Posted by
Mordy Oberstein
Back in 2016, AMP, Google’s open source project to make a speedier mobile
Here’s why I think AMP has hit a wall.
How Often Does AMP Show on the SERP?
Here’s the plan. Let’s look at some of the data related to AMP’s presence on the SERP and then get into some of the ways AMP has developed and evolved over the years. With that, we’ll be in a good position to see why I think AMP has hit the proverbial wall.
AMP was birthed within the news media universe. News publishers were the first sector to receive the gift of AMP. While initially open to anyone, Google, following in Facebook’s Instant Article footsteps, was all but entirely focused on gathering publishers under the AMP label.
It makes sense then, that since January 2017, AMP has simply dominated the News Box on the mobile SERP. At the start of 2017 Google dramatically increased AMP in the Top Stories carousel the world over. Since then, markets around the world have seen the majority of the results within the news carousel be those with the lightning bolt AMP label.
In the US, anywhere from 60% to upwards of 70% of news carousel cards contain the AMP label!
A clear and international increase in AMP within the News Box took place in 2017 and has continued ever since
While AMP has made its mark inside of the news carousel, the same is not applicable to AMP within the organic results. According to our SERP Feature Tracker, AMP entered the organic fray back on September 20th, 2016. Six months into its organic life, AMP results were found on about 15% of all page one mobile SERPs.
It was only after a year on the organic SERP that we tracked AMP results on 20% of page one SERPs. And there it has stayed for quite some time. Even as of the end of July 2018, organic AMP results appeared on only approximately 25% of page one SERPs.
While incrementally the jump from 15 to 25 percent reflects a 67% increase, AMP showing within just 1/4 of all page one SERPs is nothing to brag about. Certainly not for an initiative with the full force of Google behind it (more on that later).
What’s most telling though is the average number of AMP results found on page one. That number stands at about 1.7 AMP results as of end the of July 2018. Do you know what the average number of AMP results found on page one was back at the end of September 2016?…. 1.2!
From the very month AMP entered organic results all the way until July 2018, the average number of organic AMP results found on the page has increased but .3 percentage points!
Organic AMP data by the % of page one SERPs that contain at least one organic AMP listing along with the average number of organic AMP results on page one
The average number of AMP results found on page one since 2016 is really an odd data tidbit when you consider that AMP amidst the organic results has grown over the years, as evidenced by the percentage of SERPs that now contain at least one organic AMP result relative to 2016. The disparity (i.e., the percentage of SERPs with AMP having increased, but the average number of AMP results not moving), points perhaps to a limited number of sites having adopted AMP.
That is, the data points to bigger sites that rank well for a wide variety of queries having adopted AMP and not a host of sites across the web over having gone AMP (more on this later on as well).
Think of it like this: The commerce giant eBay is one of the sites that has adopted AMP. For all of the keywords where the site ranks on page one, and where the page has been optimized for AMP, the AMP label will appear on the SERP. The wide array of content that the site ranks well for will produce page one SERPs with the AMP label showing within the organic results. However, how many times will the site generally rank on page one for a given keyword? Once (maybe twice).
eBay is to my knowledge the only commerce juggernaut to have adopted AMP. Amazon did not go the way of AMP. Neither has Target or Walmart. So imagine the day eBay’s AMP goes online. The percentage of pages that contain at least one AMP result surely goes up, but does the average number of AMP results? No.
The data clearly shows that AMP had hit a wall! That is until…
AMP on the Rise in August of 2018
Suddenly, without warning, and after years stuck in the mud, AMP spiked. On August 22nd AMP went from having a presence on 24% of mobile page one SERPs to 33%, a dramatic 38% increase! By far the greatest one-day increase organic AMP had ever undertaken. AMP seemed to be headed on a new data trajectory that had it hovering at about the 33% mark. However, that all changed on September 6th.
Early September saw the speedy page format jump yet again from a 33% showing to appearing within 37% of page one SERPs.
While the percentage of page one SERPs has seen AMP become more prevalent, the number of organic AMP results on those pages has lagged behind a bit. Prior to the spikes discussed, and as previously noted, the average number of organic AMP results on those SERPs that contain them to begin with, stood at 1.7 as of the end of July 2018. However, even after the increases, the average still stands at around 1.7 as you can see below:
So AMP is back! No. Not in my opinion. Not at all. Let’s go into the ins and outs of how AMP has developed and return to this data in a bit.
