Successfully, and efficiently, delivering a site to visitors involves a lot of moving parts. The server has to spit something out, which gets passed over the network to the requesting device. The browser takes over from there, coordinating with the device to try and deliver the page as quickly as possible. When things go well, we barely notice and if we’re not careful, neither do our metrics.
But there is a lot of unpredictability involved and if even one step is out of sync—if the device is overtasked, the network spotty, the browser old or unexpected—the performance can suffer dramatically. Welcome to the long-tail of performance, where the real world happens.
In this talk, we’ll discuss what happens when the web bumps up against this less-than-ideal day to day reality. We’ll zero in on these “long-tail” issues and explore how to ensure your sites perform well even when things go wrong. And as we address these issues, we’ll end up making our sites more performant for everyone in the process.