Here's the reality. The Taboola news carousals at the bottom of most slow loading actual news sites are filled with total junk. One page per image websites, super slow, adblock goes crazy. I've literally NEVER had a good experience with these "Stories you may like"
Google does something with a news carousel that loads fast and the content isn't crap. I get why publishers running things like taboola think it's terrible, but as a user interested in reading some quick news without going blind, I like it.
Hacker news is the exception in terms of a quick loading / clean website.
I was a google reader user heavily - so I like the more stripped down view of things even if it's "terrible".
One other note about this study. This company's goal is not to educate but to sign up users. At least for myself, Folks who want / like a stripped down experience may just be less interested in being marketed to.
> The Taboola news carousals at the bottom of most slow loading actual news sites are filled with total junk.
This, a thousands times. For both Taboola and Outbrain it's clickbaity trash - usually misleading and low quality content. They're designed to "disrupt" so that people click on them and have in the past featured on blatantly false news stories. I'd honestly be ashamed to work at one of these sort of organisations and it further underlines how awful the web experience can be if you're not using an ad blocker.
It's no surprise to me that AMP is popular among users if news sites are not including such garbage in their AMP renditions.
AMP is a fork of HTML that pressures limited developer resources to maintain a version for Google's visitors while only being fast because of arbitrary framework limitations.
You can get a far better outcome by having search results consider site speed as a factor in rankings, which would make all publishers improve their sites overnight. The reason Google doesn't do this is because they also run the biggest ad network in the world which runs 80% of the slow ads you see on websites.
These projects within Google are common, similar to how it ranks down websites forcing web users to download an app, when their own websites do the same. AMP allows Google to keep people on Google domains, with better data collection and more control over the independent ad networks and analytics that can run on AMP pages.
I don’t like AMP, but I am so happy to see when websites complain that “leads dropped by 59%” because of it.
For “leads” to drop that much it must mean they weren’t legitimate, happy leads in the first place. I bet those “leads” came from a shit letter popup or other similar dark pattern, in which case AMP works as designed and benefits the user by shielding them from such garbage.
AMP is bad for the web in the long term. But it is a decent stop-gap solution given that nobody is willing to make their websites bearable without Google’s pressure.
EDIT: I just tried their website again without an ad blocker and sure enough, I was asked to subscribe to their shitletter.
I'm also an Amp hater and welcomed this news. Then, I went to the site. Because I really hate AMP I dismissed all three of their mobile pop ups; one after another. After reading the article I tried to use the back button to go back to hacker news; took three clicks (maybe one for each popup)? I'm guessing you are correct and they need to take another look for causation.
Agreed. However, I'm more in favour of AMP because if the reasons you mentioned. Essentially I think Google has tricked a lot of publishers into using less dark patterns and avoiding shooting themselves in the foot with heavy ads and tracking. It's not perfect but from an end user perspective I'm grateful for it.
> But it is a decent stop-gap solution given that nobody is willing to make their websites bearable without Google’s pressure.
Somehow I don’t think Google wants them to stop showing ads. Everything else is a distraction that ignores the root cause of the customer experience being shitty.
I think that AMP made Google's search results feel like they were essentially just Google's own pages. Because that's exactly what the user sees in his address bar, when he visits an accelerated mobile page: google.com/amp/<...>. And that encourages him to hit "Back" as soon as he is done with it, instead of checking out the newly discovered website. Because he never felt like he had discovered anything. He never felt that he had left Google.
The fact, that there are no AMP pages in Google's search results on Firefox Mobile, was the final reason for me to make the switch. I feel like I am browsing the world wide web again, not just a Google's snapshot of it.
- A set of enforced good practices that make sites faster and generally more user friendly
- A way for Google to get more control by introducing a semi-proprietary framework
The disturbing thing is that webmasters usually criticize the first aspect. Basically they want to bloat their sites and AMP doesn't let them do it. They don't really care about the second part. They are already using Google analytics, Google ads and optimize for Google search anyways.
It means that the ones who can do something want AMP to die for all the wrong reasons.
For that reason, I think that AMP is good (or less bad). Even better would be to do what AMP does but without the Google framework. That would make it even faster than AMP because there won't be any Google bloat.
The analysis on the google analytics data seems a bit biased against AMP.
