the ordinary course of business,” or labeled to match the requests to which they respond. No costs would be payable.
There is no material difference between the two proposals for the mechanical aspects of production, such as data formats and definitions of technical terms. Both parties suggested an express statement that backup media need not be recovered or searched. The two proposals contain similar language dealing with claw-back of inadvertent disclosures and requiring D & Os to privilege-log their own ESI. However, the D & Os would consult the court for every dispute. FDIC-R, for its part, would require “exercise of a reasonable standard of care” to invoke claw-back. Finally, the parties included similar language allowing them to omit materials disclosed pre-suit, though the D & Os add a requirement to identify those materials.
DISCUSSION
The parties’ briefing discuss three broad points: (1) whether FDIC-R must organize and label its responses to production requests; (2) whether the cost of FDIC-R’s ESI production should be shared by the D & Os; and (3) whether the D & Os should be required to submit consolidated interrogatories. Each of these are considered in turn, followed by a summary of the court’s reasoning.
I. Organization and Screening
The D & Os argue that FDIC-R must “perform a responsiveness review,” screening out irrelevant documents and labeling them to correspond to production requests; the FDIC-R counters that the “quick peek” framework it proposes is enough. When responding to a request for production, the responding party either “must produce documents as they are kept in the usual course of business or must organize and label them to correspond to the categories in the request.” R. 34(b)(2)(E)(i). The “usual course of business” alternative is only available when by the documents’ natural organization makes finding critical documents reasonably possible. See, e.g., FDIC v. Appleton, No. CV 11-476-JAK (PLAx), slip op. at 4—8 (C.D.Cal. Nov. 29, 2012) (available at Docket No. 396-1); Pass & Seymour, Inc. v. Hubbell, Inc., 255 F.R.D. 331, 334 (N.D.N.Y.2008); CooperVision, Inc. v. Ciba Vision Corp., No. 2:06-CV-149, 2007 WL 2264848 at *4 (E.D.Tex. Aug. 6, 2007). For that reason, responding government agencies sometimes cannot rely on this modality; investigation documents often lack meaningful organization, in contrast to an agency’s routine administrative records. See SEC v. Collins & Aikman Corp., 256 F.R.D. 403, 412-13 (S.D.N.Y.2009) (1.7 million documents resulting from investigation and stored in “large disorderly databases” could not be produced without labeling).
At this time, the court does not have enough information to support ordering FDIC-R to conduct organize-and-label production. The D & Os cite cases where responding parties failed to organize their productions, but do not suggest any factual basis for concluding that data that will be produced in this case is similarly disorganized. *72On the other hand, FDIC-R should not suppose that it may simply provide unsorted piles of data for the D & Os to pore over. While Ray Rivard’s declaration hints at how the data was preserved and indexed, there is no information that suggests the degree of organization that an actual production request will yield. In short, FDIC-R will not yet be required to label and organize its productions as a rule, but it may only avoid that burden to the extent its production satisfies the usual-course-of-business threshold. That could depend, for example, on whether critical organizational data (folder structures from network servers; physical sources of scanned documents) was preserved in the first place, as well as whether that organization is conveyed in the final production.
II. Cost-Shifting
FDIC-R seeks six cents per page for all ESI production beyond its initial disclosures. Yet “[f]or all discovery, including electronic discovery, the presumption is that parties must satisfy their own costs in replying to discovery requests.” Dahl v. Bain Capital Partners, LLC, 655 F.Supp.2d 146, 148 (D.Mass.2009) (citing Oppenheimer Fund, Inc. v. Sanders, 437 U.S. 340, 358, 98 S.Ct. 2380, 57 L.Ed.2d 253 (1978) and Zubulake v. UBS Warburg LLC, 216 F.R.D. 280, 283 CZubulake III) (S.D.N.Y.2003)). FDIC-R advances three rationales in support of its proposal: (1) ESI production costs are analogous to “copying costs” borne by the request- or; (2) the Westernbank ESI is “not reasonably accessible” under Rule 26(b)(2)(B); and (8) the seven-factor test applied in Zubulake III counsels cost-shifting here. I consider each point in turn.
A. ESI Production Costs as Copying Costs
FDIC-R argues it is entitled to demand an up-front contribution to its ESI production costs because (1) producing parties are not ordinarily burdened with the cost of making copies, and (2) ESI production costs are sometimes taxed under 28 U.S.C. § 1920 by analogy to its provision for copying costs. Both prongs of its syllogism are flawed. First, as the D & Os note, the fact that courts have taxed ESI expenses as costs awarded to prevailing parties at the end of a suit suggests they aren’t routinely shifted before then. (See Docket No. 395 at 23). But more importantly, FDIC-R fails to explain how the costs it identifies—scanning and performing OCR, creating static images, Bates stamping, and privilege review—are outside the realm of gathering and preparation expenses customarily borne by responding parties. See also Bills v. Kennecott Corp., 108 F.R.D. 459, 462 (D.Utah 1985) (“Ordinarily, the producing party bears the costs of reviewing and gathering documents while the requesting party pays for the costs of the copies only.”). Perhaps at most, the supply and labor costs of making and delivering recordable CDs, DVDs, or analogous media could fairly be borne by the D & Os. Cf. Dahl, 655 F.Supp.2d at 149 (requesting parties “would need only pay to put the converted documents on DVDs____”). But FDIC-R offers no evidence of what this cost would be. Regardless, it is not entitled to generalized per-page cost-shifting on this ground.
