The cancellation line that never picked up, for three years

The situation

Joao took the call from the organization's board treasurer on a Tuesday afternoon, and by the end of it he had a spreadsheet open with three years of bank statements laid out beside the original contract. The not-for-profit, a small community services organization in Ottawa, had signed up for a case management and scheduling platform when Joao was hired to overhaul its IT systems and bring its patchwork of spreadsheets and paper files onto something more current. The service was useful for about a year, helping staff track client intakes and appointments across two program sites. Then the organization switched to a different platform that better suited its grant-reporting requirements, one funders had specifically recommended, and Joao was told to cancel the old subscription as part of the transition.

The contract, signed years earlier by a since-departed executive director who had moved on before Joao was ever hired, specified that cancellation had to be done by phone, through a support line, rather than through the online account portal, an unusual term Joao only discovered when he went looking for a cancel button and found none. Joao called the number. He was placed on hold, then disconnected. He called again the following week, and the week after that, and eventually assigned a colleague to try during different hours, thinking the line might simply be understaffed at certain times of day rather than genuinely unreachable.

Manuel, the organization's board treasurer and a volunteer HVAC technician by trade, first noticed the ongoing charges on the organization's monthly bank reconciliation nearly two years after Joao's initial cancellation attempts had begun, a routine review he did every quarter as part of his treasurer duties. Nobody had disabled the automatic payment, because nobody had ever managed to reach a person who could process the cancellation on the vendor's end, and the online portal had no cancellation option at all, only a support-ticket form that generated automated replies promising a callback that never came.

By the time Manuel raised it with the full board, the organization had paid for a service it had not used in more than two years, and the monthly charge, though individually modest, had compounded into an amount well into six figures once the unused portion of the original term and the ongoing charges were added together. For an organization funded largely through grants and public donations, that was money that should have gone toward programming and staff, not a platform nobody had touched in years, and the board wanted it back.

The risk we had to size

Joao's first meeting with our office was less about the dollar figure and more about what the fight would look like and cost, because that was what the board actually needed to decide before authorizing anything further. A not-for-profit board weighing legal action against a vendor is not really asking whether the claim is strong; it is asking how much certainty that strength buys, and at what cost, to an organization with a limited operating budget and donors who expect that budget spent on programming rather than legal fees.

The core legal question was straightforward once the facts were laid out: a contract term requiring cancellation by a specific method is enforceable, but only if that method is actually available to the customer trying to use it. A vendor cannot rely on a cancellation clause while making the one channel it specifies for cancellation unreachable in practice, and continuing to bill a customer who made genuine, documented attempts to cancel under the contract's own terms exposes the vendor to a claim for the charges collected after those attempts began. That principle gave the organization real leverage, but leverage is not the same thing as a guaranteed number.

What we had to size for the board was the gap between that legal position, which was sound, and the practical path to actually collecting on it. The vendor was a mid-sized company represented by its own in-house counsel, Tariq, who had a standard playbook for exactly this kind of dispute, which meant any claim risked becoming a longer fight than the underlying dollar amount justified, particularly if the vendor disputed how many cancellation attempts had actually been made, when they started, or when the organization's obligation to keep trying by phone reasonably ended and it should have escalated sooner.

We also had to size the evidentiary risk carefully, because it directly shaped what the organization could realistically claim. Joao's early cancellation attempts had not been well documented, just phone calls with no record beyond his own memory of roughly when they happened. Only the more recent attempts, made after Manuel got involved and started keeping a log, had call times, dates, and the automated support-ticket replies preserved. That gap mattered a great deal, because it affected how far back the organization could confidently claim the charges collected were wrongful rather than simply overlooked.

There was one more factor in sizing the risk that had nothing to do with the merits at all: reputational exposure for a not-for-profit that relied on public donors and grant funders who expect careful stewardship of the money entrusted to them. A drawn-out public dispute, even a winnable one, carried its own cost in how the organization might be perceived, which is part of why a fast, negotiated outcome mattered to the board beyond the dollar figure alone.