- decrease in average positions in SERPs on mobile
- decrease in CTR on mobile
- higher number of impressions
- slight increase in total clicks
These measures are correlated. You can't just add them all together.
More impressions with the same amount of clicks means CTR (click through rate) will drop. Average position should also drop.
In this case we see a slight increase in total clicks and almost double impressions!
What if adding AMP makes Google show your site to a wider range of audience (more variance on target audience), while assigning your pages a lower weight (average position). You would get this behaviour. Lower average position but more clicks. This could lower your account creation and similar metrics because users are less interested.
What I mean is for some websites, getting more clicks would be great. If your total clicks is the goal, you would like these results (ex: page with ads). For search users, this is not good because we are shown pages that are less relevant.
I would like more explanations on mobile leads, account creation, and newsletter email sign-ups dropped. Why not show plots? Are these compared to total from previous months or percentages? We should also need to be able to compare both websites (mobile vs AMP) to see if there are any major differences.
I am not an SEO expert or web designer but I personally use AMP on my modest website. I don't even have a desktop version. Only AMP. It is simpler that way and took only an afternoon (for a css noob) to switch. What made me switch was their amp-img carousel lightbox. They are simple to use and just work. I tried a plethora of css/js carousel or album viewers and chose AMP. Bonus page speed and better mobile search cards.
Very much agree with this comment. It’s quite likely that the AMP pages received incremental impressions in lower positions, which reduced the overall average position but increased clicks.
The screenshot of rankings from another tool (looks like Accuranker) with a few +1 ranking improvements after disabling AMP also seems insignificant. Often this kind of fluctuation is very normal. Without knowing the baseline level of ranking fluctuation, it’s hard to read too much into this.
This is the danger of analysing ‘totals’ without segmentation to better track incrementality.
Any test needs to be properly controlled to form clear conclusions and I don’t see enough rigor here.
I would however commend the article on its advice to avoid simply ‘disabling’ AMP after using it for a period of time. There is cleanup to be done as the article touches on, and I suspect many may not be aware.
Over and over again I'll click a link from Twitter or elsewhere, and in mere moments I'm thinking to myself…waaait a sec, is this another crappy AMP page?
It never looks right. It's always "off" in some weird (and sometimes very obvious) way. Nearly every time I go to the real page on the real website, it looks better and functions better. Since I use an ad blocker, it's never a problem for me that the real site might want to load heavier ads on mobile.
AMP is a plague on the open web. It's offensive, I never request it, and it never solves any problem for me. I'm glad I use DuckDuckGo as my search engine so I never get kicked to AMP pages from search. As a web developer, I've vowed never to implement AMP on any of my client sites and will explain to my clients why if asked (and so far I've never been asked).
So another problem with AMP delivered via Google is it tends to utterly booger your analytics... first pageview (via the Google cache) is viewed as a different domain than your site. This can play havoc with optimization efforts.
In terms of performance... this is a situation where the AVERAGE AMP site is much faster than the AVERAGE site not using AMP... but that's a rigged game given the volume of non-optimized WordPress junk and god-knows-what coming from the typical advertising server (images, video, scripts).
A well-optimized site run by knowledgeable and balanced adults will see a moderate improvement in speed; this is assuming you leaned out the pages already and kept the advertising cruft on a short leash. You can generally beat AMP's load times with some elbow grease.
They don’t actually say why they think leads declined.
They list a few reasons why they think it didn’t increase, e.g. they’re already fast on mobile. But they don’t explain a 59% drop in leads.
Crucially they note that impressions and clicks actually increased with AMP. That points to an unflattering explanation: the AMP format is blocking techniques they use to turn clicks into leads.
So why don’t they discuss the reasons for the drop? Perhaps because they know those techniques are user hostile.
I have experienced AMP only from the user perspective and always hated it since day one. Until now I don't really understand where it came from and why people would willingly enable it on their site. The first thing I do when I see it is consider whether to go somewhere else or open the actual underlying link.
It already sounds toxic. Instead of loading your page suddenly I load a google wrapper that includes some parts of your page. Doesn't that sound like a red flags for anybody? Does it sound less toxic if I call it "free and open google wrapper build on existing technologies"?
