developer

Bryce Adams and Metorik

August 7, 2018

Show Notes

And be sure to check out my new show, Creator Toolkit, on Apple Podcasts or wherever you get your podcasts.

Transcript

Intro: Hey everybody and welcome to episode 88 of How I Built It. Today we’re talking to Bryce Adams, creator of one of my favorite tools, Metorik. Bryce tells us about the path the led him to making Metorik, his view on ecommerce, GDPR, and how he’s using some cutting edge tools. We’ll get to all of that and more, but first…

I have a new podcast coming out this week called Creator Toolkit. If this is the show where we talk to the carpenters of the world, Creator Toolkit is the show that tells you what hammer you should us. We’ll talk all about how to build specific types of projects and how to make certain decisions when building on the web. The first episode drops Thursday, August 9^th and covers hosted vs. self-hosted. If you stick around until the end of the show, I’ve included a short preview.

Sponsors: This week’s episode is brought to you by Creator Courses and Pantheon. We’ll hear about Pantheon a little bit later.

Creator Courses is a website dedicated to teaching you how to build on the web. Their catalog of courses is continually growing and it’s becoming the best place to learn how to build specific projects with task-based objectives. You will always learn by doing. Currently, you can learn how the new WordPress editor with their Introduction to Gutenberg course. You may have seen the notification to try it in the latest WordPress update. This course will teach you everything you need to take full advantage. Head over to buildpodcast.net/gutenberg, and use BUILDIT at checkout for 40%.

Joe Casabona: Hey, everybody. Welcome to another episode of How I Built It, the podcast that asks, “How did you build that?” Today, coming all the way from Melbourne, Australia– am I saying that right? It’s Bryce Adams of Metorik.

Bryce Adams: Perfect.

Joe: Bryce, how are you today?

Bryce: Hello, hello. I am really good. I feel very spoiled to get to do a podcasting to you at 11:00AM my time. Thank you for meeting me late on your end.

Joe: Oh, no problem. My wife is a night shift nurse, so once the baby goes to bed I’ll get an hour or so of extra work in around that time. But it’s my pleasure, I’m glad we were able to sync up.

Bryce: Yeah, me too.

Joe: I’m always careful to say “Melbourne” the right way, because my brother lived in Melbourne, Florida.

Bryce: I think everyone in Melbourne, Australia knows Melbourne, Florida. Especially when we talk to anyone from America. You just say, “I’m from Melbourne,” if they don’t hear your accent, maybe you’re typing or something like that. They’re like, “Melbourne, Florida? I’m close!” Or, “I went there once.” It’s so strange to me because I’ve never been. I don’t even know what’s going on there. “I know it so well.”

Joe: There’s a college that’s younger than my dad and that’s about it.

Bryce: That sounds like a great place.

Joe: It’s nice. It’s about an hour from Disneyworld. That’s interesting, because when he said, “I’m going to Melbourne,” I’m like “You’re going to Australia for college?”

Bryce: Not quite.

Joe: He became friends with somebody from Australia and they corrected our pronunciation. So I am always mindful to say it the right away.

Bryce: You did it perfectly. It would be quite challenging, as someone from Melbourne, to go and study at that college in Melbourne, Florida. Just thinking of the conversations I’d have. It would be like, “It’s nice to meet you, where are you from?” “I’m from Melbourne!” “You don’t sound like it.”

Joe: The closest thing we have near me is there’s a Moscow, Pennsylvania.

Bryce: Oh, wow.

Joe: Not nearly as cool as Moscow, Russia. That’s everybody’s geography lesson for the day. Today we’re going to be talking about a tool that I’ve been using for a while that I’m a huge fan of, called Metorik. But Bryce, why don’t you tell us a little bit about who you are and what you do?

Bryce: Sure. I’m Bryce. My background is that I was traveling for a while, I was nomadic. Then I got into the WordPress world and made some plugins. A lot of them were free. Eventually I made a couple that were paid ones on CodeCanyon. That was my first foray into making money from coding. That was really cool and challenging.

And then I ended up getting a job over at Woo. This is back in probably 2014, and I was doing support for WooCommerce. Really challenging but valuable experience for me. It was my first time doing WordPress and working with WordPress and coding as a career. I really enjoyed that time. Eventually we joined Automattic, did the same thing there.

But I ended up leaving back in 2016 and started to build this product, Metorik. I always struggle to describe it, which is awful. Because it’s changed so much from what in my head I would describe it as before. But these days I describe it as a co-pilot for WooCommerce stores and anyone running WooCommerce e-commerce stores.

