Can Google AMP make the web faster without taking over the web?September 25, 2018
Google’s Accelerated Mobile Pages project has been controversial since it first rolled out in 2015. AMP is an open-source standard for a specialized type of webpage that offers a huge speed boost to sites that use it, and those pages load instantly from search engines like Google and (now) Bing. But AMP has many critics who have spent years contending that Google is using AMP to take over the web. In particular, Google shows AMP results in a news carousel at the top of its mobile search results, creating a huge incentive for news organizations — including The Verge and Vox Media — to use AMP.
For the past three years, AMP has been led by Google’s Malte Ubl, who was the project’s final decision maker. That’s changing now, though: Google is changing AMP’s governance model, which is how the open-source community likes to describe the creation of committees. Ubl has always said that AMP is part of the open web, and his goal in changing how AMP is run is to make that promise explicit, by involving non-Googlers in making decisions about the future of AMP — and by extension, the future of the web.
I spoke to Ubl on the Vergecast this week to talk about where AMP has been, where it’s going, and how all those changes will affect the web we all use every day. Here’s a small excerpt, where we talk about whether AMP would have been as successful if not for Google’s push towards adoption.
This excerpt has been edited for clarity.
Nilay Patel: Do you think if you had developed AMP in a vacuum — you didn’t have the carousel, and you didn’t have Google’s infrastructure to do instant loading in search — do you think AMP would have won as a web format on the merits? Or do you think it’s growth was accelerated by the things that Google was able to do on the search page?
Malte Ubl: I think that there were two things to this equation: we have some experience at Google with trying to move ecosystems based on search incentives. And they don’t actually have, even though I don’t want to go into details, the greatest track record.
Nilay: Wait, now I want you to go into detail.
Malte: I think you can’t just say that because Google says something, people are going to do it. In the end the ROI has to be right. It actually has to be great, it actually has to be something that people want. There has to be a path toward participating that doesn’t involve throwing everything away.
On the other hand, if I had done this by myself, I would have certainly struggled to get the same kind of publicity. We have done over 25 road shows around the world. We have invested in making our documentation internationalized. If you look at most web frameworks out there, it’s all English. Not everyone around the world speaks English, but everyone around the world should make webpages, right? So, there’s all these reasons why adoption is where it is, and not some incentive system.
Nilay: But you had the incentive system. I just want to be clear, you give the incentive system some credit, and you’re saying Google’s additional resources carried you the rest of the way.
The only reason I ask this, is I’m curious about what evidence you have — independent of adoption — that AMP is a success. Do you have user data that says people are happier and spend more time on AMP? Do you have data that says people are happier with the search product? Is there something outside of adoption, which is driven in part by incentives, and in part by the other work you’re describing, is there some other data that tells you AMP is a success?
Malte: There’s no way this could have launched at Google as a product if it wouldn’t have been for those user benefits. In fact, one of the core original goals of the project was to make trade offs driven by user experience. Unfortunately, that’s not how web development works in general.
There’s two other main things that people think about. One obviously is business success, and one is developer experience. How does it feel like as a web developer to work on this? There’s a trade off between these three. And while we try to make the developer experience and the business outcomes as good as possible, we always prioritize user experience, and it comes to a fundamentally different outcome.
It’s important to understand what AMP came from. We talked to all these publishers, and they came to us, and said it’s difficult out there, and the web isn’t really doing well. You have to remember, this was 2015, people were singing the death song of the web. Things were not in great shape.
Nilay: I wrote an article on 2015 headlined “the mobile web sucks.”
Malte: We probably printed that out, and pinned it on the wall.
Coming from there, what we heard from publishers was like, “Okay, we know we have to do something, but if we’re the only ones stopping pop up ads, then we’ll lose money, everyone else makes more money, and users still hate the web.”
There was agreement that you needed industry-wide effort to say we’re all going to make this jump and prioritize user experience, so that in the end, obviously everyone’s happier and everyone has the right business outcome. Because users are happier, and in the end that always comes to that’s good for business. Right?