Well sure, Google isn't dumb. Adding in enough benefits (speed, carousel placement) to your Trojan horse ensures there are plausible reasons to bring it into the fort.
Publishers certainly wouldn't give up the most important bits of their page (the top), and cede left/right swipe hijacking (on carousel loaded pages) if there weren't some perceived benefit.
The article heavily implies sales leads...collected email addresses to try and convert into sales.
They sell managed wordpress hosting.
However, their site doesn't appear to have much in terms of collecting leads. No newsletters, no trial accounts, etc. They just have paid plans you can sign up for. I guess the "contact us" page could be considered a lead generator.
So, I'm as confused as you are. Lead generation would be pretty bad if you don't collect leads :)
> Our mobile leads dropped by 59.09%. Our newsletter email sign-ups from mobile dropped by 16.67%. Our account creations from mobile devices dropped 10.53%.
This is a totally uncontrolled test. How do they know leads wouldn't have dropped anyways? Without an A/B test you can't draw any firm conclusions from this.
Please don't turn off AMP actually. Do I really want to sign up for your dumb site's newsletter, or get push notifications? Well, so far I've wanted to do that Literally Zero times in the past, so I'm gonna bet on no.
Front end engineers are not selected for their skills in understanding algorithms and data structures. It's quite easy to speed up a web site if you understand how CPU works on a low level and how Javascript is compiled to machine code, but it requires a lot of knowledge.
Haha what? You absolutely do not need to understand how a CPU works to make a fast site. 90% of the job is getting rid of shitty third party ad and tracking code, something most front engineers don't control. Beyond that most is just sensible practices.
In depth knowledge of the DOM would serve you a lot better than knowing about CPUs. What triggers a repaint, how you can avoid it, and so on.
The obsession with data structures and algos is getting out of control, cargo culting this idea around is going to lead to requiring an L5 to change a button color.
Very little knowledge is needed to know how to speed up a website (or keep it from becoming slow). Almost every front-ender I've met, including those who never got beyond inserting jQuery snippets, knows enough.
The reason you're being downvoted is not necessarily because the first part of your broad sweeping generalization is always wrong. At least at the more junior levels, frontend focused engineers don't always have the same level of DS&A fundamentals as backend focused engineers. However, many engineers usually run into their first performance related issue in the first few years of their on the job experience, at which point they learn to speed things up.
It's in the second part of your statement where you're completely and totally wrong. Speeding things up on the whole does not require a lot of knowledge, and it does not require understanding how the cpu works at a low level and how javascript is compiled to machine code. Knowing how that stuff works can help you squeeze the last 1% of optimization out of code, but in practice, you get the first 99% of speed from decisions:
- Which frameworks do you use?
- What "add-ons" that are key to business do you end up embedding in the front-end and how much page size and slow down do they add?
- Where are your hot loops? Are you doing anything expensive inside them?
- Do you have things that are synchronous where they could be in parallel?
- How much eye candy are you adding to the page? Does it all need to be there?
- How much CSS are you using? Are you using it in a manner where CSS optimizations can do heavy lifting?
- What does the frontend and the backend API contract look like? Are there places where excess requests are occurring, and could they be rolled up so that there is less waste?
You may have noticed that many of these decisions boil down to architectural concerns as well as product and business level decisions, which are tangentially related to the labor of front-end engineering. I don't want to lob ad-hominems at people, but I find this kind of attitude one of the most tiresome parts about certain parts of the engineering community. There's this haughty, holier than thou mentality that places data structures and algorithms at the very top of engineering skills. There's a giant world out there where those skills are not at the top of the hierarchy, and actually are least useful because any sufficiently advanced development there ends up being commodified and available as open source software or as a paid SaaS (IE AWS).
Based on this short, flippant comment, it's obvious you actually have no idea how to actually speed up a web site because you have no idea about what the actual top ten things are that you'd do to speed it up in any kind of commercial production usage. What's even worse is that instead of trying to figure out something you don't know, you're instead making up an answer that sounds reasonable but is actually completely wrong and something any experienced frontend or full-stack engineer would understand is poor judgment and an ineffective approach. This is an extremely dangerous attitude to allow into an organization, because you end up with a culture where people are focused on one-upping each other and attempting to look elite as opposed to pragmatically arriving at the right approach for the problem, specific to all of its constraints. I've worked on these kinds of teams before, and it ends up being a miserable waste of time for everyone involved, and I've endeavored to work on teams that don't behave this way and to create teams that don't suffer from this.