It’s there to help them with all the things that they can’t do through WooCommerce itself, and also, a lot of things that maybe they never thought to do with WooCommerce. It starts with things like reports, but then goes further and lets you segment your data. Finding customers that haven’t ordered in a while, or trying to figure out what your highest selling products are.

And then I kept taking it further than I originally planned, and started doing things like integrations. I had an integration with Zendesk and then HelpScout, and a few other support platforms. And then Google Analytics. More recently a big focus of mine has been on something I’m calling Metorik Engage, which you might have played with yourself, Joe. It’s like e-mail automation for WooCommerce stores. Made really simple, but still really, really powerful.

My advantage there is that I’ve leveraged this segmenting system I built for Metorik, that I described as an infinite segmenting system. The idea is you have all your data, your orders, your customers. And then in Metorik you can say, “Show me all the orders from one of these cities, and was for this amount, and had these products,” and you can stack as many rules as you want.

With Engage the idea is, “Now that you’ve segmented those orders, or those customers, or even those WooCommerce subscriptions. Let’s e-mail those customers automatically.” For me that’s a pretty exciting direction I’m heading in now.

Joe: That sounds great. We’ll totally dive into that. It sounds super interesting, I think the way you described it is naturally really good. You’re the founder, so–

Bryce: Thanks. You’d hope so.

Joe: What drew me to it was WooCommerce is free, it’s a free e-commerce platform.

Bryce: Yeah.

Joe: But I think one of the big things it’s lacking as a big boy e-commerce platform is reporting. There’s not a whole lot built into it. I heard of this from Brian Krogsgard, who says that he raves about it. I signed up a little while ago and I wasn’t doing that well with sales, so I’m like, “I can’t really justify reports for the 5 customers I have.” But this year it’s a lot better for me.

Bryce: Awesome! Congrats to you for that.

Joe: Thank you.

Bryce: Yeah.

Joe: One day I was just bouncing around WooCommerce, and I was like, “I need to figure out how many people from Pennsylvania have bought my product.” Because in Pennsylvania you still need to pay the sales tax on digital products.

Bryce: Exactly.

Joe: And I couldn’t do it. I’m like, “All right. I’m just going to go ahead and use Metorik. I’m just going to do that now. Pay for the year.” Totally worth it.

Bryce: I appreciate it. I think it’s interesting you mention that because that’s the idea. I love when people say to me, “What do I do with it?” And I’m like, “Either that means you’re just getting started and your store doesn’t have that much data, and that’s totally okay.” I don’t feel like Metorik’s a product every single WooCommerce store should have.

A lot of my customers feel like that, and I appreciate that. But then of course when you’re doing a few orders a month there just isn’t as much value you can get out of it. Perhaps you can get a little bit, but I can’t guarantee to them they’re going to get the price of the subscription out of it.

And I don’t do anything for free. Just because– we can talk about that later. But it’s something that I’ve struggled to accept since it compromises the sustainability of the product. I always say to people, “I don’t want you to pay for it if it’s not worth it for you.” I’m not trying to be the winner here and make more money to get the better end of the deal.

I want our customer to feel like it’s a win-win situation. Of course I’m winning because I’m getting a customer and revenue, but they’re winning because they’re saving so much of their time. Like, you ran into that issue. If it saves you even an hour of time that’s probably worth it, at least just for the month. So that’s my approach to it.

Joe: Without a doubt. That’s exactly how I justified it. I’m like, “I could either dig through or find some less than reputable plugin to do this one thing for me this one time, or I can pay for the year and save myself time now and around tax time.” And just saving my time around tax time is going to be clutch for me. I can send this to my accountant and be like, “Run whatever reports you want.”

Bryce: Yeah. And that’s it. Everyone wants to do things differently. Someone might struggle at first using Metorik because it might be a little overwhelming for them, and I say, “I could just make all those decisions for you.” I’ve even done talks at WordCamps about making decisions, not options for customers. I’m a big believer in that.

But I think running an e-commerce store, an actual business online, is so much different to a simple plugin that I install on someone’s site for adding a contact form or something like that. For me an e-commerce store, your store is the heart of your business. So I don’t want to make all these decisions for you and pretend I know exactly what you want, and how you want to get it.

