Core Web Vitals – the beginning of the end for AMP?

Core Web Vitals – the beginning of the end for AMP?

This week Google announced a May 2021 rollout for incorporating their Core Web Vital metrics as page experience signals in search results.

Core Web Vitals is an evolving set of metrics for measuring key aspects of the user experience on any given page, including performance. This aims to provide a clear, concise set of guidelines for sites to follow ensuring sites deliver better experiences for their users.

Initially, Core Web Vitals includes three metrics measuring loading, interactivity and visual stability – these metrics are Largest Contentful Paint, First Input Delay and Cumulative Layout Shift. I'd consider these as giving you a tip-of-the-iceberg view into your site performance rather than a complete view, but it makes for a good starting point.

What does this mean for search results?

Starting in May 2021 Google will introduce three changes to search results – a change to ranking, the "Top Stories" space and a new visual indicator.

Ranking change

Page experience using Core Web Vital metrics will be added to the mix when it comes to signals used for ranking search results. Google's documentation states "Great page experience doesn’t override having great page content.", so page content is still the most important aspect of your page when it comes to ranking. But, now where two pages have a similar rank based on the page content then the page experience will influence the search results.

Don't forget that optimising your page experience has a bigger impact beyond search too – performance and user experience have a huge impact on user engagement.

Core Web Vital metrics measuring loading, interactivity and visual stability join existing signals for mobile friendly, safe browsing, HTTPS and no intrusive interstitualsls
Core Web Vitals joining other page experience search signals

Top stories

Search results on mobile often display a "Top Stories" carousel which, until now, have required pages to support the AMP framework in order for them to qualify for this prime spot on the results page. With Core Web Vitals this AMP requirement will be removed – pages which met the Core Web Vitals criteria will be eligible for the carousel regardless of the technology used to implement it.

Carousel of 3 pages on Google search results within a carousel
Example of a top stories carousel showing AMP pages

Visual indicator

Google have also announced that they will be experimenting with a visual indicator on search results to highlight pages that score well when it comes to page experience. Exactly what form this will take is unknown, but an obvious point of comparison can be made with the existing AMP ⚡️ indicator which shows up against AMP pages within search results.

Screenshot of a Google search result showing the the AMP documentation website result with a lightning bolt icon after the site name
Example search result showing the AMP lightning bolt icon displayed after the site name

Why could this spell the end of AMP?

Now that the technology-agnostic Core Web Vitals initiative provides an alternative route to "Top Stories" and a visual indicator on search results without using AMP, does this mean we no longer need to support AMP pages anymore?

I think it's fair to say AMP has a significant image problem. Despite efforts by the AMP team, it is still perceived as an effort by Google to force publishers to follow a very specific set of Google-approved standards and services in order to obtain prominent spots on search results pages and gain access to automatic preloading and caching.

AMP was introduced at around the same time as the likes of Facebook Instant Articles and it's through that viewpoint that most people have approached AMP – as a lightweight alternative mobile version of their full pages done for the SEO benefits rather than as a choice by developers.

The AMP team have tried to shift this perception, ditching the "accelerated mobile pages" definition and distancing themselves from Google. They now describe AMP as "a web component framework to easily create user-first websites" but it has remained hard for AMP to shake its original image.

Screenshot of the AMP website with the headline "AMP is a web component framework to easily create user-first websites."
Screenshot of the AMP website in November 2020

Implementing AMP as an alternative mobile version adhering to different standards introduces a whole set of challenges keeping the two versions in sync with new functionality and maintaining consistency that hasn't made it popular among developers. In many ways this has only become worse as AMP became increasingly feature-rich – AMP pages went from slimmed down versions of pages to fully capable if (and only if) using the building blocks supplied by the AMP framework.

The "accelerated" in AMP's name showed the important of performance in AMP's original pitch. The user-facing advantage of AMP was sold as being performance, with the "Top Stories" slot and visual indicators being your reward to delivering a performant page. However it's entirely possible to serve a performant page without using AMP but it's never been permitted (until now) for pages to compete based on performance alone. With the additional features added to AMP over the last few years it's no longer safe to assume that AMP is "fast by default" as it's still possible to overload a page and create a slower user experience which only further emphasises the importance of letting pages compete based on performance rather than framework.

So, what happens next?

My expectation is that we'll start seeing sites moving away from creating AMP versions of their pages, and instead investing in site-wide performance as part of Core Web Vitals. Users stand to benefit from an improved user experience regardless of how they access a page, whilst development teams can drastically reduce their maintenance effort.

Where AMP has been successful is in dangling the carrot of SEO as a motivation for building a faster version of our pages with an improved user experience. For many publishers, this was probably their first introduction to the idea that performance is critically important. For some sites, achieving the Core Web Vital benchmarks may not require any additional investment, but instead switching existing time away from AMP. This serves as a useful anchoring mechanism compared to starting entirely from scratch making Core Web Vitals feel like a cheaper investment.

However, a potential risk is that people optimise purely against the Core Web Vital metrics without then assessing if this does deliver the right user experience – there are still other metrics and factors which can impact your page performance. Use Core Web Vitals as a starting point, not a destination.

My hope is that this helps AMP start to change it's perception. Instead of being chosen by a business for SEO, could it now be recommended by developers and designers for its library of proven and documented components? The Bento AMP approach promises the ability to cherry-pick specific AMP components without committing to the full framework, allowing you to combine AMP with JavaScript frameworks like Vue or React.

I can't exactly call myself a fan of AMP, but it does provide a different approach to most other frameworks and has made multiple contributions to open web standards. This competition and contribution is great and should continue.

AMP and non-AMP websites will have to start competing with each other on speed to gain SEO boosts, whilst AMP components will have to compete on quality with components from other non-AMP frameworks. Hopefully, this competition translates to faster websites, improved web standards and a better user experience for all users.

AMP is dead, long live AMP? ⚡


Cover photo by Paweł Czerwiński on Unsplash.