Making an application scale and go faster is often seen as a wizardly task. We read the micro-optimisation tricks posted in tech blogs and apply them with unconditional trust and great hope, and then wonder why performances haven’t improved that much (“Wait, I even replaced ‘print’ with ‘echo’!!!”). In this talk we’ll see how we can take easy, practical steps we can apply over and over that really make a difference, by analysing what our application does under the hood, measuring how and where the different resources are used, eliminating the real bottlenecks and restructuring critical components to handle growing loads.