Any engineer that gave this kind of a response to an interview question of "We've got a web app that's slow on the frontend and exhibiting XYZ symptoms -- how would you determine the root cause and diagnose it" would be rejected on the spot by me. That's the kind of attitude that can cause engineering organizations millions of dollars a year in engineering resource misallocation.
As a professional community, we need to evolve away from attitudes like yours. They symbolize an idealized, fictional world that is anything but the pragmatic reality of what good software engineering is.
[+] [-] privateSFacct|7 years ago|reply
Google does something with a news carousel that loads fast and the content isn't crap. I get why publishers running things like taboola think it's terrible, but as a user interested in reading some quick news without going blind, I like it.
Hacker news is the exception in terms of a quick loading / clean website.
I was a google reader user heavily - so I like the more stripped down view of things even if it's "terrible".
One other note about this study. This company's goal is not to educate but to sign up users. At least for myself, Folks who want / like a stripped down experience may just be less interested in being marketed to.
[+] [-] lol768|7 years ago|reply
This, a thousands times. For both Taboola and Outbrain it's clickbaity trash - usually misleading and low quality content. They're designed to "disrupt" so that people click on them and have in the past featured on blatantly false news stories. I'd honestly be ashamed to work at one of these sort of organisations and it further underlines how awful the web experience can be if you're not using an ad blocker.
It's no surprise to me that AMP is popular among users if news sites are not including such garbage in their AMP renditions.
[+] [-] manigandham|7 years ago|reply
You can get a far better outcome by having search results consider site speed as a factor in rankings, which would make all publishers improve their sites overnight. The reason Google doesn't do this is because they also run the biggest ad network in the world which runs 80% of the slow ads you see on websites.
These projects within Google are common, similar to how it ranks down websites forcing web users to download an app, when their own websites do the same. AMP allows Google to keep people on Google domains, with better data collection and more control over the independent ad networks and analytics that can run on AMP pages.
[+] [-] tyingq|7 years ago|reply
[+] [-] Rjevski|7 years ago|reply
For “leads” to drop that much it must mean they weren’t legitimate, happy leads in the first place. I bet those “leads” came from a shit letter popup or other similar dark pattern, in which case AMP works as designed and benefits the user by shielding them from such garbage.
AMP is bad for the web in the long term. But it is a decent stop-gap solution given that nobody is willing to make their websites bearable without Google’s pressure.
EDIT: I just tried their website again without an ad blocker and sure enough, I was asked to subscribe to their shitletter.
[+] [-] codazoda|7 years ago|reply
[+] [-] kentt|7 years ago|reply
[+] [-] drb91|7 years ago|reply
Somehow I don’t think Google wants them to stop showing ads. Everything else is a distraction that ignores the root cause of the customer experience being shitty.
[+] [-] krn|7 years ago|reply
The fact, that there are no AMP pages in Google's search results on Firefox Mobile, was the final reason for me to make the switch. I feel like I am browsing the world wide web again, not just a Google's snapshot of it.
[+] [-] zavi|7 years ago|reply
[+] [-] emayljames|7 years ago|reply
[+] [-] GuB-42|7 years ago|reply
- A set of enforced good practices that make sites faster and generally more user friendly
- A way for Google to get more control by introducing a semi-proprietary framework
The disturbing thing is that webmasters usually criticize the first aspect. Basically they want to bloat their sites and AMP doesn't let them do it. They don't really care about the second part. They are already using Google analytics, Google ads and optimize for Google search anyways. It means that the ones who can do something want AMP to die for all the wrong reasons.
For that reason, I think that AMP is good (or less bad). Even better would be to do what AMP does but without the Google framework. That would make it even faster than AMP because there won't be any Google bloat.
[+] [-] setquk|7 years ago|reply
I can’t understand why the hell anyone would want to use it anyway. The experience is terrible.
[+] [-] soared|7 years ago|reply
If I'm reading this correctly they switched from regular pages with CTAs to AMP pages without CTAs. Seemingly this test was designed to fail.
