as someone who tracks rank on a daily basis and in great detail. i can tell you that rank [and traffic] changes can very clearly be attributed to specific event

as someone who tracks rank on a daily basis and in great detail. i can tell you that rank [and traffic] changes can very clearly be attributed to specific events or actions such as new inbound organic links or new on-page changes. for web properties that have few changes which occur daily, it is very easy to see the direct impact from deploying something specific within the following week or two. there is no ambiguity as to where these things come from. our rank is very steady across thousands of keywords over time and any statistically significant changes are never a mystery. to say that the boosted position/visibility of AMP can somehow get lost in the "noise" of rank fluctuations from other factors is demonstrably false. i dunno what kool-aid you guys are selling over there, but i'm certainly not buying it. the mission of AMP is fully Google-centric, not user centric. it is possible and easy to build pages faster than AMP without any of AMP's additional artificial limitations. AMP's carrot is that you get much greater visibility and faster loading as a result of pages being hosted on google's servers and loading via a pre-established TLS/TCP session. you can read more here: https://www.reddit.com/r/javascript/comments/78fevc/amplify_automatically_convert_plain_html_to/ https://www.reddit.com/r/javascript/comments/78fevc/amplify_... real-world AMP pages are not what they're billed to be: https://imgur.com/a/OPVdN https://imgur.com/a/OPVdN https://imgur.com/a/mt6s6 https://imgur.com/a/mt6s6