A bit over a year ago I decided to disable Accelerated Mobile Pages (AMP) Project on this site. I have a long history with AMP and at some point I was one of it’s main critics. I’ve since accepted that it has a role to play, with no valid alternative. At the same time, I am happy to NOT have to support it on this site.
I did a deep dive into my site’s analytics one month after I disabled the AMP where I did not see any meaningful (negative or positive) impact.
A year later results are roughly the same. My site is doing just fine and I get plenty of traffic from Google.
The following traffic chart shows that my Google traffic does not look any worse than the traffic from a year before.
It’s a bit hard to see due to one big day with a lot of traffic, so I did a zoomed-in version for only 6 months out of the period.
Blue line represent traffic without AMP and orange line represents traffic with AMP. As you can tell, my search traffic has improved since I disabled AMP. A lot of it can be attributed to me writing additional articles in the last year. Still, I feel very confident that disabling AMP did not have a negative impact on my search traffic.
Disabling AMP also took the frustration around it out of my life, so it was a good choice for me. I would highly recommend small time bloggers and publishers like myself to avoid wasting their time on AMP.
you are actually looking pageviews. amp doesn’t really provide so much pageviews. i recommend looking into sessions or users before deciding whether it has an effect or not.
Just keep your mobile site fast and you don’t need amp. There are many non-amp ways to keep a site fast.
I’ve ran your article through WebPageTest* and Lighthouse, and I’m getting some pretty high Time to interactive numbers (~15 seconds; the recommendation is to keep it under 5 seconds). I’d recommend checking if this can be improved.
*https://www.webpagetest.org/result/180803_4Y_b81b2d9a9b572aca98d81179a3741576/
WebPageTest and similar tests actually don’t measure the speed thy measure some code performance on which thy can predict speed potential (which is quite wrong). Do you really think that this blog is so slow ~15s?
I think 15 seconds is a bit high, but my site does have a lot of work to do in terms of optimizing first load on mobile devices. Following loads are not as bad.
I did not feel like WordPress AMP plugin solved the issue much better. Personally as a user I prefer to wait a few seconds longer to get non-AMP version of the site, unless the site is full of annoying ads and pop-ups (which my site is not).
The situation with WordPress AMP plugin reminded me of this joke: