 Hi, everyone. Thank you for having me. So as the title of my talk spoiled the end of the movie, we wanted to improve our mobile traffic. So what we wanted to do here is that we identified the problem and tried to fix this. So my presentation is about not the end process, because it's already here. We increased our blood traffic. But it's also how, as many developers are in the room, they can relate is how we identify the cause, what kind of solution we try to be able to put in place, and basically how we also measure the results. Little about me. I'm the only Danish speaker, and I'm not even Danish. I live in Denmark for 10 years. I work for a company called Falcon IO. Been working with WordPress for as long as I remember. 10 years, and that's my lifespan memory. There's not going to be any question, but as mentioned, if you have questions after this, you can either reach me on Twitter or talk to me later. So just to provide some context, what does Falcon IO do? It's a SaaS company. We provide platform to handle social media in a single platform, so from publishing, advertising, engaging, et cetera, et cetera. That's context. Context is important here, because part of the problem is that why did we want to increase our mobile traffic in the first place? This is marketing technology, landscape, 5,000 graph. It's something that tells us that this is all our competitors. We're one small dot in the space, and we have, as a brand, to be vocal, to deliver our message fast and better. And working in marketing, one of my job is to be sure that our message can be in front of the line. I also have technology in my tool set. We use innovation as a pillar of growth. So we measure everything, and technology is a way for us to achieve that goal. We identify biggest room for opportunities. There's many projects that we can launch, but we are just so many people who are able to do that. So we use data to measure both impact, to measure what problem, and we test fast, fail fast, and fix fast. At least we try to. One opportunity that we saw was that in 2017, our mobile website, marketing mobile website, was not optimized at all. We had it responsive and all that. It was working, but it was extremely slow. And we're not pushing any paid or social content to it. We're just trying to get some organic from SEO. But it represented only 13% of our overall website sessions. So one of the things that we wanted to work with is AMP, Accelerated Mobile Pages, which is a framework made by Google. And a year ago, it was still very experimental. We implemented it in February 2018. It was not even 1.0. Then it was 0.3 or something, and the framework was evolving super fast. And I'm going to dig into it. First is why AMP? These are numbers that are taken from Alberto Medina's talk from World Camp Europe last year. He works at Google. 53% of mobile site visitors just leave the site if it loads above three seconds. And you lose conversion for every second of delay. And from 5 seconds to 19 seconds load on mobile, you can link it to two times of revenue. I invite you to look at his talk on WordPress TV. So AMP looked like a pretty promising solution for us. So everything that we wanted for it is the framework that built super fast mobile pages. There was an AMP WordPress plugin that was maintained by Google. So that was great. It promised more conversion from mobile, and it promised also more organic traffic from Google, meaning that your content, if it was built on AMP, ranked higher in search in Google. There were also some downsides. AMP was, and probably still is, a fairly rigid framework. I needed to build templates from scratch. Imagine working with a WooCommerce template. I needed to build a template that matched our branding, that matched our setup, and things like that. It also, because of its nature, it's limiting a lot of the assets that you can load within WordPress. You can only have some limited external tracking. We had, for sure, Analytics, Google Tag Manager, but anything else that was part of our tech stack didn't couldn't fire. So we needed to have a lot of work around. And then the last point that I got from my company, Google, is a bad corporate machine. That's for another talk that I'm not going to do. AMP HTML. So the framework works as they have their own markup and their own setup, meaning that some HTML tags are the same. You're still going to have divs, headers, H1, whatever. But some tags are really AMP specific. So you're going to have to learn or relearn some way of doing mobile development. Some tags are prohibited for obvious reasons, like objects and embeds. So you'll need to work with components that I'm going to start talking about. And if you miss a required attribute, you're dead. It's just breaking. So if you're not checking your work and just validate, because the validation happens externally on Google website, so if something breaks, you will not know about it. And it will just fail to be indexed. Components. Components are basically a library that Google is maintaining and validating. It works like Wordpress Plugins Library. It works like a bit like Altitbox, JS plugins. Their open source contribution are encouraged. And you can try to build your own components, if you can. I've tried, and I gave up quite fast, because it requires a lot of ramping up. Result. So this is the layout that we have. Thankfully, our non-AMP normal native responsive layout is fairly easy. So with AMP, I just did minimal adaptation so that it looks fairly the same. The experience is not compromised. And the result at least on speed level was quite great. For us, going from 11 seconds to 5 seconds load average on mobile guaranteed us that we lose less visitors. And for results. So this is a pretty neat graph that I showed for my colleagues to know what we could do with one implementation. So in February 2018, we implemented AMP. And already from June, while we were still fixing, implementing SEO fixes and stuff like that, we started to see a big spike. And the thing with AMP and Google is that the more indexed you get, the more incremental it gets. So by September, October, where I applied for this talk, we were at around 800. And by the end of the year, we're already at plus 2,000%. And for those of you who are maybe skeptical, it's not like one visit that came into 1,000 visits. In January, we had around 1,000 visitors on organic mobile. And now we're above 26,000. So our key learnings. We are one and a half developer in marketing. So this was made by me. And it was made during one sprint. For you who are working in agile and scrum and stuff like that, it represented something like two weeks of having a workable prototype. We implemented it only on our blog because of tracking issues. I made one template that got implemented into 500 posts that got indexed. So that was also the idea for me to get this working as a template that can be reproducible for a lot of content. We saw an uptick in 5% increase in overall mobile traffic, not only organic. But we're still working around the limited tracking that AMP is offering. Final thing that I wanted to point out is that if you want to start working on WordPress and AMP, the new AMP plugin 1.0 has been released at least in December and haven't really gotten with it. But there's a paired mode. There are AMP for Gutenberg. It's quite now on a really stable release. So I'm looking forward to see what this can drive me to. Thank you very much. It felt like I sped up. I still have some time. But if you want to have questions, if you have questions and want to reach me out, please do. Thank you.