B. “Not Reasonably Accessible”
FDIC-R alternatively contends that the Westernbank ESI is not reasonably accessible within the meaning of Rule 26(b)(2)(B), justifying cost-shifting as a protective measure. That rule provides that “[a] party need not provide discovery of electronically stored information from sources that the party identifies as not reasonably accessible because of undue burden or cost.” The party resisting discovery has the initial burden of showing inaccessibility, after which the discovery proponent must establish “good cause” for requiring production. Id. In other words:
Rule 26(b)(2)(B) takes a categorical approach: it invites the classification of [ESI] as either “accessible” or “not reasonably accessible.” While cost and burden are critical elements in determining accessibility, a showing of undue burden is not sufficient by itself to trigger a finding of inaccessibility. For example, the sheer volume of data may make its production expensive, but that alone does not bring it within the scope of Rule 26(b)(2)(B). Rather, the cost *73or burden must be associated with some technological feature that inhibits accessibility.
Chen-Oster v. Goldman, Sachs & Co., 285 F.R.D. 294, 301 (S.D.N.Y.2012) (footnote omitted). See also W.E. Aubuchon Co. v. BeneFirst, LLC, 245 F.R.D. 38, 42-43 (D.Mass.2007) (unindexed, unsearchable filing system was not reasonably accessible, and so irrational that it bordered on intentional obfuscation). Here, FDIC-R has not hinted at any technical problem hindering access to the data the D & Os would have it search. To the contrary, both Rivard’s statement and FDIC-R’s proposed protocol suggest that relevant ESI has already been loaded into a retrieval system (DMS) that is both searchable and organized into meaningful databases.
FDIC-R argues, superficially, that another decision from this district “recognized that high production costs are sufficient to render ESI ‘not reasonably accessible’” under the rule. (Docket No. 391 at 13) (citing Rodríguez-Torres v. Gov’t Dev. Bank of P.R., 265 F.R.D. 40 (D.P.R.2010)). The plaintiff there had alleged age and sex discrimination; among the failure-to-promote claims ultimately found timely by the court, the highest-paid position she sought had a $51,400 annual salary. See 704 F.Supp.2d 81, 91 (D.P.R.2010) (opinion granting summary judgment for defendants). In its discovery order, the court held that an estimated $35,000 cost of production, exclusive of privilege review costs, was excessive “in this type of action.” 265 F.R.D. at 44. But while the court cited Rule 26(b)(2)(B), its rationale did not address accessibility so much as proportionality, which is not relevant to Rule 26(b)(2)(B)’s purpose and function. Cf. Chen-Oster, 285 F.R.D. at 303 (court may block disproportionate requests even when ESI is reasonably accessible under Rule 26(b)(2)(B)). In short, I reject the contention that Rule 26(b)(2)(B)—and its shifting burden to justify production requests—kicks in any time that discovery implicates both (1) electronically stored information and (2) large volumes of data, even where the volume renders review costly.
Because FDIC-R has not shown that access to the Westernbank data is hindered by any unique technological hurdles, it has failed to trigger Rule 26(b)(2)(B). It is therefore not entitled to categorically label the DMS databases “not reasonably accessible.”
C. The Zubulake Factors and Proportionality
FDIC-R argues that several of the “Zubulake factors” favor cost-shifting. The plaintiff in that ease alleged that she suffered employment discrimination because of her gender, as well as retaliation. In discovery, she sought e-mails that were largely stored on archived backup tapes which were “only accessible through costly and time-consuming data retrieval.” Zubulake III, 216 F.R.D. at 281. The court applied a seven-factor analysis to determine the extent to which the requesting plaintiff, rather than the responding defendant, should bear the cost of production:
1. The extent to which the request is specifically tailored to discover relevant information;
2. The availability of such information from other sources;
3. The total cost of production, compared to the amount in controversy;
4. The total cost of production, compared to the resources available to each party;
5. The relative ability of each party to control costs and its incentive to do so;
6. The importance of the issues at stake in the litigation; and
7. The relative benefits to the parties of obtaining the information.
216 F.R.D. at 284. Yet as the D & Os note, the court tailored these factors to allocate the cost of retrieving data that is not readily accessible—‘“[t]he responding party should always bear the cost of reviewing and producing electronic data once it has been converted to an accessible form.’ ” (Docket No. 395 at 24) (quoting Zubulake III, 216 F.R.D. at 290-91). In light of the accessibility analysis discussed above, I am persuaded that the Zubulake analysis does not apply per se. Notwithstanding, the court is required, “[o]n motion or on its own,” to:
*74... limit the frequency or extent of discovery otherwise allowed by these rules or by local rule if it determines that:
(i) the discovery sought is unreasonably cumulative or duplicative, or can be obtained from some other source that is more convenient, less burdensome, or less expensive;
(ii) the party seeking discovery has had ample opportunity to obtain the information by discovery in the action; or
(iii) the burden or expense of the proposed discovery outweighs its likely benefit, considering the needs of the case, the amount in controversy, the parties’ resources, the importance of the issues at stake in the action, and the importance of the discovery in resolving the issues.