Why Does Google Want AMP so Badly?
The first question we have to ask is why Google even wants AMP to begin with. If we want to know why AMP is seemingly going nowhere fast, this is the place to start. For that, let’s just go with what Google themselves have said:
“Speed has always been a cornerstone of Google Search and we’re thrilled to be working alongside others in the tech industry and publishers around the world to bring the Accelerated Mobile Pages Project to life.” (Dave Besbris, Google VP of Engineering)
At its core, and despite some of the other optics (more on that very shortly), AMP, at its heart, is meant to make the mobile web faster and easier to navigate. We all remember sites that would load awkwardly on mobile with ads overlapping content and link buttons placed at random… AMP was created so that we wouldn’t have to deal with that anymore.
The Other Optics Behind AMP’s Creation
While I have to believe Google created AMP so as to help move the mobile web forward, there are other and less than desirable optics out there. Due to the timing and nature of AMP’s coming onto the scene, some have speculated that Google’s AMP aspirations are less than noble.
I want to say at the outset, that while there may be a tinge of truth to these optics, I do not think they represent the main thrust behind Google’s AMP aspirations. I take Google at its word here. Sure some of the reasoning below was likely a slight factor, but I wouldn’t define Google’s goals here in those terms.
With that, here are some of the optics that I’m sure Google wishes would go away.
AMP Was Made to Rule the News Traffic World
As I mentioned above, the early days of AMP were aimed almost exclusively at news publishers. This created the optic that with AMP Google was merely trying to keep up with Facebook, who had launched Instant Articles. That is, a common perception is that Google didn’t want to lose user traffic to Facebook’s faster-loading news content. It’s the timing here, of course, that creates the optic. Facebook’s Instant Articles appeared in the News Feed in the fall of 2015. Just a few short months later, in early 2016, AMP made its debut. With AMP’s initial focus on news content, the connection was easy to make.
Cached AMP Pages – Who’s in Control Here Anyway?
There’s another less than desirable optic out there, and it stems from how AMP works. To load AMP results as quickly as Google does, the search engine caches the page. When doing so, Google initially showed their own URL when displaying such pages. Meaning, if you clicked on a news publisher’s AMP page, the URL at the top of the page itself, was Google’s, not the publishers. Only in January of 2018 did Google finally present a peace offering to publishers in the form of a plan that seriously dealt with the AMP URL issue.
The obvious optic here is that Google was interested in a power grab. Not showing the publisher’s URL was interpreted by many publishers as Google saying that the web, from the SERP down to the page, belongs to Google. The URL issue was a sore spot with publishers for quite a long time, and I don’t think they’re over it even today.
Is AMP an Open-Source Project?
Most everybody views AMP as a Google property. In reality, it’s an open-source project. Yet, no one really refers to it as such. We’re barely into this post and I’ve been guilty of referring to AMP as being under the Google umbrella multiple times already. Here again, we get into a bit of an “optic sticky point.” If AMP is an open-source project, then why is Google pushing it so hard? Why is Google so prominently behind it? I’ve seen estimates that up to 90% of the contributions to the AMP Project have come from Google engineers. You don’t have to be a political genius to realize how this might look.
How Google Has Pushed the World Towards AMP
It’s not my style to just put my thoughts and theories out there. I want it to be a learning experience, I want you to see not what I have to say, but how I got there. We’re not ready to hear why I think Google’s AMP has hit a wall. We have some good background in knowing why Google went AMP and the optics surrounding the move. Now let’s continue the build-up by looking at how Google has gone about trying to create its world of AMP.
A Plethora of AMP Options
In its task of getting the world hooked on AMP, Google, over time, has released batches of new AMP abilities. The first significant round of AMP updates came in October 2017, about a year after AMP first launched. This batch included elements such as:
- The ability to add “scroll-bound animations” which let you animate an element that overlays the page via scroll movement.
- A sidebar that helps users make their mobile page more responsive.
- Support for native video analytics.
Most recently, in May 2018, Google gave AMP sites a way to offer targeted content based on a user’s location. At the same time, the AMP Project put forward tools to help sites track user data while conforming to GDPR.
