How do you convince stakeholders that "Error Budgets" are better for business than 100% uptime?

All answers to this question.

You have to frame it in terms of "Opportunity Cost." In 2023, I started telling my stakeholders that chasing 100% uptime is like buying a tank to drive to the grocery store—it’s expensive and unnecessary. An Error Budget defines the point where users actually start to get frustrated. If we are within our budget, we should be taking risks and deploying features. If we exceed it, we stop and focus on reliability. This aligns the business's desire for speed with the SRE's desire for stability.

   Answered 2025-11-17 by Cynthia Higgins

Cynthia, how do you handle it when a "minor" incident consumes the entire budget for the month, but the business still wants to push a critical, high-revenue feature?

   Answered 2025-11-19 by George Hall

• George, that’s where the "Policy" comes in. If the budget is gone, the policy says no new features. If you break that rule once, the budget becomes meaningless. However, you can negotiate a "one-time exception" if you also commit to a "Reliability Sprint" immediately afterward. The key is making the data visible to everyone so the trade-off is a conscious business decision, not an engineering complaint.

     Commented 2025-11-20 by Cynthia Higgins

We use SLIs (Service Level Indicators) that actually matter to users, like checkout success rates, rather than just "server is up." It makes the conversation much easier.

   Answered 2025-11-22 by Donna Hayes

• Exactly, Donna. Business leaders don't care about CPU usage; they care if customers can buy things. Focusing on user-centric SLIs is the secret to getting executive buy-in.

     Commented 2025-11-23 by Paul Rivera