Forty Thousand Emails Standing Between a Charity and Its Budget

The situation

The number Tom kept coming back to was 30,000 dollars, the amount a local IT consultant had quoted just to search and export the emails the other side's document request was asking for, against a dispute over a grant repayment worth somewhere between 8,000 and 35,000 dollars in total. Producing the documents, in other words, risked costing as much as or more than the claim itself.

Tom ran a small not-for-profit organization in Goderich that had received a grant several years earlier from a funding partner, later spending the money on a program the funder claimed had not matched the terms of the original agreement. The funder filed a Small Claims action seeking repayment, and as part of the standard document exchange that comes with civil litigation, sent a request asking Tom's organization to search its email systems using a long list of search terms and produce every message that matched, spanning five years of correspondence among staff, board members, and volunteers.

Run literally, the search terms produced roughly forty thousand hits across the organization's shared email accounts, a volume driven mainly by common words in the search list, like the funder's own program name and a few generic terms tied to the grant, that also happened to appear constantly in the organization's day-to-day, entirely unrelated correspondence. Reviewing forty thousand emails for relevance and privilege, even at a modest hourly rate, would have consumed a meaningful share of the organization's annual operating budget.

Before Tom brought the file to us, Biniam, a board member with a background in office administration, trying to help, had already attempted to respond to the request directly, offering informally to hand over a folder of what he considered the relevant emails without going through a proper search process, and separately telling Tesfay, the funder's lawyer handling the claim, over the phone that the organization would simply produce everything it had. Both moves, made with good intentions, had narrowed the room to negotiate a sensible scope by the time Tom sought legal advice.

The legal problem

Small Claims Court does not come with the kind of formal, court-supervised document discovery a larger civil action has. Its rules require each side to produce, ahead of the settlement conference, copies of the documents it actually intends to rely on at trial, not to run a search across every record it holds and hand over anything that might conceivably touch the dispute. The funder's request, asking for a search of five years of email using dozens of terms, went well beyond what the Small Claims process would ever have compelled on its own. That mattered a great deal to how the request could be answered, because it meant the starting point for any negotiation was not 'how do we comply with this' but 'how much of this were we ever actually required to do.'

That did not make the request something Tom's organization could simply ignore. A funder pursuing a legitimate claim was entitled to ask for the documents that would let it prove its case, and a court holding a settlement conference would take a dim view of a party seen to be stonewalling a reasonable request for relevant records, even one made informally rather than through a motion. The organization's real leverage came from the gap between what the funder had asked for and what a court would likely consider necessary and proportionate for a dispute of this size, not from a right to refuse outright.

The funder's initial search term list had not been built with that gap in mind. Broad, common terms generated an enormous number of false hits, and nothing in the request acknowledged that a Small Claims dispute over an amount in the low tens of thousands of dollars could not reasonably justify a document review costing more than the claim itself. That imbalance was the core of the legal problem: complying with the request as written would have imposed a cost on Tom's organization wildly out of proportion to what the litigation was actually about, while an outright refusal risked looking uncooperative at the very settlement conference meant to resolve the case.

The board member's earlier phone call complicated the negotiating position further, because opposing counsel had a reasonable basis to believe the organization had already agreed, informally, to produce everything. Walking that back without looking evasive, after a verbal commitment however casually made, needed to be handled carefully, distinguishing between a good-faith intention to cooperate and a binding agreement to a specific, unworkable process.

There was also a genuine relevance question buried in the volume problem. Somewhere in those forty thousand hits sat a real, and fairly small, set of emails that actually discussed the grant's terms, the program's implementation, and the funder's concerns, the documents that would matter to deciding the case. The task was not to avoid producing documents but to find a search approach that surfaced the emails that mattered without forcing a review of everything that happened to contain a common word.

What we did

1. Reviewed the full search term list against the organization's email systems to understand precisely why the hit count was so high, rather than accepting the forty thousand figure at face value. That review identified that a handful of overly broad terms, mainly the funder's own program name and a couple of generic words tied to the grant, were responsible for the bulk of the results, which meant the real problem was narrow and fixable rather than a sign the request itself was unreasonable in principle.
2. Clarified directly with opposing counsel that the board member's earlier phone offer was not a binding agreement to a specific production process, explaining plainly, in writing, that the organization intended to cooperate fully but needed a workable method to do so. We put this in writing early and deliberately, because leaving an ambiguous verbal exchange unaddressed would have let it harden into an assumed baseline that made every later negotiation harder to have.
3. Proposed a revised set of search terms that narrowed the broadest, most generic words to combinations tied specifically to the grant program's name paired with the funder's own account, and to the timeframe genuinely in dispute, aimed at capturing the same relevant emails with far fewer false hits. The goal was never to exclude anything real, only to stop pulling in years of unrelated day-to-day correspondence that happened to share a common word.
4. Ran a test search using the proposed revised terms before sending them to opposing counsel, confirming the hit count dropped from roughly forty thousand to a range in the low thousands. That gave us a concrete, defensible number to bring to the negotiation rather than an estimate or a promise, which mattered because opposing counsel had no reason to trust an unverified claim that the new terms would work.
5. Negotiated the final scope directly with opposing counsel over a series of calls and letters, trading a slightly broader date range for a tighter set of terms, since the funder's side had a legitimate interest in confirming nothing relevant had been excluded by an overly narrow search. Offering that concession voluntarily, rather than waiting to be pushed into it, kept the negotiation collaborative and avoided the cost of a contested motion neither side actually wanted to fund.
6. Retained the IT consultant only for the narrowed set of documents rather than the original full volume, cutting the anticipated review cost dramatically once the search terms were fixed and confirmed by the test search. Waiting until the scope was settled before spending anything on the consultant's time meant the organization never paid to search records that were going to be excluded anyway.
7. Reviewed the final narrowed set for privilege before production, withholding the small number of documents that reflected internal legal advice the organization had received about the grant, and provided a log identifying what had been withheld and why. That step protected the organization's confidential advice from disclosure while still meeting its obligation to account openly for everything it was not producing.

The outcome

The final negotiated search terms produced roughly nine hundred documents, a fraction of the original forty thousand hits and a volume the organization could review and produce for a cost that, while still real, no longer threatened to exceed the value of the claim itself. The review and production process was completed within the organization's usual monthly operating budget rather than requiring an emergency draw on reserves.

The concession the organization made was accepting a slightly broader date range than it initially wanted, covering a few additional months on either side of the core dispute period, to satisfy the funder's legitimate concern that a too-narrow search might miss something relevant. That was a real trade-off, not a clean win: producing more months of records than Tom's organization believed were strictly necessary, in exchange for search terms that were otherwise proportionate.

The underlying repayment dispute continued after the document exchange was resolved, proceeding toward the usual Small Claims resolution process. Whatever the eventual result on the merits, the document production issue itself was contained rather than allowed to become a separate, costlier fight layered on top of the original claim, and the organization avoided the outcome it had feared most going in, spending more to comply with a document request than the underlying dispute was ever worth.

What you can learn from this