Other changes have included sending users who click on a Google Ad (formerly AdWords) to an AMP landing page and speeding up ad load time on AMP pages (more on AMP and ads in a bit).
The point is, Google has tried hard to improve ads and make its implementation more and more advantageous.
AMP Stories & AMP in Gmail
There’s one last “AMP element” that I want to discuss, and that’s AMP Stories. AMP Stories are
An AMP Story presents a series of highly visual content that is blended with text in order to tell a story. If you’re not familiar with the format, the easiest thing for me to do is just to show you what it looks like. Here’s an AMP Story from the Washington Post:
In July of 2018, Google took AMP Stories to a new level by offering a way to monetize them! Opening up the format to all, the AMP Project released an ad version of AMP Stories.
What AMP Stories represents is Google trying to make AMP both a powerful and trendy marketing device. The format brings AMP to users who are likely to consume Instagram and Facebook Story content. It’s
AMP Stories were not the only way Google sought to push AMP beyond page formatting per se. In the winter of 2018, Google announced that AMP was coming to Gmail. Long story short, AMP within Gmail was designed to make emails more interactive. For example, you could click on an image, see some deeper content and return to the previous “page” much like you can in a web browser.
The point is, here again, we have Google taking AMP outside of its original universe, and we have to ask why? Why is Google taking AMP away from its core and moving it into more peripheral areas?
Questions now,
AMP and Wordpress
30%. That’s the percentage of sites on the web that are powered by WordPress. As chance would have it, in February 2018, Google announced that it was establishing an engineering team to partner with WordPress with the goal of making the mobile web faster. It all makes good sense, with so many sites being powered by WordPress and with the CMS being competitively slower.
At the same time, the focus on mobile speed makes you wonder if this has to do with AMP. Well, you don’t have to wonder. Alberto Medina, Develop Advocate in the Content Ecosystems Team at Google, called on WordPress experts to help his team by “expanding capabilities of the AMP Plugin to enable delightful user experiences.”
How delightful! Here’s yet another optic issue for AMP. We have the giant that is Google sending an engineering team to the CMS that is responsible for 30% of the web to help push AMP. Optically, you’re almost forced to ask, at least in the back of your mind, if this is about a better and faster web, or about getting 30% of the web hooked on AMP?
Again, I’m not saying any ulterior motives were at play. Rather, I’m offering how the move optically looked to a lot of people. My point is not to call Google out on anything (especially since I don’t necessarily think they did anything wrong). All I’m pointing out is how the move made people perceive Google vis a vis AMP.
The Great AMP Debate: Site Traffic & AMP Ad Revenue
Let’s have a look at one last and very large piece of the AMP puzzle before jumping into the answer to our question (i.e, Has AMP hit a wall?). There’s been a large amount of both debate and controversy over AMP’s ability to earn revenue and gain site traffic. In simple terms, there’s been a real question as to whether or not AMP is actually good for sites.
The AMP Ad Revenue Question
As we already touched on above, there was an issue with AMP pages loading too fast. So fast that the ads could not keep up with the site’s content. As discussed earlier, Google tweaked things a bit in order to address this. Now AMP ads are able to keep up.
That said, AMP had ad revenue PR trouble out of the gate. You might have thought that the initial issues would be related to ads on AMP pages lagging as was just mentioned. According to the Wall Street Journal, however, the issue had to do with control over ad formats. As early as October of 2016, reports were coming out that publishers were not happy being limited in regards to what ad formats they could and could not utilize. For the record, many publishers were quite happy with AMP vis a vis ad revenue.
Forget the complaint. It doesn’t matter nearly as much as how Google responded. Instead of embracing the complaining publishers, Google retorted that with proper implementation, AMP should earn all the ad revenue in the world. I’m not a psychologist, but I’m pretty certain that’s not what these publishers wanted to hear. And so a continually contentious relationship between Google and news publishers began.
Again, whether or not AMP is or is not actually good for ad revenue is not the issue. The fact that it’s a contentious debate, the fact that AMP ad revenue is viewed as potentially problematic is the real issue. When dealing with AMP everyone argues back and forth about the technical abilities of the project. Who cares? We’re talking about widespread AMP assimilation (or the lack thereof rather), and that’s all about optics.
