The Exit Clause That Cost a SaaS Company $2M in Cloud Credits
Archi4 min read
The decision#
When a mid-sized SaaS company got a renewal offer from its cloud provider, the headline number looked great: $2 million in credits over three years, plus a 15% discount on compute. The catch was buried in an exit clause. Leaving early meant paying back a prorated share of the credits and covering data egress at a rate that would have cost more than the savings. The CTO, who asked not to be named because the contract is still under negotiation, walked the team through the math. They declined the offer and stayed on a month-to-month plan. I spoke with her about how the decision got made.
Q: What did the offer actually say?
A: It was a three-year commitment with $2 million in credits spread across the term. The discount applied to compute and storage, not to data transfer. The clause that mattered was in section 7. If we left before the term ended, we had to repay the unamortized credits and pay egress at list price for all data moved out during the migration. For us, that was about 40 terabytes. At list egress rates, the exit fee alone was north of $300,000. The credits were not free money. They were a lock.
Q: How did you model the trade-off?
A: We built two scenarios. In the first, we took the deal and stayed for three years. We saved about $1.4 million in compute and storage after accounting for the credits. In the second, we took the deal but had to leave after 18 months because of a product pivot or an acquisition. In that case, the repayment plus egress wiped out the savings and added about $200,000 in cost. We weighted the scenarios by likelihood. The probability of leaving early was not trivial. Two of our last three infrastructure decisions had changed within 18 months. Once we put a number on the exit risk, the deal looked less like a discount and more like a bet against our own flexibility.
Q: What did the provider say when you pushed back?
A: They offered to cap egress at a fixed rate for the first 12 months. That helped, but it did not address the credit repayment. They also offered to convert the credits into a reserved instance pool, which would have made them non-transferable. We asked for a shorter term with a smaller credit and a defined exit ramp. They said the pricing model did not support it. That was the moment the conversation shifted. We were not negotiating price anymore. We were negotiating risk allocation, and they were not willing to move.
Q: Who else was in the room?
A: Our CFO, our head of platform engineering, and a outside counsel who specializes in cloud contracts. The CFO was the one who flagged the repayment clause. She had seen it before at a previous company and knew how it could distort a budget. The platform engineering lead ran the migration estimate. Counsel read the termination language line by line. The three of them disagreed at first. The CFO wanted to take the deal and manage the risk. Engineering wanted out. I had to make the call, and I sided with engineering because the exit cost was asymmetric. A $1.4 million upside did not justify a $200,000 downside on a decision we might have to make under pressure.
Q: What did you do instead?
A: We stayed on a committed-use plan with a one-year term and no credits. We pay a little more per hour, but we can leave with 60 days notice and no repayment. We also moved about 15% of our steady-state workloads to a second provider to keep a migration path open. That costs us roughly $8,000 a month in duplicate tooling and engineering time. It is insurance, and we treat it as such. The savings from the original offer would have been larger on paper. The flexibility is worth more to us right now.
Q: What would you tell another CTO facing the same offer?
A: Read the termination section before you read the pricing table. The discount is the sales pitch. The exit clause is the contract. Model the cost of leaving at 12, 18, and 24 months, not just at the end of the term. If the exit cost is more than one quarter of the expected savings, the deal is probably a trap for a company that changes its mind. And get someone who has negotiated these clauses before. The language is not complicated, but the implications are easy to miss when you are looking at a big credit number.
More articles