This post was originally posted back in 2016 – I decided to leave it live, but you can read more about whether AMP is really needed in 2020 here. A lot has changed over the years and having AMP may not be the right decision for you!
A special thank you to Carol at Fidose of Reality for being our “guinea pig” for our AMP case study and allowing us to show her data so you can see the impact AMP can have!
Update June 2020: Carol kept AMP on her website until the end of 2019 at which time we removed it because we heard that it could affect ad earnings. Upon removing AMP – her RPM and monthly earnings did increase greatly.
This is case study to give you an idea of what CAN happen if you begin using AMP on your website – this is not a guarantee it will be as successful for you due to the many variables involved.
The data here is a low level overview of one website over a very short time.
AMP Case Study Back Story
First we clearly need to see what was going on before AMP was included on the website – this way you can see the “before” and “after”. You of course will not need to do this – we just want to prove how quickly this can happen for you!
First, you should have proof that there were no AMP pages indexed on the website (we would have preferred to use the Search Console for this, but it was not hooked up prior to our installing AMP – so we are using Analytics instead).
- We visited Google Analytics > Behavior > Site Content > All Pages
- We changed the dates to November 1 – November 30, 2016 to only give us results for that time frame since we did not install AMP until December 2, 2016.
- In the SEARCH BOX we typed in /amp – this will now return any pages that had visits that had AMP appended to the end of the url.
You can see in the screenshot below that during November there were absolutely no pages for the /amp query.
Adding AMP To The Website
December 2, 2016 we implemented AMP on the website.
Once we added AMP and customized it the way we wanted to – we went into the Search Console and submitted a new sitemap so that Google can see quickly that the site had some new stuff going on.
Then we waited.
Now waiting can mean a couple of days – or a couple of weeks. It depends on your website, the coding, the traffic etc. Fidose is a fairly large site that gets a lot of traffic so we do believe that helped them get their AMP pages indexed a little quicker – actually a lot quicker.
The first day we installed AMP they were getting traffic to those pages – you can see the screen capture below for December 2, 2016 – with the /amp search query. You can see the had 14 AMP pages indexed the same day (possibly more – but 14 pages received traffic that day).
What Has Happened Since Installing AMP
Now let’s take a peak at what has happened since we implemented AMP until the date of this original post (12/8/16) – you can see below that they now have 115 AMP pages that have received traffic in just 6 days.
(In Google Search Console it shows a total of 281 AMP pages being indexed!)
Summing Up The Numbers
Improved Mobile Click Thru Rates
The click thru-rates on AMP pages are often much higher than traditional pages.
We took a couple of samples from the Search Console of AMP pages that are receiving traffic so we can compare the AMP version versus the regular version.
EXAMPLE 1: You can see for this post there was a total of 237 clicks for the dates December 2 – December 6, 2016. You can see that the AMP version received many more clicks and was shown a lot more in the search results (impressions).
EXAMPLE 2: In this example there was a total of 184 clicks for same dates. You can see that the CTR for the AMP page is 2% higher and resulted in more clicks. What is really interesting here is that the impressions (or times the page was shown in the search results) – was actually just a bit lower but they still received more clicks.
More Traffic
Below is a screenshot that just shows one example of a page from the website. This report is for the same dates (12/2-12/6/16) and is only pulling data from ORGANIC TRAFFIC (so we excluded referral, social and direct traffic sources) . You can see that the AMP page version is almost 40% of all the traffic for this page – while the general page shows a decrease in visit for mobile users.
We also included the device column so you can see what traffic each specific device type brings in terms of traffic. Because this website has almost a half of their overall website traffic from mobile – we knew the AMP would benefit them – and we are thrilled with the results after only 6 days!
Improved Page Speed
If there is anything you take away from this is that website visitors want a FAST experience and that is what AMP pages do. Below you can see the exact same page compared in terms of time, page size and resource requests.
The NON AMP PAGE – or the original page took almost 11 seconds — has a whopping 8.77MB of information to download and pulls in 260 requests from other websites — that is a lot to ask of a mobile visitor.
The AMP PAGE – as you can see greatly reduces this by removing a lot of the “fluff” of a website and skimming it down the basics for the fastest loading experience.
FINAL UPDATE – 1/19/17
So we wanted to do one more website update to look at the AMP results. We implemented AMP on December 2, 2017 and are going to do a review from the full month of December 2016.
Below you will see a screenshot that shows the AMP page views for the full month as well as the fact that they now have 311 AMP pages indexed.