First There were Ads, Then There was Traffic – The AMP Debate Lives On
The debate over AMP’s ad revenue vitality continues to this day. Sure, Google has sped up ads and so forth. However, site traffic has become a “new” point of contention. Remember those publishers that complained AMP’s limited ad formatting is hurting them. Guess what? Years after AMP rolled out they’re still complaining about the same thing, except with a twist.
These publishers are not seeing AMP result in increased traffic. According to one study, only 1 in 3 publishers saw an increase in traffic as a result of AMP. So what? If traffic hasn’t fallen off, why is this a problem? Because an increase in traffic would supplement the lost revenue these publishers claim results from being limited in what type of ads they can show with AMP. With no traffic gains, all these publishers are left with is less ad revenue.
Why It’s All Led to AMP Hitting a Wall
OK, I freely admit, that was a ton of set-up. But in my totally biased opinion, it was time well spent.
At this point, we have a ton of context, a few questions, and the ability to formulate an answer as to why AMP has hit a wall. To get there, I’m going to ask you to take your SEO hat off and don the fine dress of a marketer, because I don’t care how awesome AMP is, how wonderful it is for the web, and how glorious Google is for implementing it. All I care about are the optics of AMP, true, not true… irrelevant. Many a product
Always Make a Good First Impression
Why does the data show that AMP has hit a wall? It’s pretty simple. AMP is that guy who picks up a **** wearing a **** n’ Roses tank top with a chrome chain tucked into his torn ripped black jeans which is topped off with a metal-studded leather jacket, ripped boots, and hair that looks like Keith Richards just got electrocuted (actually that’s kind of how Keith always looks).
The first rule of marketing (and ******, which is the ultimate form of marketing)… always make a good first impression. AMP did not do that. Google, to their fault, was not careful to make sure they were optically clean. Pulling out of the gates like a stallion on steroids, AMP gave off the impression that it was Google’s answer to Facebook’s Instant Articles. Then Google asked the whole world to adopt the format. You do see why that equation doesn’t work, right?
AMP as a way to change the underbelly of the mobile web and its success in doing so rests entirely on the idea that it is completely altruistic. AMP’s entire brand identity must be wholly pure if Google is going to ask the world to change. The timing of its release and the urgency Google placed on AMP’s head stole AMP’s sanctity. Instead of it being angelic, AMP was just another initiative by another big tech giant, and when was the last time any of those were holy?
If you want proof that what I’m saying is true look no further than the AMP Project. Why is this not a Google project? Why is this an open-source initiative? Why didn’t Google just go ahead and tell everyone they have to go AMP. If Google wanted an AMP only universe there are easier ways of creating that than an open-source initiative.
Big Dogs **** Control & AMP Took It Away
Not referring to Google here. I’m talking about the news publishers. Asking them to give up control over their site (asking is not the right word, get to that in a moment), is a non-starter. I’m referring to Google caching a site’s AMP pages and then serving up a Google URL.
Again, think of this from a marketing perspective. If you’re going to ask someone to give over the keys to the Maserati, you better come off as having the other person’s best interests in mind. But how “trustworthy” did Google come off from the start? Already there was a perception that AMP was more a Facebook foe than a hero of the mobile web. Imagine you see someone wreck their car, walk out of the wreckage, and then ask to borrow your car so they can get home… that’s what Google asked news publishers to do in a way. I’m obviously being a bit dramatic, but here comes Google, fresh off the perception that AMP is all about responding to Instant Articles, asking publishers to adopt the format knowing that their own URL won’t be placed in front of the user’s eyes. That’s a lot to ask even without AMP having a negative optic surrounding it.
I’m not saying Google did anything malicious. I’m merely pointing out how AMP may have appeared to already skeptical publishers. In all honesty, the only thing Google was probably guilty of was not realizing how AMP was coming off. Had Google been a bit more self-aware, I’d have to imagine they would have pushed AMP’s release off until the URL issue was resolved, as the last group of folk to hand over the keys to the palace has to be news publishers.
To make this worse, Google has essentially limited its news carousel cards to AMP results (think back to the data at the start of the post). This all but forced publishers to adopt the format. Again, another way Google took control away from publishers via something thought to be an open-initiative. More, it created a general impression among many that Google didn’t care if AMP was your “thing” or not. It came off to some as if Google was saying, “Like it or not, you are going to adopt the AMP format if you ever want any news article traffic.” An obvious problem.
