3 ms·

I implemented AMP on my own project and we implemented it at some clients ... all regret it. The bounce rate, time on site, average page views of AMP pages is

by franze 10y ago

I implemented AMP on my own project and we implemented it at some clients ... all regret it. The bounce rate, time on site, average page views of AMP pages is always worse than on the responsive version. And as we now have responsive webpages + an AMP version to maintain the overall project costs got higher. Additional the AMP pages are already falling behind as devs and management just hate them. And outside of the news vertical we haven't seen any major positive traffic impact (and as stated above, the usage values of the AMP-traffic is crap.) Also we spoke with some users and they all were confused of "what happened to our site...". AMP is a horrible idea with an even worse implementation.

fixermark 10y ago
Question: Why do you not then just switch off the AMP support and stick with responsive pages? It sounds like you just generated an A/B test with some pretty easy-to-interpret results.
- gcb0 10y ago
  because nobody used amp because they assumed it was a good idea. it's 100% because google gives better seo position if you use it. so no test needed to know it's garbage. everyone already knew.
- franze 10y ago
  a) for deleting amp you need to implement some ping logic https://developers.google.com/amp/cache/update-ping#remove-amp-content https://developers.google.com/amp/cache/update-ping#remove-a... b) deleting page(variations) big scale is always a traffic risk c) deleting AMP it's already discussed (by literally everybody), so it will happen (soon...ish) d) the one reason not to do it is currently market monitoring. even though there is currently no major traffic benefit doesn't mean there will not be one in the future. so it's just a question of when the pain of AMP gets greater than the market-surveillance / curiosity / traffic bet. Basically everybody who currently has AMP pages is the experiment guinea pig for the whole industry. My bet: Most of the AMP implementations won't survive 2018.
zlynx 10y ago
Did you consider that higher time on site may be a false indicator? It could be your users being forced to sit and wait for complete page loads. Or time spent trying to find the real next page link.
chickenfries 10y ago
Absolutely agree. I work in publishing, our readers are confused by the changed design of the site (Who the hell knows about AMP or Instant Articles besides people in the industry? The average reader doesn't care). We have to maintain variations of article templates for our responsive website, AMP, Facebook Instant Articles and Apple News now. Product owners and editors think implementing these things is cheap and easy because of all the marketing but unless you have something like a basic wordpress setup, it's plenty of work to implement and adds a lot of things to maintain and requires developers to remember which subset of HTML AMP uses, what the JSON API for Apple News is like... at least Facebook Instant Articles use rss...

by franze 10y ago

fixermark 10y ago

Question: Why do you not then just switch off the AMP support and stick with responsive pages? It sounds like you just generated an A/B test with some pretty easy-to-interpret results.

gcb0 10y ago
because nobody used amp because they assumed it was a good idea. it's 100% because google gives better seo position if you use it. so no test needed to know it's garbage. everyone already knew.
franze 10y ago
a) for deleting amp you need to implement some ping logic https://developers.google.com/amp/cache/update-ping#remove-amp-content https://developers.google.com/amp/cache/update-ping#remove-a... b) deleting page(variations) big scale is always a traffic risk c) deleting AMP it's already discussed (by literally everybody), so it will happen (soon...ish) d) the one reason not to do it is currently market monitoring. even though there is currently no major traffic benefit doesn't mean there will not be one in the future. so it's just a question of when the pain of AMP gets greater than the market-surveillance / curiosity / traffic bet. Basically everybody who currently has AMP pages is the experiment guinea pig for the whole industry. My bet: Most of the AMP implementations won't survive 2018.

zlynx 10y ago

Did you consider that higher time on site may be a false indicator? It could be your users being forced to sit and wait for complete page loads. Or time spent trying to find the real next page link.

chickenfries 10y ago

Absolutely agree. I work in publishing, our readers are confused by the changed design of the site (Who the hell knows about AMP or Instant Articles besides people in the industry? The average reader doesn't care). We have to maintain variations of article templates for our responsive website, AMP, Facebook Instant Articles and Apple News now. Product owners and editors think implementing these things is cheap and easy because of all the marketing but unless you have something like a basic wordpress setup, it's plenty of work to implement and adds a lot of things to maintain and requires developers to remember which subset of HTML AMP uses, what the JSON API for Apple News is like... at least Facebook Instant Articles use rss...