ORDER
STEPHEN L. CROCKER, United States Magistrate Judge.In this civil action, plaintiff David Haka contends that his former employer, Lincoln County, violated state and federal law when it eliminated his position in retaliation for his having filed a false claims complaint alleging improper child support billing by the county. Two discovery issues need to be addressed by the court.
The first, involving plaintiffs dissatisfaction with defendants’ privilege log, seems to have resolved itself, at least for the time-*578being and hopefully forever. See dkts. 37-39 and 41-44. Accordingly, I will deny plaintiffs motion to compel (dkt. 37) without prejudice because it is moot at this time.
The second, involving the exploration of electronically stored information (ESI), requires a bit more analysis. On February 15, 2007, plaintiff served a Request for Production of Documents on defendants, seeking all “documents, notes, memos, e-mails and meta-data of any Lincoln County official ... or any other Board member or officer regarding the reorganization or restructuring of the Lincoln County Child Support Agency.” Defendants report that documents responsive to plaintiffs request exist in electronic format, specifically on two external hard drives containing about four terabytes1 of data extracted from the computers of several Lincoln County employees and officials and from Lincoln County’s servers. (These drives contain an exact replica of ESI that was produced by Lincoln County in compliance with a federal grand jury investigation into the false claim allegations.) However, because of the cost involved in reproducing the hard drives in their entirety, defendants suggested to plaintiff that they attempt to agree upon search terms that could be used to cull relevant documents from the tremendous amount of data stored on the drives. In spite of their good faith efforts to reach an agreement, the parties have been unable to agree upon search terms broad enough to satisfy plaintiff that they will catch all relevant documents but narrow enough to satisfy defendants’ financial concerns. As a result, defendants filed a motion for a protective order (dkts. 31-34) which plaintiff opposes (dkts. 35-36).
Defendants, in their motion, suggest three alternative solutions: (a) limit plaintiffs last search to the e-mails on the hard drives, with plaintiff paying the cost (estimated to be $27,000 or higher), except for defendants’ privilege and relevance review; (b) outsource plaintiffs unredacted search to a named third-party reviewer (Midwest Paralegal) at plaintiffs expense (conservatively estimated at $55,000 + ); defendants would review for privilege and relevance (at an additional cost of $10,000) prior to disclosure to plaintiff, with the possibility of shifting some cost back to defendants on this last phase; or (c) outsource the search to a third party reviewer of plaintiffs choice at plaintiffs expense, with the same filtering and cost-shifting steps as suggestion (b).
Defendants argue that the amount at stake in this lawsuit, coupled with plaintiffs broadly over-inclusive search demand and the extraordinary cost of compliance compel protection pursuant to Rule 26(b)(2)(C). Invoking the seven factors identified in Zubulake v. UBS Warburg, LLC, 217 F.R.D. 309, 322 (S.D.N.Y.2003) and the seven factors listed in the Advisory Committee Notes to Rule 26(b)(2), 2006 Amendments, defendants contend that plaintiffs discovery requests are not at all tailored to the issues of this lawsuit; the amount of ESI to be searched is breathtaking; and plaintiff has not used other available discovery techniques to obtain relevant information or to narrow his ESI search request. According to defendants, they already have spend $25,000 providing Rule 26(a) disclosures, and would have to spend another $49,000 to complete the ESI search requested by plaintiff. Defendants calculate plaintiffs potential recovery of damages in this lawsuit to be significantly less than six figures. Defendants’ attorneys aver that they have discussed these concerns at length with plaintiffs attorneys but cannot agree on whether or how to limit the ESI search.
Plaintiff takes a different view of the dispute. Plaintiff defends the breadth of his requests, which he has narrowed during discussions with defendants, and predicts might be narrowed further upon receipt of defendants’ updated privilege logs. Plaintiff argues that defendants, as a public entity and its officers, have a statutory obligation to preserve and disclose these records to people like him, and their inability efficiently to do so is their fault, not his. As to the relative equities, plaintiff repeats his claim of that he was laid off under circumstances implying *579retaliation, and that he does not now have enough money to pay the extraordinary costs of searching defendants’ ESI. Plaintiff proffers that he has found a vendor who has quoted a price of about $27,000 to search defendants’ ESI (although it is not clear from the “preliminary pricing estimate,” Exh. B-3, that this is an apples-to-apples comparison with defendants’ low bid). Plaintiff characterizes some of defendants’ assertions of undue burden as speculation, but he reasserts his willingness to compromise with defendants to ensure that no one bears an inappropriate burden during discovery in this case.
Both sides make valid points. Although the issues at stake in this lawsuit are important, the potential damages are low, so that the cost of engaging in the ESI search plaintiff needs is disproportionate to the available recovery. On the other hand, defendants’ ESI, particularly their e-mail, is a potentially fecund source of relevant information that is not easily obtained from other sources. I say “potentially” because no one can say for sure what’s there without looking, but if defendants or their agents made any unguarded statements tending to show animus toward plaintiff, then they likely did so in their e-mails to each other. (The odds of any defendant in a civil lawsuit remembering and admitting to such statements while being deposed are low). So is it worth it to spend tens of thousands of dollars to explore ESI that might reveal a smoking gun but is equally likely to reveal nothing much? Neither litigant has deep pockets: plaintiff, a wage-earner with minimal resources, is suing a sparsely populated rural county with minimal resources. Although the county has statutory record-keeping obligations, it is entitled to obtain fair reimbursement for production costs. To the extent the public has an interest in the outcome of this ease, it can be vindicated through the U.S. Attorney’s investigation without the need to impose additional costs on the county.
Having considered all the relevant factors, I conclude that fairness and efficiency require the parties to proceed incrementally, limiting the initial search to the e-mails stored on the hard drives. Plaintiff is required to narrow his search terms to the narrowest set with which he is comfortable. Any additional searches shall occur only by joint agreement or court order. The parties will divide the costs of performing plaintiffs term searches of the e-mail 50/50. Defendants will pay 100% of the cost of a privilege/relevance review, which they are entitled to conduct prior to the disclosure of the search results to plaintiff. If the parties cannot agree on which company should perform the search(es), they shall hire the lowest bidder.
Accordingly, defendants’ motion for a protective order (dkt. 31) is granted in part and denied in part in the manner and for the reasons stated above.
. A terabyte is 1024 gigabytes, more than a trillion bytes, equivalent to 500 billion typewritten pages.