AMP’s Performance Perception
On top of all of these poor optics, since AMP’s inception, there have been questions as to just how advantageous adopting the format really is. If you’re planning on creating universal adoption, having your “beta” group constantly complaining doesn’t help matters. Having this debate go on for years is a nail in the coffin. Imagine your brand had so much debate about how effective your product really is, you’d be pretty beat up. I would imagine it would be hard to sell someone on your services with PR like that.
But that’s exactly what Google did.
Calling All Sites – AMP for Sale!
This is the point where AMP hits that wall and hits it hard (in my opinion). After all of this. After all of these situations that have created “poor optics” for AMP, Google chugs right along and asks all site managers and
What have these people heard about AMP until this point? At a minimum, the general perception of AMP is a mixed bag of nuts. Between publishers complaining about ad revenue, site traffic, and control over their URLs, what’s your average SEO or site manager supposed to think? That’s not to say people won’t choose AMP anyway, they surely have. Rather it means that before they do some concerns need to be addressed. That already puts AMP at a disadvantage.
Pile on the fact the creating AMP pages is not as simple as it might seem, and AMP has itself a serious problem. If a site looking to optimize for AMP is already concerned about how things might turn out from a technical standpoint (i.e., will the page look nice and function correctly, etc.) having these sort of meta concerns doesn’t help things.
AMP isn’t the only game in town. There are other ways to make a mobile page faster (i.e., choosing a “faster” font, etc.). Add to the mix that going AMP may not be a “site’s” top priority and you can see why AMP hasn’t grown the way Google may have expected.
The Impact of Google’s Behavior, or Lack Thereof
Part of the reason why AMP hasn’t resonated is that Google hasn’t really done much in terms of damage control. Instead of dealing with the problem head on, Google’s sort of taken this course where it attempts to just showcase AMP and all it can do. It’s hard to blame them for doing so, I’m certainly not. That said, doing as such meant that the issues and optics have lingered on.
Instead of admitting that AMP has hit a wall, Google has tried to push AMP into new arenas. That’s why I believe you’ve seen things like AMP Stories and AMP in email. That’s Google all but saying, AMP has not gone the way we thought it would, we need to expand it through other means. And as creative as these means are, they are not the core identity of AMP and as such don’t have much of a chance of turning things around. How many AMP emails or AMP Stories have you seen? Exactly.
What Google Should Have Done with AMP
I think the overall mistake Google made was that it forgot that an initiative to change the mobile web is similar to a marriage, it’s an all-encompassing partnership. It’s easy for a mammoth player like Google, who lives and breathes the web, to forget that the “little guy’s” voice is a mighty big commodity in a project like AMP. The real issue is that Google never seemed to overcome the initial perceptions many held about AMP. Overcoming a poor brand perception is not an insurmountable task. The issue is, Google never tried to do so in a way that was connective.
Instead of interacting with the “AMP community” Google kept rolling out new features to solve problems, or to make AMP look more enticing. They never really sat down with the folks that mattered in any substantial way in order to hash this out, in order to foster a true open-project environment, namely one of partnership.
I mean all you need to do is look at how Google handled the AMP ad crisis… it published a post all about how AMP ads are awesome. But that doesn’t help to build relationships with the publishers that are already unhappy. Word of mouth is more potent than any blog post, I don’t care how “big” you are. Let’s take this to an absurd place. Imagine the CEO of the Wall Street Journal is having lunch with the CEOs of the NY Times and CBS News. During this power lunch, the CEO of the Journal says something like, “This AMP thing is not working for us.” To which the CEO of the Times says, “Yeah, but Google put out a really great blog post about how AMP ads do really work.” Do you think the CEO of CBS News who has yet to try AMP (in this imaginary world) is going to say, “Oh they have a blog post, it must be OK despite the CEO of the Wall Street Journal saying otherwise!”
There have been some gestures of trying to appease the “AMP community,” such as making AMP part of the open web standards. But even here, Google didn’t surge forward and engage the community. In fact, in this case, Google avoided commenting if the move to universalize AMP was a response to community critique.
AMP Advice Well Taken? Is Google Starting to Make a Real Perception Change?