I want to give you that flexibility while still guiding you there and saying, “These are numbers that I think are important, but if you feel otherwise, that’s okay. You can figure out whatever numbers you want. You can segment your data however you want. It’s your data. That’s totally okay.” That’s my approach there.

Joe: That’s fantastic. You totally mentioned decisions and options, that’s part of the WordPress development mantra. Their development principles. But you’re right, you’re running somebody like me that’s only digital products. Or there’s people who are selling physical stuff that might need more information.

Bryce: They need different data, and it’s not like I can just make decisions for you guys separately. It’s not just the digital products and the physical ones, but then the physical ones might have subscriptions. And then some of those physical ones that have subscriptions also have a digital element to it.

But that’s the beauty of WooCommerce and WordPress and why we all use it. It’s not because WordPress makes these decisions for us. It’s the complete opposite. It gives us that freedom. So I wanted to take a leaf out of WordPress’ book and WooCommerce’s book there, and give people the freedom to get the numbers that matter to them.

Joe: I think that’s incredible. So, you left Automattic in 2016. What gave you the idea for this product, or this service?

Bryce: Yeah well my advantage was that I was talking to not even hundreds, but thousands of WooCommerce stores every month. Especially because my job was WooCommerce support. I didn’t start in development, I grew into that. But I started by talking to customers and them telling me the problems they were having.

Back in end of 2014-2015, this was way before Automattic. I had that idea and I was still in support but I’d been learning to program on the side, and been making a few plugins here and there. I felt confident enough to take a crack at it and make something really simple but that solves the problem I was trying to solve.

Which was a lack of reports and KPIs, just knowing “What’s your average order value?” That wasn’t something that you could easily find at the time. And so I built something, a really rough beta back then. I pitched it to the Woo team but it wasn’t a really a good fit at the time, and it kept getting delayed, and then I started doing development there as my job. My spare time I didn’t really want to spend doing more development.

So I just forgot about it. But slowly over time as we joined Automattic and I realized no one, at least from outside, was interested in doing it. I just thought, “OK. This is something I really want to do. I’ve got the idea, that’s pretty much it. But I’m confident that I can build something that could create value for me and for customers.” Definitely really, really tough decision.

A lot easier to look at in hindsight and say, “It was the right decision.” At the time I really wasn’t sure. Especially in the WordPress world, working at Automattic is considered the best job you can have by a lot of people.

Joe: Right, yeah. “You’ve made it.”

Bryce: Yeah. Of course there are so many other amazing companies that I would in a second be happy to work for in the WordPress world. Like Human Made, and even what Pippin’s doing, Easy Digital Downloads, Sandhills. So there are so many of these great companies to work for. But at least for me, working at Automattic, I wasn’t wishing for another company to work for. I wasn’t thinking about that.

It was really tough to do. But I just said, “I want to do this. I’m fortunate that I’m young, I don’t have kids.” I have a dog, it’s a little responsibility there. But I did feel like I was in a position where it was possible. So I said, “How long will it take me to build it? How long will it take me to get that first customer? How long will it take me–?” Importantly, to not replace my salary at Automattic. That was not the goal at the start.

But, “How long will it take me to break even, to live the life I’m living right now without dipping into my savings?” I set a budget and timeline and said, “If it doesn’t work out,” I left around August-September 2016. I said, “I’ve got until March or April 2017 to get to the point of just breaking even.”

I didn’t mind if I was breaking even for another year after that. But at least get to that point where I can breathe. I think anyone in that situation is thinking of the same thing. Like, “How can I get to the point where this is a sustainable activity?” And then I can worry about making a business and making money and benefiting from it.

Joe: Absolutely. I maybe picked self employment at the exact wrong time from that perspective. I had a 3 month old at home.

Bryce: That’s hard. But it’s also like, “When is the right time? When they’re six months, when they’re a year?” It’s just going to keep getting harder. So you take that risk. And I think, at least, the sooner the better. At least for me, I was just thinking, “I’ve got the money there just to support me building this for a while. I’ll just do it. And if it doesn’t work out I’ll get a job.” I’m not giving up that much. And it’s not that I’m really struggling in that situation.

Joe: Yeah, absolutely. And my wife was incredible she was very supportive. I said, “I’ve got basically six months to make it so that we’re not dipping into our savings. Otherwise I’ll find a job.”

Bryce: Yeah. It’s not the end of the world. It doesn’t have to be. It doesn’t have to be like that “Make it or break it,” thing. It can be just like, “OK, I tried this. It doesn’t work. I’ll try a different job. And then I can try this again in a few years if I still want to do it.”