What we did

1. Reconstructed the cancellation timeline from available records. We worked with Joao to pull phone bills, calendar entries, and email records that could corroborate even the early, undocumented calls, cross-referencing call durations against the organization's phone provider records to establish a credible pattern of repeated, good-faith attempts going back further than Joao's own memory alone could reliably support in a demand letter.
2. Sent a formal notice distinguishing documented from undocumented periods. Rather than claiming the full three years outright and inviting a dispute over the weaker early evidence, we structured the demand to lead with the strongly documented eighteen months, treating the earlier period as a secondary claim supported by weaker evidence, which made the vendor's response harder to dismiss wholesale on evidentiary grounds.
3. Framed the claim around the vendor's own contract term. The letter did not argue the contract as a whole was unfair, an argument that invites a much broader and slower fight; it argued the vendor had failed to honour the cancellation method the contract itself specified, a narrower and more defensible position that put the vendor in the position of explaining an unanswered phone line rather than defending its billing practices generally.
4. Gave the board a cost-and-certainty comparison before escalating further. Because predictability mattered to the organization as much as the outcome itself, we set out, in plain terms the board could act on, what a negotiated resolution was likely to recover against what a lawsuit might additionally recover, and what each path would cost in time and legal fees, so the board's decision to proceed was made with real numbers in front of it rather than a vague sense of the risk.
5. Opened settlement discussions before filing anything in court. We contacted Tariq, the vendor's in-house counsel, directly with the documented timeline, the supporting records behind it, and a specific recovery figure attached, giving him a defined two-week window to respond before litigation became the default path. That kept costs contained while the claim and its supporting evidence were still fresh, complete, and easy for the vendor to verify against its own records.
6. Held the line on the automated support-ticket evidence. When the vendor initially suggested the online form satisfied its cancellation obligations, we pointed to the organization's own submitted tickets and the form's own auto-reply language, which explicitly promised a callback that never came, closing off that line of defence early in the negotiation before it could gain traction.
7. Kept Manuel's board reporting current throughout. Because the board had asked for predictability, we provided short written updates after each significant exchange with the vendor, translated into plain terms rather than legal language, so Manuel could keep the full board informed without the process feeling opaque, which mattered as much to the client as the eventual number recovered.
8. Flagged the organization's other vendor contracts for review. While the dispute was still active, we suggested Joao pull the organization's other recurring service agreements to check for similarly restrictive cancellation terms, a small additional step that cost almost nothing to take at the time but let the board get ahead of a systemic risk across the organization's other contracts, rather than treating this dispute as an isolated incident and waiting for the next one to surface the same way.

The outcome

Tariq, on the vendor's behalf, confirmed the cancellation, effective retroactively to the date of Joao's earliest well-documented attempt, and agreed to refund the charges collected from that date forward in full. It also credited a portion, though not all, of the earlier and less documented period, reflecting the evidentiary gap we had flagged for the board from the very first meeting rather than something that emerged as a surprise later.

In total, the organization recovered close to the full documented amount and a meaningful share of the disputed early period as well, landing in the upper part of the amount the board had been weighing against when it authorized us to proceed. The organization did not recover every dollar billed over the full three years; the earliest charges, where the only evidence was Joao's recollection with no supporting record, were not part of the settlement, a concession the board accepted once it saw clearly that the alternative was a contested trial over facts that were genuinely hard to prove years after the fact, with no guarantee of a better result.

The matter resolved through negotiation, without a lawsuit ever being filed, in a little under four months from the first meeting to the signed settlement. For a board that had told Joao predictability mattered as much as the final number, that mattered a great deal: the organization knew what it was getting, and roughly when, rather than facing an open-ended fight against a vendor with considerably more resources to sustain one over years if it chose to.

Joao also used the process as an occasion to review the organization's other recurring service contracts, and found two more with similarly restrictive cancellation terms, which the board flagged for renegotiation well before either was due for renewal, turning a costly lesson into a small change in how the organization now signs contracts.

What you can learn from this