The Financial Times product and technology team writes about site performance, tech strategy, team structure, and engineering practice at FT.com.
A former FT developer says the team kept FT.com fast using user-centered metrics. Another commenter doubts the performance claims, saying graphs can be picked to fit any conclusion.
I'm super pleased to see the FT towards the top of the charts. It's a couple of years since I worked on the project[1] to make it faster and it's really great the team have been able to keep it that way.
We bet on measuring site speed with user centered metrics early on which was going against the grain at the time - removing above the fold CSS, are you crazy!? It took a lot of demos to convince people that what we were doing was faster, and that they really needed to trust their own eyes even when the tools disagreed!
Keep it up, folks.
[1]: https://medium.com/ft-product-technology/designing-a-sustainable-front-end-toolset-for-ft-com-f37c59d27eeb
Graphs.
I was here https://medium.com/ft-product-technology/a-faster-ft-com-10e7c077dc1c when they did this.
Its easy to find graphs that support your conclusion.