The below screenshot is from the Search Console – we ran a comparison from December 1 – December 15th compared with December 16th – December 31st so we can see the growth. We also are showing this so you can see that the original URL (the non-AM
P page) is getting less traffic because of the AMP pages – the AMP pages are generating additional traffic. You can also see the AMP pages are doing very well in organic search results.
Final Thoughts
So I hope you found this AMP case study interesting! Clearly the results for this client were good – but results could vary!
I am often asked if “AMP is good for our website”? Here are our general thoughts!
1) If a large part of your traffic is mobile traffic – implement AMP!
2) If you do not have a lot of traffic or minimal traffic then no need to implement AMP
3) Image and video heavy websites probably do not need to bother with AMP right now.
4) If you have a lot of keywords ranking well whether your traffic is low or high, AMP might have some value.
We have implemented AMP on several websites over the last few months and we have been very happy with the results.
We will add to this post with a future edit to see how well the website continues to do!
Have you implemented AMP yet? If so, what are your thoughts?
Great to see a solid case study to move much of our guesses into the realm of how Google regards AMP pages. I too, have implemented AMP on several sites and find it has been a bit up and down, but more up overall. It is still so new. As Google continues to test and make improvements, we expect to see it of significant value for many niche sites whose traffic is largely mobile. For others, going AMP may mean that they gain better mobile user engagement.
Thanks @disqus_EFVpekJGou:disqus — we have seen a couple of websites that did not have that big affect, mostly smaller sites that had minimal organic rank to begin with. But if a website has rank — it is so worth it from what we have seen. It will be interesting to see where this goes! 🙂
I had my eye on AMP last year and after seeing Carol’s blog post on BlogPaws.com and your case study I decided to take the plunge. Here’s a little about my blog and my results. To answer your questions:
1. A large part of my traffic is mobile.
2. Traffic to my site is pretty good.
3. Not an image or video heavy site.
4. I have a lot of keywords ranking well.
AMP has only been up on my site for 6 days and here are my results compared to the trailing 6 days:
Sessions have increased 11.21%
PageViews have increased 6.35%
However, on the not so good side my bounce rate is up 4.29%. Also, after peeking in my Search Console I can see that the AMP pages have a higher position, but the CTR is mixed. Some page CTR’s are up and some are down. AdSense has been stripped from all the AMP pages so that revenue is also obviously down.
One questions if you don’t mind. What can I do to decrease my bounce rate on AMP pages? I’m assuming the increase in bounce rate has to do with the fact that the sidebar, navigation and footers have all been stripped from the original page. Is there anything you have done to mitigate the bounce rate?
Overall I’m enjoying positive results from my AMP experiment. One more question: do you know how I can get my WordPress pages converted to AMP? Right now I’m using the AMP plugin by Automattic to convert my posts to AMP. Thanks and let me know if you have any questions.
Hi Colby, those are pretty good results! I would not worry too much about the bounce rate, by nature blogs have high bounce rates and 4% is not that bad of a difference. Also depending how your search results for AMP pages are being shown, they could be “swiping” through articles which could show as a hit……even if they are not really viewing the full page. There are just too many variables that can cause bounce rate issues…..it is hard to pinpoint without knowing your site well.
But do make sure you do have a menu on your amp pages so people can click through to other pages if they so wish. You should also have a “go to regular” website link somewhere…usually in the footer to try and encourage people to go to your regular website. I would focus more on time on the page since that is what is more important here — are they really taking the time to read the article?
Also, you can add adsense to your AMP pages…I am assuming you are using WordPress – if so you can use Advanced AMP ads plugin to do this.
Thank you. I have a navigation menu at the top and a “go to regular” website link on the bottom of the pages. My avg time on page is up 4.73%, but my avg session duration is down 13.89%. So it does seem they are taking more time to read the page, but are clicking to other pages less often. I’m going to see if I can work on improving my bounce rate and CTR. By the way, I loved the case study. Thank you for sharing.
Looks like your Bounce Rate is insanely high – 95%. There’s no point in having more visits if they don’t convert or even see 2 pages of your site. You might want to reconsider if enabling AMP was worth it.
Hey Chris! Great point – but this is a blog which typically have high bounce rates so this was not a concern for them. I would not recommend AMP for a service base business that needed to worry about conversions – but for a blog with a lot of scripts and code it made sense as it greatly decreased load time. We only recommend it for sites that have a high load time even with mobile and just cannot get it any faster.