97 Things Every SRE Should Know

Chapter 86. The Performance Emergency Fund

Dawn Parzych

LaunchDarkly

SREs rely on concepts such as error budgets to manage changes across the organization, whether that means determining whether a release can move forward or identifying where to make improvements. Error budgets are related to availability, but of course, you need to know not just whether you’re available but also the quality of that availability.

If you’re not thinking about the quality, then you’re only getting part of the picture—but how do we define quality? I’d argue that one of the most important ways is through performance. How do you feel when it takes what seems like forever to load? Or performance on your smartphone is slower than on your laptop?

Everybody in an organization should care about performance. If a site loads too slowly or inconsistently, you run the risk of lost customers and lost sales. Nobody wants that.

Numerous studies have shown that faster-loading pages result in higher revenue, increased user engagement, and a decreased bounce rate. Slow-loading pages can also be an early indicator of a problem. Catching a problem due to slowness is better than waiting for a failure.

Just as you can create an error budget, you can do the same for performance. A performance budget is a clearly defined limit on performance metrics used to guide design and development. You can have multiple budgets for ...