An Uptime Promise Nearly Signed Away by a Trusted Friend

The situation

Ayesha was reading the final contract draft for the third time when she noticed a paragraph that had not been there two weeks earlier. The uptime guarantee her company had built its enterprise pitch around — 99.9 percent availability, with escalating service credits for every hour below it — had been rewritten. The new language capped total credits at a small flat amount regardless of how bad an outage got, and pushed the measurement window out so far that a full day of downtime could vanish inside a monthly average. She scrolled back through the document version history and saw the change had been made by Jomar, the company's head of client operations, three days after the customer's procurement team had sent their comments.

The company Ayesha ran had been built out of a family trust her parents set up when the software platform — a scheduling and dispatch tool for field service businesses — first started generating steady revenue. It had grown into an established business doing between roughly $1 million and $5 million a year, and this deal, with a national logistics operator, would be its largest customer by a wide margin. Jomar was not just an employee. He had joined in the company's earliest days as a favour to Ayesha's father, and he and Ayesha had been close friends for over a decade before either of them worked in software.

Adnan, the company's operations lead and the person who had actually built the escalating credits model in the first draft, was the one who flagged the discrepancy to Ayesha after noticing the customer's negotiator had stopped pushing back on service levels unusually early. When Ayesha asked Jomar directly why the clause had changed, his answer was that the customer had threatened to walk if the original terms stayed in, and that he had made the call himself to save the deal rather than bother her while she was travelling.

That explanation did not sit right. Ayesha had been reachable the entire time, and nothing in her email showed any communication from the customer threatening to walk. The contract was scheduled to be countersigned in five business days, and the company's board — effectively her parents, as trustees — expected the deal closed on the terms that had been presented to them a month earlier.

What made this urgent

The financial exposure was real but not the biggest part of the problem. Under the weakened clause, if the platform went down for an extended stretch during a peak dispatch period, the customer's credits would be capped at a few thousand dollars regardless of how much revenue or reputational damage the outage caused them — which sounded good for the company on paper, until the reverse was considered. A capped, watered-down service level clause is a two-way signal: it tells a sophisticated enterprise customer that the vendor does not actually stand behind its own uptime numbers, and it removes the vendor's own incentive to prioritize that customer's infrastructure when something breaks. Ayesha's technical team had spent months building monitoring and failover specifically to make the original 99.9 percent commitment defensible. A contract that quietly abandoned it undercut that investment for no clear benefit to either side.

The more pressing issue was internal. Jomar had authority to negotiate operational terms in customer contracts — that was his role — but changing a commercial term this significant without telling the president, on a deal of this size, went beyond what his mandate covered. Under general contract and agency principles, though, the question is not only what authority the company actually gave Jomar, but what authority the company appeared to give him. If Jomar had been put forward to the customer as the person who negotiated these terms, the customer would generally be entitled to rely on that, and the company could be held to the concession even though it was never authorized internally. Correcting the term promptly and in writing, before anything was signed or acted on, was what could still cut that off. Once signed, the calculus changes; an executed contract is very hard to walk back without the customer's cooperation.

Five business days did not leave much room. The customer's procurement team would see any last-minute change as either a sign the vendor was disorganized or, worse, a sign the earlier draft had been a bait-and-switch to win the deal. Ayesha also had to figure out, quietly, whether Jomar's unilateral edit was an isolated lapse in judgment under deal pressure or something that had happened before in smaller contracts nobody had caught. And she had to do all of this while deciding how much to involve her parents, who trusted Jomar personally and would be uncomfortable hearing that a family friend had gone around their daughter on a term this size.

What we did

1. Reviewed the full negotiation history against the executed authority Jomar actually held. His role covered day-to-day operational terms — support hours, onboarding timelines, integration scope — but the company's own internal approval process required commercial terms above a set threshold to go through Ayesha or the board. The service level and credits clause fell squarely in that category, which gave the company a clean basis to reopen it before signature.
2. Drafted a short, factual message to the customer's procurement lead correcting the draft. Rather than accusing anyone or explaining the internal breakdown, the message simply noted that the service level terms in the near-final draft did not match the terms the parties had discussed and agreed to earlier in the process, and proposed reverting to language closer to the original while remaining open to reasonable adjustments.
3. Proposed a middle structure instead of simply restoring the original clause outright. Rather than presenting the customer with an ultimatum, we built a revised credits schedule that kept the 99.9 percent target and the escalating credit structure for sustained outages, but softened the measurement window slightly and added a defined process for disputing an outage classification — addressing the customer's underlying concern about unpredictable penalty exposure without gutting the commitment.
4. Had a direct, documented conversation with Jomar before the customer call, not after. Ayesha needed to know, and needed it on record, whether the change reflected genuine customer pressure or a judgment call made without authority. Jomar acknowledged there had been no explicit threat to walk — only a sense that the customer seemed hesitant — and that he had made the edit unilaterally believing it would smooth the deal along.
5. Set out, in writing, the boundaries of Jomar's authority going forward. This was framed as a process fix rather than a punishment, since treating it as discipline would have made an already difficult personal conversation harder without making the company any safer. Any commercial term change above the existing threshold, on any customer contract, now required written sign-off from Ayesha before it left the building. The document was short, specific, and applied to future deals rather than relitigating this one.
6. Negotiated the revised clause directly with the customer's counsel over two calls. Going in with a specific, justified proposal rather than simply restoring the original language unprompted gave the customer's side room to accept a compromise without feeling steamrolled. The customer accepted the revised measurement window and dispute process in exchange for a slightly higher credits cap than the original draft had proposed, landing close to, but not identical to, where the deal had started.
7. Confirmed final language and closed the contract inside the original signing window. We circulated the finished draft to both sides for a last check before signature, since a rushed final read after weeks of back-and-forth is exactly when small errors slip through unnoticed. The extra step of correcting the clause added roughly a week to the process rather than derailing the signing entirely, since the customer's own team had never actually pushed for the weaker terms in the first place.

The outcome

The contract closed with a service level clause that restored the substance of the original 99.9 percent commitment and escalating credits, adjusted with the measurement window and dispute process that had not been in the first draft. The customer's procurement lead, once shown the corrected terms, did not push back meaningfully; the earlier weakening appears to have come entirely from inside the company rather than from any real customer demand. The deal was worth a significant new revenue line for the business, and it closed roughly a week later than originally planned, a delay the customer accepted without friction.

The harder cost was between Ayesha and Jomar. He kept his role, and there was no allegation of dishonesty about the business terms themselves — his account of trying to save the deal under perceived pressure was consistent throughout. But the trust that had let Ayesha not double-check his commercial edits for a decade did not fully return. The two agreed to the new sign-off process, and Ayesha's parents, once told a plain version of what had happened, supported it without drama, though the conversation was not an easy one.

The company came out with something it had not had before: a written, board-approved threshold for what commercial terms an operations lead could concede without escalation, applied to every customer contract afterward, not just the ones involving family. Ayesha has described the episode since as the moment the business stopped running on personal trust alone and started running on process that happened to also protect the personal relationships inside it.

What you can learn from this