Joe: I always like to ask about the research, but it sounds like you had a pretty nice gig built into your full-time job doing research, talking to customers about their frustrations.

Bryce: Yeah, definitely. Probably the biggest disadvantage I had was that at Automattic it wasn’t really possible to do side projects like that. So it’s not like I could work on it in that time when I was there. But ideas are something you can have, and understanding the problems that people are having and thinking of ways to solve it definitely helped.

As I was touching on before, it’s funny how it worked out. Metorik is so far from that beta I built back in 2014-15. It’s so far from that now, just because, as you start it turns into something else based on customer feedback and everything like that. It’s funny because I don’t think I’ve ever researched what it is now, but I definitely felt like I had some understanding going in of what I was starting to build.

Joe: That’s great. I mean it’s cool that you’re flexible enough to change it in ways that add a lot of value for your customers, too. So let’s get into the title question, “How did you build it?” I’m very keen on this question because it sounds like– I made my first web site 15 years ago, I was in high school, I learned how to program. I’ve been programming for about that time too. And it sounds like you, within the last five years learned how to write code. Is that accurate?

Bryce: To some extent, yeah. I had my first WordPress site a long time ago. I don’t know, 10 years. It was a Mac news website. It was called It’s All Mac. That was pretty fun. I don’t know even what version of WordPress it was. But back then I had to dabble with changing the color of something, basic CSS. Never really got further.

I remember 2011, or 2012 I was starting University here and I really, really hated it. It was my first week and I had this idea, “How great would it be for a stack overflow app, but for Universities. Where you could join the community of your University and ask questions, and answers.” So I was like, “I’ll build that since I’m already here at Uni. I’m in the right environment.”

I remember even back then it was PHP, I was using some framework. Can’t remember what it was called. Definitely not WordPress. Something else. But I had no idea what I was doing, I remember being so overwhelmed and so scared of the concept of how to change things and writing PHP and just how it all worked.

I think I had a little bit of background definitely with CSS and basic stuff and understanding how the web works. But I definitely wasn’t a very good programmer back then. I ended up leaving University after that week. It didn’t go very well. I just couldn’t do it.

I tried and I was like, “No, I’ll go travel for a bit.” Came back, tried again. I think I lasted another week that time and then I was just like, “This is not working.” That was 2012 and I ended up trying to learn more about WordPress and coding and that’s where I started that journey.

Joe: Awesome. That’s a very cool story. I totally did the traditional four years of University and then I got my Masters degree in software engineering. I was very, “Academics, academics.” And that was the right way for me.

Bryce: Yeah. I don’t think there’s any right or wrong way to do it. That worked for you. Maybe I’d be a much better programmer if I’d also done it that way. I think it’s just the person who’s doing it and like, “What works for them? What’s the best way for them to learn and for them to grow in other ways?” Not even just with programming but as a person.

Joe: Absolutely. I mean I think if I maintain the same mindset that I have now in 17 years I’ll tell my daughter, “Go to college or don’t. As long as you can do what you love, or do what you want to do and support yourself. That’s fine.”

Bryce: Exactly. It’s also the world we’re in. Especially 10 years ago, I’m thinking back to when I was going to Uni, I definitely felt a lot of pressure to go. It wasn’t optional where people were like, “You can go if you want.” I remember back in school, meeting with the career counselors and they’re trying to help you figure that out.

It was never an option not to go to University. No one ever said, “You don’t have to go if you don’t want to. You can just figure things out, or travel.” Of course they can’t recommend you that in their position, but maybe in a few years, or maybe already now that’s starting to change. I hope so.

Joe: Here in the States I think that the next big bubble to burst is going to be higher education because it’s so expensive. Now people are leading the charge for trade schools.

Bryce: The student loans, and stuff.

Joe: Right. I’ll say that, “Yes. Getting a formal education in software engineering has made me maybe a better software architect than a lot of my contemporaries starting out.” But lots of people are self-taught, and if you find the right resources to self-teach, then that’s fine too.

We’re covering a lot of different life topics here. So how did you build Metorik? It’s very nice looking, I mean it looks really nice. It seems very powerful. I’m very curious about your development stack.

Bryce: That’s a great question. Well it’s Laravel behind the scenes, which is pretty familiar now to a lot of WordPress people because it’s the biggest PHP framework in the world right now. It’s amazing, the community is amazing. It’s all open source so it’s a very similar feeling to what you get working in the WordPress world. I do feel like one thing I am missing out is that I haven’t really gone to many Laravel conferences or communicated a lot with the community there.