It seems that Google and I think alike on this one (or at least we’re starting to). As I mentioned, one of the big PR obstacles that
Just recently, on September 18th, Google announced that AMP was taking a more democratic approach to its execution and operation going forward. Up until this point, one Google engineer had been the final word on all things AMP. No longer. Now AMP will be governed by a committee that is said to have representatives not just from Google, but from the companies/sites that have adopted AMP.
In other words, Google is now seemingly making AMP a truly communal project that is operated with shared governance. I think Google is starting to understand that if it is going to increase AMP’s presence, if this thing is going to work long-term, AMP must actually belong to the community in more than just name.
The question is… is it too late? Has AMP’s perception been cemented in stone? To the best of my knowledge, there still seems to be a tremendous air of doubt around AMP. If you look at a lot of the big names in SEO and their presentations on AMP, the question as to if AMP is a good idea is still a *** topic. Meaning, there is still quite a bit of uncertainty around AMP despite Google’s best efforts, though of course, nothing is really ever set in stone, at least not completely.
But there’s one nagging question left… the August and September AMP upticks… how has AMP hit a wall if it’s spiked to new heights?
AMP’s Data Resurgence – Is AMP a Ranking Factor?
Is AMP a ranking factor? AMP as a ranking signal has been the source of some serious SEO debate for some time and until now I haven’t really had an opinion on the matter. We’ve thrown the idea around here at Rank Ranger more than once with our team being pretty split on the notion. So when I say that I do think AMP is a ranking factor, please know that it’s a personal opinion, and does not reflect our official position, since as I mentioned, there’s a big split among us.
Why then do I think AMP is a ranking factor (which of course is a very controversial idea if taken to its logical conclusion)?
It’s because AMP’s August/September uptick didn’t happen in a vacuum, not even close. Both spikes, that of August 22nd and that of September 6th both coincided with quick, though significant, Google algorithm updates. In other words, the only time in the last two years that we’ve tracked AMP truly significantly increasing its SERP presence is when Google made algorithmic changes.
The Rank Risk Index showing increased rank fluctuations that paralleled AMP’s data spikes
More, if AMP has really caught on like wildfire than shouldn’t we see more sites adopting the AMP format? But we don’t, not really. The average number of AMP sites has not kept up with the percentage of SERPs that contain a minimum of one organic AMP result. With more and more SERPs containing AMP shouldn’t there be more AMP sites on those SERPs? The fact that we don’t indicates that Google does not have more AMP sites from which to choose from.
So how is more AMP making its way to page one? AMP as a ranking factor is an easy explanation. It’s not that there is more AMP adoption, but that AMP adoption is now more heavily rewarded (which in turn could produce what Google wants with more sites choosing to make the move to AMP).
In fact, just recently, we’ve seen organic AMP’s data slide down a bit. Around October 12th, organic AMP, overall, showed to be on a bit of a downtrend. In the US that’s meant that AMP has lost about two percentage points (moving from showing on roughly 36% of page one SERPs to about 34%).
AMP on page one of the SERP has undergone a slight decrease in mid-October
Oddly enough, there was yet another algorithmic event that coincided with the change in AMP’s showing. On October the 13th we recorded a moderate increase in rank fluctuations. For the third time in less than two months, a shift in AMP’s data trajectory coincided with a change in Google’s algorithm.
AMP’s downtrend coincides with yet another Google algorithmic event
So are we really seeing an AMP revival or is the data I presented at the onset of this article more about Google increasing the importance of AMP as it relates to ranking? If anything, it could mean AMP has hit a bigger wall than ever.
For the record, I am not conclusively saying AMP is a ranking factor. All I am saying is that after having seen the pattern I’ve blabbered on about for the last few paragraphs, I have enough to feel comfortable to say that I think (as in, I believe, I speculate, I hypothesize) that AMP is a ranking factor.
AMP Fallen and Can’t Get Up?
I have no way of predicting whether or not AMP will live on or fade into the black. No one does. What I can say is that we are very far away from Google giving up on the project. That does not seem at all like a possibility at this point in time. That alone makes things a bit interesting. If I am correct, and AMP has hit a wall, and if Google is still *** on the format, that should make for some “drama.” How will folk respond to Google’s AMP efforts? What will Google do to ensure AMP’s future? How far will it go to do so? What tactics will it take up? Will Google continue to recognize the issues facing AMP and act accordingly? How will that play itself out? Will Google’s more democratic position affect a real change?
What do you think? Let me know!
About The Author