[+] [-] lpasselin|7 years ago|reply
- decrease in average positions in SERPs on mobile - decrease in CTR on mobile - higher number of impressions - slight increase in total clicks
These measures are correlated. You can't just add them all together.
More impressions with the same amount of clicks means CTR (click through rate) will drop. Average position should also drop.
In this case we see a slight increase in total clicks and almost double impressions!
What if adding AMP makes Google show your site to a wider range of audience (more variance on target audience), while assigning your pages a lower weight (average position). You would get this behaviour. Lower average position but more clicks. This could lower your account creation and similar metrics because users are less interested.
What I mean is for some websites, getting more clicks would be great. If your total clicks is the goal, you would like these results (ex: page with ads). For search users, this is not good because we are shown pages that are less relevant.
I would like more explanations on mobile leads, account creation, and newsletter email sign-ups dropped. Why not show plots? Are these compared to total from previous months or percentages? We should also need to be able to compare both websites (mobile vs AMP) to see if there are any major differences.
I am not an SEO expert or web designer but I personally use AMP on my modest website. I don't even have a desktop version. Only AMP. It is simpler that way and took only an afternoon (for a css noob) to switch. What made me switch was their amp-img carousel lightbox. They are simple to use and just work. I tried a plethora of css/js carousel or album viewers and chose AMP. Bonus page speed and better mobile search cards.
[+] [-] le205|7 years ago|reply
The screenshot of rankings from another tool (looks like Accuranker) with a few +1 ranking improvements after disabling AMP also seems insignificant. Often this kind of fluctuation is very normal. Without knowing the baseline level of ranking fluctuation, it’s hard to read too much into this.
This is the danger of analysing ‘totals’ without segmentation to better track incrementality.
Any test needs to be properly controlled to form clear conclusions and I don’t see enough rigor here.
I would however commend the article on its advice to avoid simply ‘disabling’ AMP after using it for a period of time. There is cleanup to be done as the article touches on, and I suspect many may not be aware.
[+] [-] jaredcwhite|7 years ago|reply
It never looks right. It's always "off" in some weird (and sometimes very obvious) way. Nearly every time I go to the real page on the real website, it looks better and functions better. Since I use an ad blocker, it's never a problem for me that the real site might want to load heavier ads on mobile.
AMP is a plague on the open web. It's offensive, I never request it, and it never solves any problem for me. I'm glad I use DuckDuckGo as my search engine so I never get kicked to AMP pages from search. As a web developer, I've vowed never to implement AMP on any of my client sites and will explain to my clients why if asked (and so far I've never been asked).
Just say no to AMP.
[+] [-] emayljames|7 years ago|reply
[+] [-] sixtypoundhound|7 years ago|reply
In terms of performance... this is a situation where the AVERAGE AMP site is much faster than the AVERAGE site not using AMP... but that's a rigged game given the volume of non-optimized WordPress junk and god-knows-what coming from the typical advertising server (images, video, scripts).
A well-optimized site run by knowledgeable and balanced adults will see a moderate improvement in speed; this is assuming you leaned out the pages already and kept the advertising cruft on a short leash. You can generally beat AMP's load times with some elbow grease.
[+] [-] franze|7 years ago|reply
wait until they are kicked out from the google cache (see your trusted google search console, best the useable old one).
then remove the references to the AMP page from the non-AMP pages. then delete the AMPs.
usually takes 2 to 3 days for a few thousand pages.
[+] [-] abalone|7 years ago|reply
They list a few reasons why they think it didn’t increase, e.g. they’re already fast on mobile. But they don’t explain a 59% drop in leads.
Crucially they note that impressions and clicks actually increased with AMP. That points to an unflattering explanation: the AMP format is blocking techniques they use to turn clicks into leads.
So why don’t they discuss the reasons for the drop? Perhaps because they know those techniques are user hostile.
[+] [-] amelius|7 years ago|reply
[+] [-] erikb|7 years ago|reply
It already sounds toxic. Instead of loading your page suddenly I load a google wrapper that includes some parts of your page. Doesn't that sound like a red flags for anybody? Does it sound less toxic if I call it "free and open google wrapper build on existing technologies"?