Definitely my fault, because I’m building this product and I really don’t have as much time as I used to when I was an employee at a company where I was involved in the community. But I definitely see a lot of it just from building it and being on the outside. It definitely has a lot of similarities to WordPress. So Laravel is behind the scenes, it’s the PHP framework.

For the actual front end I don’t really do anything with Laravel there. I’ve got an API that Metorik will talk to. So it’ll say to the WordPress API, “Give me all the orders. Give me all the customers.” On the front end it’s all Vue.js, which may be familiar. But it’s very, very similar to React in that sense where it’s a complete UI framework that gives you complete control of the reactivity of your app, and building these really amazing experiences.

And the stuff I’m building now I never even thought would be possible. It wasn’t even a matter of, “Is it possible for me to build this?” But I didn’t even think technically it was possible. And yet I’m doing it now, and it’s honestly easy with Vue.js. I’m not trying to be humble with that, I’m just being honest. It is really, really simple. I’ve taught it even to a few people around me, where they’ll ask about that. And I’ll try to explain the concept to them, and in doing that I even see again how simple yet powerful it really is.

Joe: It’s funny, I outline the things I want to learn this year, and Laravel and Vue are both on my list of things to learn.

Bryce: Perfect. I’m very happy to hear that.

Joe: Another question I’d like to ask the developers is what’s your environment? Are you a Sublime guy? Or VSCode?

Bryce: VSCode, for sure.

Joe: That seems to be the popular answer.

Bryce: I think I’ve gone through a few different ones while I built Metorik. I’m thinking about it now and it feels like there’s nothing before VSCode, but I definitely only started with VSCode 6-9 months ago, something like that. Before that I used to use PhpStorm, but then everyone wants to go simple and use Sublime sometimes.

I definitely switched between them. But these days I can’t even think of using anything but VSCode. Just an amazing product. So many great plugins for it. It’s such a pleasure to work with.

Joe: It’s so pretty and works really well. I was using Atom before that and Atom was good, but it got bogged down a little bit too easily for me. So, cool. VSCode.

Bryce: Yeah, love it.

Joe: What about your local development environment? I think the most common answer probably is “Local by Flywheel,” among WordPress people. But this is not a WordPress powered app. Are you doing local development?

Bryce: Technically I have used Flywheel. It really came in handy for testing. There is a component of Metorik which is a WordPress plugin, the Metorik Helper plugin. Really lightweight, it’s really there to improve the syncing of data, and also to do some things like track where customers come from, how long they’re spending on the site before checking out. Those kinds of things.

So I did use Flywheel Local recently, just to test that plugin with different PHP environments. With 5.2, 5.4. Amazing for that, to be able to spin up those environments so quickly and run it all at the same time. I’m running 5.2 and I’m running 7 and everything, so that was really cool.

Joe: And you can change it on the fly, too. If you set up a custom environment you can just change the version, right there, on the fly.

Bryce: It’s amazing. I wish I had that when I was doing WordPress plugins every day back then. But that’s all right. Before it used to be HomeSit which was a VirtualBox favoring thing. But these days I’m using Laravel Valet. It’s really, really great because what it does is you don’t actually create a virtual environment or VirtualBox on your computer.

But rather it’s running the stuff needed to power the app, just on your computer. So you have MariaDB or whatever using my sequel, you’re just running that on the Mac. Normally using Homebrew. It’s so easy to set up and the best thing is I don’t have to tweak it at all. I never even think about it.

When you asked me I had to stop and think for a second, just because it’s honestly not something that’s part of my day-to-day anymore. While with Vagrant stuff I felt like every week I was searching how to fix some problem and waiting an hour for something to happen. Nightmare.

It’s a really, really nice experience. Don’t get me wrong, I do run into issues occasionally. It’s normally when there’s a big OS X update, there’ll be something that messes up the configuration for it or how nginx handle things. Normally you can solve it by reinstalling it. And because it’s so quick, there’s no setting up a virtual environment, you can try a whole bunch of solutions really, really quickly.

Joe: Awesome. So, I’m going to ask this because it’s timely as we record this. People will probably be well sick of it by the time this episode comes out but you mentioned you have the Helper plugin and it’s sending data to your app. How have you been affected by GDPR–?

Bryce: GDPR. Sorry, I can’t help it.

Joe: You totally telegraphed that question.

