I don't get the presidential part. Lots of statements about response time being better, then the guy with nearly double the load time won the election? Not that I'm arguing against page speed, just found that odd in the info graphic..
They are somewhat confused. In the Amazon section, they write "A 1-SECOND DECREASE IN AMAZON'S PAGE LOAD TIME WOULD ANNUALLY COST [...]" instead of "increase".
Back in the days I used Opera Mobile for surfing on the phone, when there were no data-plans or 3G/4G as there is now. A compressed page with pictures were just a few kBs, and loaded fairly quick.
But now, since the mobile net is so fast, there's little optimization done for mobile other than layout. Loading the largest newspaper in my country's mobile version is a whopping 1.5MBs with all the pictures. 2.7MB in normal. I think this is sad.
I was under the impression part of that is because of how Opera Mobile worked, they have some kinda "Turbo" mode that compressed websites and reduced downloading times.
I've just left my current mobile phone provider (Vodafone) because of the continual problems with their website. Throughout the call to request my PAC (unlock) code every employee I spoke with didn't believe me when I said that I wanted to leave because of their site. I was offered increasingly more desperate financial 'offers' without acknowledgement of my actual complaint. Some big business' still don't relate web performance with user happiness - disappointing.
Interesting blip in the "Conversion Rates by page load time" graph. Wonder what the reason is for the conversion rate for 0-1 seconds being markedly lower than 1-2, 2-3, and 3-4?
Also, that graph really irks me. Not enough contrast, poor layout, and confusing axis labels which depend upon color. Also, is the 0.025 on the y-axis a percentage or a decimal (I think it means 2.5% conversion rate?)