R. 26(b)(2)(C). I would therefore focus on the proportionality considerations under this rule. But frankly, the parties’ broad claims about their respective discovery proposals are too speculative to merit a ruling at this time. FDIC-R’s affidavit describes the nature of the Westernbank data and itemizes some of the bulk costs, but does not shed any light on the effort in this case—particularly with respect to building responsive searches—that will be required to respond to particular requests. Likewise, it extols the policy virtues of its Relativity proposal, though it never articulates how using a contractor that charges $450 per gigabyte will reduce the net burden on FDIC-R. (See Docket No. 406 at 10). For their part, the D & Os point to two decisions where FDIC-R was ordered to provide discovery in a certain way. (Docket No. 395 at 19-22) (discussing Appleton and FDIC v. Klein, No. 1:12-CV-0896-RLV (N.D.Ga. Dec. 13, 2012) (available at Docket No. 395-1)). But the Appleton order came after discovery attempts showed that FDIC-R’s response was entirely unworkable. And as FDIC-R notes, the Klein court did not fully explain the considerations persuading it to adopt the defense’s ESI protocol; thus, that precedent does not weigh heavily in my consideration here.
In sum, this is less a situation where the scales are evenly balanced, and more one where the court has been given nothing to place on either side. Until the parties take affirmative steps to conduct discovery—perhaps after test runs, for instance—there is no ground for the court to dramatically alter the defaults under the Federal Rules of Civil Procedure. Cf. Zubulake v. UBS Warburg LLC, 217 F.R.D. 309, 324 (Zubulake I) (S.D.N.Y.2003) (declining to issue a cost-shifting order for inaccessible media until a test run was conducted; “by requiring a sample restoration of backup tapes, the entire cost-shifting analysis can be grounded in fact rather than guesswork.”).
III. Consolidated Interrogatories
Unrelatedly, FDIC-R also seeks an order requiring the defendants’ written discovery requests to be consolidated in light of the number of people it has sued. By default, Rule 33 permits each party to serve up to twenty-five written interrogatories. FDIC-R would have the defendants agree on fifteen questions in common, and leave each defendant with a bank of up to fifteen questions limited to “individual issues” facing each defendant. This, it argues, would spare it the burden of “responding to duplicative, overlapping requests on the same issues from each Defendant.” (Docket No. 391 at 17).
At best, I find FDIC-R’s request to be premature. As it is not complaining about any specific requests for interrogatories, its abstract argument about the burdens it will face are difficult to assess. Moreover, the default rule strikes a balance by correlating the potential number of interrogatories a party faces with the number of parties in the suit, which seems to be a reasonable proxy for the case’s overall complexity. Gross abuses, such as those narrated in the eases FDIC-R cites, may justify tinkering with this balance, as could stipulations or partial settlements that narrow the scope of the case. Since there is no such demonstration here, I am not persuaded that the court should issue such an order at this time.
IV. Summary
To give the parties a roadmap, the court will issue an order reflecting the reasoning discussed above. In short, FDIC-R will neither be categorically required to organize *75and label its productions, nor permitted to produce documents without adequate organization. No cost-shifting is justified at this time, though trial runs showing disproportionate costs, or abusive discovery tactics, could warrant reconsideration.
Without explanation in their briefing, the D & Os would have FDIC-R propose initial search terms. But as former directors and officers of Westernbank, the D & Os are more likely to have an idea of what documents they are looking for in a particular request. Therefore, the court will require the requesting party to propose search terms first—though since FDIC-R oversaw the loading of ESI into DMS, it is expected to provide active assistance, and should anticipate consulting its technically—skilled staff or contractors as necessary.
In addition to specifying a protocol for negotiating searches, the order will explicitly require FDIC-R to produce and continue producing materials subject to its initial disclosure obligations. But because the parties’ briefing only touched on the terms for producing FDIC-R’s searchable data in the DMS system, the order will not address production by either (a) the other parties, or (b) any of FDIC-R’s other sources of discoverable material, whether digital or physical. Those sources of production remain subject to discovery under the ordinary rules.
The parties’ agreed technical procedures and vocabulary will be incorporated with only stylistic changes, along with their general claw-back language. However, the language duplicating Fed.R.Evid. 502 and Fed. R.Civ.P. 26(b)(5)(B) will be simplified. Finally, rather than endorsing the suggestion that the parties will bring all privilege questions to the court (see Docket No. 363-1, ¶¶ 13-15), language is added emphasizing the parties’ obligation to meet and confer under the local rules.
CONCLUSION
For the foregoing reasons, FDIC-R’s request for a change in the number of interrogatories is DENIED. A separate order will issue containing the court’s ESI protocol.
IT IS SO ORDERED.