What is Google AMP?
AMP, or Accelerated Mobile Pages, is a framework to build web pages introduced by Google in early 2016. Aimed to improve the loading speed for web pages on mobile, AMP’s benefits have been tempered by some major drawbacks that have slowed its adoption, despite the fact that Google has given priority in mobile search results to AMP pages.
To convince your users that sticking around is worth their time, it is important to give them what they want and expect, in the most effortless way possible; this is the reason we built the Branch deep linking service. If you’re a content producer, then AMP contains the promise of huge improvements to your user experience. Of course, if your website already scrupulously follows best practices and doesn’t contain garbage code, AMP likely won’t give you much. But for everyone else, this is a great opportunity to make a major upgrade with fairly minimal investment because a lot of the heavy lifting has already been done for you.
- This is an example of what reading in Pinterest Pin with AMP is like: https://branch.io/img/glossary/pinterest-with-amp.gif
- This is what reading a Pinterest Pin without AMP is like: https://branch.io/img/glossary/pinterest-without-amp.gif’
Visibility on Google and Others
Since the list of AMP supporters already contains a who’s who of the most important names in both tech and publishing (with the notable absence of Facebook and Apple, of course), it is probable that AMP-enabled pages will continue to get increasingly special treatment in future. Even if your site doesn’t need the speed and usability boost AMP offers, it may still be worthwhile simply to take advantage of the improved visibility for your content.
One of the drawbacks of AMP is the limited and strict framework that doesn’t present a mechanism for users to click to the content publisher’s web properties from an AMP page. Instead, the AMP page “steals a brand’s traffic” by always taking users back to Google search results. This concern was first voiced by Alex Kras’s post Google May Be Stealing Your Mobile Traffic and amounted huge backlash from the web and publisher communities alike.
The second wave of criticism was brought on by the publisher community, as AMP pages don’t monetize well. As Google started showing more and more AMP pages in search results, more publishers adopted the standard — upon adoption, publishers discovered that these pages don’t actually help them grow mobile traffic or revenues.
Google AMP pages are slowly creeping into the open web despite all the criticism. Over 2B mobile web pages were released by the end of 2017. While it will be very hard for individual organizations to push back on Google for such a dramatic change to the web standards, solutions like AMP Journeys are available for businesses to turn AMP traffic into mobile app installs and mobile user engagement.
How AMP Works Under the Hood
Since you’re still reading, you probably want to know how all this magic actually works. As mentioned, AMP pages exist in parallel to the traditional pages we already visit. This means that in practice the same content will often (if not always) be available in two different versions, connected together with a special tag in the header section of each one. Googlebot already uses this tag for indexing, and it’s conceivable that future mobile browsers (at least, those not named “Safari”) will start to look for it, too.
The specification itself is an evolution of familiar web standards, which come together to form three main ingredients, AMP HTML, AMP JS, and AMP Cache.
AMP HTML is essentially a form of traditional HTML. There are a number of custom attributes to specify, a few troublesome tags that are banned outright, and a sizable collection of new elements that are available to provide pre-configured interactivity without impacting performance. In other news, styling is still accomplished with standard CSS.
If AMP JS is the engine in your car, then AMP Cache is the interstate trucking company that delivers fuel to your local filling station. You won’t even need to think about it unless everything suddenly stops working. Under the hood, this part of the system improves performance by hosting all resources in a single location and optimizing their delivery. While technically optional (AMP pages are hosted on a normal server just like every other HTML file), every AMP implementation comes with an implicit agreement that all content must be available for crawling and caching by third parties. Anyone can use a CDN to set up and run an AMP Cache, but only content in Google’s cache — which Google has stated can be used without restriction and at no cost — is currently getting preferred search results treatment.
AMP’s Secret Speedy Rendering Sauce