How about Discourse Support Google AMP?

I don’t think anyone has properly identified a use case yet.

AMP sounds nice on the surface, but even light research uncovers a litany of issues, particularly if you decide to try and disable AMP down the line.

  • What would you use AMP for in a community setting? The first post in a topic?
  • How would discourse interact with the amp cache?
  • How would readers see subsequent responses?
  • If it’s only for content being replicated from a wordpress site, wouldn’t the native AMP plugin there be easier?

AMP gets placement priority, but it’s not a magic SEO bullet for communities. Discourse already has a snappy mobile experience, AMP can’t beat that. If it doesn’t add something meaningful then why use it?

1 Like

I have done my own testing with AMP on my site, and haven’t really seen much advantage in terms of driving more traffic to AMP pages. After splitting up my pages into AMP and non-AMP I found the sum of traffic to the two versions of the pages roughly equalled what my traffic was before the split.

1 Like

Do you run ads?

If so, was there any impact on revenue?

To be honest, I never analyzed the results there, but I do run adsense on both versions of the pages. This is only one section of a very large site, so I never took the time to parse out the ad revenue just to that area before and after.

we want amp on discourse to launch opur forums

AMP still is a Google-only “standard” and brings very little (if not takes away) to our vision of the open web.

Unless someone from the community works on this or AMP overcome its many flaws and becomes an approved standard, there’s very little chance we will work on it.


Thanks for not supporting AMP. Their Signed HTTP Exchanges and “Portals” just make things worse, imho. If Google succeeds, it might be the end for the open WWW.

Some alternate viewpoints:

I don’t think AMP can be considered “open” technology. It’s a cunning scheme by Google, and other tech companies with resources to host AMP caches, to pillage/appify the WWW and divide the spoils.

It causes web publishers to lose control of their stack, because Google forces them to load a restricted version of the site on Google’s own servers (or the servers of one of their co-conspirators). It’s the end of server-side logs, because visitors never reach your server.

In the long run, the less control you have over your stack, the more at the mercy of those large companies web publishers will be. You end up sending your content onto their platform, and it is no longer really the open, decentralized WWW.

Google puts a back button on your site, so that visitors are more likely to go back to the Google SERPs than to explore deeper into your site (continuing this trend).

Signed HTTP exchanges spoof the URLs so that visitors don’t even know that they aren’t on your real website.

You have to load JavaScript from Google’s servers for it to be valid AMP.

If users don’t want to load Google’s JavaScript, the spec punishes them by forcing an absurd 8-second page loading time (not an exaggeration).

“Portals” are harmful to smaller publishers, because they allow large sites (that typically send traffic) to only show previews of the linked-to sites. It’s the new “can you open all external links in a frame?”

That’s part of the reason why it’s so bad. Google is not playing by the rules any more, and they are treating the Web as if it’s a Google product.

The only way Google can get people to use it is to strong-arm web publishers into adopting their format under threat.

AMP isn’t even faster than a hand-optimized page. Preloading-on-hover techniques (like in Gatsby.js) make loading pages as fast as an SPA, so I don’t think that the speed argument is convincing. It’s mainly about the business interests of some large tech companies that have the resources to host AMP caches.

A little extra traffic in the short term might sound attractive, but I don’t think it’s worth selling out the future of the Web. Also, I’ve read some comments from people who haven’t noticed much increase in traffic after implementing it. See also:

Sorry for the rant. Maybe Google will eventually win and all of our “websites” will become little more than pieces of content hosted on the networks of a few companies’ AMP caches, but I think that there is still a chance to stop it.


I am really glad that Discourse don’t consider to use AMP. With AMP users click a few times more to enter a website.


I have made a paid topic for anyone will to take this and do this as a paid project for me, [PAID] Convert discourse topics(forum) to Google AMP automatically

New information about AMP:

In summary, it claims that Google falsely told publishers that adopting AMP would enhance load times, even though the company’s employees knew that it only improved the “median of performance” and actually loaded slower than some speed optimization techniques publishers had been using. It alleges that AMP pages brought 40% less revenue to publishers. The complaint states that AMP’s speed benefits “were also at least partly a result of Google’s throttling. Google throttles the load time of non-AMP ads by giving them artificial one-second delays in order to give Google AMP a ‘nice comparative boost.‘”


This video offers a great summary about Google AMP and why it really isn’t beneficial after all.


So not only is supporting amp needlessly complicated, it’s actually to your own detriment.

Pretty much everything that I’ve heard from anyone claiming to know about improving SEO beyond the most obvious of practices (with a couple of exceptions) has seemed like total snake oil. The only people who seem to want to do things totally against all good sense (e. g., get an email address and start sending email there immediately, no need to validate that it belongs to them and they know how to type) are those who claim to know about marketing. (Again, with a few exceptions.)


I’ve worked in SEO before I started focusing on my own projects, and there are techniques that work really well, but the industry is full of scams and snake oil. If you hire an SEO company, you might get ripped off, but if you find the right person to do it, or learn enough about it, it can mean the difference between success and failure of a project (especially in industries like ecommerce or travel). It’s possible to stumble on the right formula by accident, but it’s much more effective to improve it systematically.

I’m sure that’s true, and I’ve worked with a few people who knew a lot. One I paid a fair amount of money and wasn’t sorry. (even though it didn’t pay off)


And I’m sure that there are people who actually know something about marketing too. I know that I don’t know about it (I see products in the forum space that are much inferior and much more expensive than cdck hosting; how can I market to people who buy those products?) , but the things I’ve seen client’s marketing people do seem just crazy.

SEO (in-house) tends to be good for that. I’d try to place my site in Google for any query I could find that’s related to starting and managing forums. Most people will do some research before they start. If they don’t search Google themselves, they will ask in a forum and someone else will search Google and then paste a link (further boosting the rankings). Maybe try this technique.

1 Like

The new AMP is “we will start penalizing you in search results if you don’t use it”, aka…

Which is weird, because it seems like a race to the bottom where the only way to win is static HTML files? :thinking:


Will Discourse start using it then (just in case any penalties are faced)? :thinking:

Just to follow up on what I said in November…

TL;DR: Publishers are looking to move away from Google AMP.


There is another nice “new” technologie - also working for non-amp pages:
Signed exchanges (SXG)

1 Like