Accelerated Mobile Pages (AMP) is Google’s technology for creating faster mobile web experiences.

I have been using AMP on my websites for the past few years for one reason only: SEO.

While Google has said that they don’t favor AMP pages, I had reason to believe that AMP would give me an edge with mobile rankings. First off, Google isn’t always entirely clear about the factors in the search algorithm, and certain “unconsidered” factors may be considered by the next update. Second, even if Google doesn’t directly favor AMP pages, they may favor the supposed benefits of AMP pages (i.e. faster load times, better user experience, etc.)

Using the AMP for WP plugin, it is relatively easy to create an AMP version of your website within minutes, so why not, right? The work is minimal so even a small edge in SEO would be worth it.

That has been my logic for the past few years, but throughout that time AMP seems to have more downsides than upsides.

The one benefit is cleara chance to increase rankings (i.e. using AMP for SEO). That’s really the only reason I use AMP.

The Downsides of AMP

Over time, the downsides start to overshadow the sole benefit. Here are a few:

  1. Technical issues related to AMP
  2. Design issues related to AMP
  3. Lower conversion rates on AMP
  4. Redundant work (making changes on pages and their AMP counterparts)

The biggest issue for me is that, overall, AMP pages are worse than regular pages. They don’t perform as well when it comes to generating conversions (which is the main goal of my sites). Furthermore, my team and I have to do double work when optimizing AMP pages. We may use one ad management tool for regular pages and another for AMP pages. We may use one email optin tool for regular pages and another for AMP pages. The list goes on and on.

I dealt with these issues for a while because I felt that AMP gave me a small EDGE in SEO.

Today, I’m putting that theory to the test with two of my sites.

Does AMP Help With SEO and Is It Worth It?

I am disabling AMP on two of my smaller sites to determine if AMP is really worth it. I will be looking at:

  1. The SEO Impact – Does disabling AMP hurt SEO?
  2. Conversion Rate – Does disabling AMP improve conversion rates and website profits overall?

Search traffic is only useful if you can monetize it. If hypothetically, AMP increased traffic by 10% but decreased conversions by 30%, it’s not worth it.

That said, I was always nervous to disable AMP and see a drop in rankings. Time to put it to the test.

AMP for SEO Case Study: Background

I am removing AMP on two sites, both of which get ~30% of pageviews on AMP pages.

AMP Pageviews

Before removing AMP, I wanted to make sure site performance and Core Web Vitals scores wouldn’t suffer.

I obsess over site performance so my sites are already highly optimized.

Ironically, the non-AMP pages performed better on mobile than the AMP versions.

Here is the AMP pagespeed score for a sample page (blog post):

AMP Pagespeed Insights

Here is the non-AMP pagespeed score on mobile devices:

Looking for tools that will make you a better marketer? Check out this hand-curated list of 50+ digital marketing tools that will take your marketing to the next level (seriously!) Check out the List
Non-AMP Pagespeed Insights

This is an important prerequisite to the test. If your non-AMP pagespeed score is much lower than your AMP pagespeed score, you may not want to disable AMP, as performance will drop and SEO results may suffer.

You may also want to analyze your own metrics. You may find that AMP is beneficial for your site. For example, on some sites I saw that the average time on AMP pages was double that of non-AMP pages. On other sites, I saw that time spent on AMP pages was 50% lower than non-AMP pages.

Removing AMP should be a purposeful decision driven by the goal of increasing your bottom line.

The Process of Disabling AMP

I use AMP for WP so disabling AMP is quick and easy. This guide provides a good explanation of how to do it.

Basically, you need to remove the reference to AMP pages from your main pages (i.e. rel=”amphtml”) and noindex your AMP pages.

Disabling AMP for SEO

Theoretically, Google should remove the AMP pages from their index and replace them with their non-AMP counterparts. I say “theoretically” because that is the point of this case study. I am trying to figure out what the repercussions of disabling AMP are before disabling it on my other websites.

AMP for SEO Case Study Results

I disabled AMP on August 12, 2021. I will continue to report back in the coming months as data rolls in.

I will be looking at:

  1. Overall site traffic
  2. User experience
  3. Conversion rates
  4. Overall profitability

1 Month Update (September 2021)

It has been one month since I disabled AMP and I am happy with the results so far.

AMP pages have been removed from the Google index.

Indexed AMP pages

Traffic has remained stable as Google is now redirecting users to the non-AMP pages.

Webmaster Tools Traffic

Here’s an example of how this looks for a specific page:

AMP Removal Traffic Case study

I have started removing AMP from some of my bigger sites where the stakes are higher. I am still not removing AMP from all sites at once. I will be testing the waters progressively with sites getting 10k monthly visits, 20k, 50k, 100k, and so on.

4-Month Update (December 2021)

A lot has happened in the SEO world since August 2021 when I initially launched this study, including multiple core updates.

At this point, there are no AMP pages in the index.

Here is the traffic graph (blue line is AMP traffic, orange line is non-AMP traffic):

amp case study december

It is clear when AMP was disabled (August 12, 2021). The AMP traffic drops off right away and the non-AMP traffic increases.

  • Total Search Traffic from July 12 – August 12 was 5,605 (pre-AMP removal)
  • Total Search Traffic from August 13 – September 13 was 5,235 (1-month after AMP removal)
  • Total Search Traffic from September 13 – October 13 – was 5,585 (2-months after AMP removal)

So, basically, traffic remained about constant and AMP traffic was redirected to the non-AMP pages. This was the best-case scenario and makes a strong case for removing AMP from your site IF it is causing some of the issues discussed earlier in the case study.