[+] [-] asdfologist|7 years ago|reply
[+] [-] tyingq|7 years ago|reply
Publishers certainly wouldn't give up the most important bits of their page (the top), and cede left/right swipe hijacking (on carousel loaded pages) if there weren't some perceived benefit.
[+] [-] fatjokes|7 years ago|reply
[+] [-] anothergoogler|7 years ago|reply
[+] [-] izacus|7 years ago|reply
[+] [-] tyingq|7 years ago|reply
They sell managed wordpress hosting.
However, their site doesn't appear to have much in terms of collecting leads. No newsletters, no trial accounts, etc. They just have paid plans you can sign up for. I guess the "contact us" page could be considered a lead generator.
So, I'm as confused as you are. Lead generation would be pretty bad if you don't collect leads :)
[+] [-] drb91|7 years ago|reply
[+] [-] Lizzo|7 years ago|reply
[+] [-] gwern|7 years ago|reply
What happened to total leads/signups/creations?
[+] [-] freedryk|7 years ago|reply
[+] [-] jchw|7 years ago|reply
[+] [-] Markoff|7 years ago|reply
[+] [-] xiphias2|7 years ago|reply
[+] [-] dang|7 years ago|reply
[+] [-] untog|7 years ago|reply
In depth knowledge of the DOM would serve you a lot better than knowing about CPUs. What triggers a repaint, how you can avoid it, and so on.
[+] [-] hellisothers|7 years ago|reply
[+] [-] mercer|7 years ago|reply
[+] [-] yowlingcat|7 years ago|reply
It's in the second part of your statement where you're completely and totally wrong. Speeding things up on the whole does not require a lot of knowledge, and it does not require understanding how the cpu works at a low level and how javascript is compiled to machine code. Knowing how that stuff works can help you squeeze the last 1% of optimization out of code, but in practice, you get the first 99% of speed from decisions:
- Which frameworks do you use?
- What "add-ons" that are key to business do you end up embedding in the front-end and how much page size and slow down do they add?
- Where are your hot loops? Are you doing anything expensive inside them?
- Do you have things that are synchronous where they could be in parallel?
- How much eye candy are you adding to the page? Does it all need to be there?
- How much CSS are you using? Are you using it in a manner where CSS optimizations can do heavy lifting?
- What does the frontend and the backend API contract look like? Are there places where excess requests are occurring, and could they be rolled up so that there is less waste?
You may have noticed that many of these decisions boil down to architectural concerns as well as product and business level decisions, which are tangentially related to the labor of front-end engineering. I don't want to lob ad-hominems at people, but I find this kind of attitude one of the most tiresome parts about certain parts of the engineering community. There's this haughty, holier than thou mentality that places data structures and algorithms at the very top of engineering skills. There's a giant world out there where those skills are not at the top of the hierarchy, and actually are least useful because any sufficiently advanced development there ends up being commodified and available as open source software or as a paid SaaS (IE AWS).
Based on this short, flippant comment, it's obvious you actually have no idea how to actually speed up a web site because you have no idea about what the actual top ten things are that you'd do to speed it up in any kind of commercial production usage. What's even worse is that instead of trying to figure out something you don't know, you're instead making up an answer that sounds reasonable but is actually completely wrong and something any experienced frontend or full-stack engineer would understand is poor judgment and an ineffective approach. This is an extremely dangerous attitude to allow into an organization, because you end up with a culture where people are focused on one-upping each other and attempting to look elite as opposed to pragmatically arriving at the right approach for the problem, specific to all of its constraints. I've worked on these kinds of teams before, and it ends up being a miserable waste of time for everyone involved, and I've endeavored to work on teams that don't behave this way and to create teams that don't suffer from this.
Any engineer that gave this kind of a response to an interview question of "We've got a web app that's slow on the frontend and exhibiting XYZ symptoms -- how would you determine the root cause and diagnose it" would be rejected on the spot by me. That's the kind of attitude that can cause engineering organizations millions of dollars a year in engineering resource misallocation.
As a professional community, we need to evolve away from attitudes like yours. They symbolize an idealized, fictional world that is anything but the pragmatic reality of what good software engineering is.
[+] [-] walshemj|7 years ago|reply
You need to understand how html css and js work and not cut and paste megabytes of cruft.