Bryce: Well, it’s a great question. Definitely something really relevant to Metorik, and something that I did have to invest quite a lot of time into like everyone else. But also something that I didn’t have to invest as much time into because I’ve always built Metorik from the start as being privacy-centric, privacy focused and really trying to respect not just my users’ privacy but their customers’.

Because an interesting thing is that I’m not just what typically every store is, which is a data controller. I’m also a data processor in that I’m processing your data and the data of anyone using Metorik. So there are different responsibilities there, and different ways to handle things. Fortunately it’s not as much work as you think because– I’m trying to think of a good example.

It’s not like I’m a quoting app or an invoicing app where I’ve got all your data that’s sensitive to your store, what’s your business, and everything like that. Metorik doesn’t have too much data about individuals. Like, about you. It’s just your name and your e-mail and things like that. The general stuff.

Most of the data is your actual store’s data which I process, don’t control. A lot of things are taken care of through how WooCommerce handles the data. Especially with the WooCommerce 3.4 latest release that added that GDPR compliance. Little things like removing customers details, no problem. It’ll remove them in Metorik. Delete a customer, they get deleted in Metorik.

It hasn’t been that difficult to comply. But it’s been really insightful to see the impact it’s having on the industry. For me, I really got a lot out of it, because I’ve improved my privacy policy. That’s been completely revamped. Which everyone’s doing. Sending the e-mail to everyone, “I got a new privacy policy update.” It is something that I think everyone knows that feeling.

When you’re starting a product, especially on your own and you’re bootstrapping it, you’re not going to go spend $10 thousand dollars on lawyers. Or invest all this time in worrying about your terms of service, how the words are formed and if you’re using the right grammar and things like that. That’s not how you build a sustainable bootstrapped company at the start. It was really nice to have a chance to revisit that now, where I had the resources, and try to improve it.

Joe: That’s a really great point. It did make me revisit a few things. Luckily I use mostly WordPress tools and they’ve made it easy to be at least as GDPR-compliant as I’m willing to be as a stubborn American.

Bryce: Fair enough.

Joe: At least, outside of the EU.

Bryce: Nice save.

Joe: You know, you made it into my privacy policy. I said, “I send data to Metorik.”

Bryce: Exactly. And that’s the extent of it. One of the things that I found really helpful, and I saw a lot of similar products do the same, is to make a GDPR help doc or page. Where it says, “These are how we’re tackling all the things.” I’ve answered a lot of frequently asked questions. Of course, I’m not really getting asked by American stores. But I’ve got a lot of European customers, so I was getting asked everyday.

I just said, “Yeah, no worries. Don’t worry. I’m compliant, I’m on board. Here are some of the answers that you’re looking for.” Basically everyone wants to know the sub-processor who I’m sharing data with, and it’s really easy for me because I don’t share with anyone. There’s no reason. I don’t need to, and I won’t.

For example, I’ve got something you’ve probably seen with Metorik where it’s a global search and then there’s just general searching where you can just type in a name and find all the orders or costumers or products that match it. It could be a lot faster if I use something like Aloglia. I’d love to. It would be really, really fast.

Besides the cost reasons, I don’t feel comfortable with the idea of sending not just my customers data, but my customer’s customers’ data to these third parties. So I’ve tried to build everything in-house. I don’t even back-up or store actual customer data to external services. Of course I’ve got a bunch of backups and stuff in place, but my focus is more on backing up the Metorik data. I can always get the data from your store again.

I can always sync with your store and get everything, if you’re a small store, in a couple minutes. If you’re a big store maybe a few hours. I’m willing to save you, “I’m sorry, you have to wait a few hours for that data to come in.” If it was a catastrophic failure and I have to now go and get it from your store again. I’m okay doing that if it means that your data doesn’t touch a third party. For me that’s a priority. It’s made GDPR easy.

Joe: That’s fantastic. As a store owner it gives me peace of mind, because you’re not taking my data and then selling it off to somebody else or giving it to somebody else.

Bryce: No. That’s my nightmare. Of course, if Metorik was free, I don’t know how else you’d make it sustainable. “If you’re not free, you’re the product.”

Joe: Exactly, right.

Bryce: “If you’re not paying, you’re the product,” Right?

Joe: “If you’re not paying, you’re the product.” Right. That’s for sure. When all that Cambridge Analytica stuff came out about Facebook, I was shocked more at how shocked people were.

Bryce: Yeah.

Joe: I’m like, “What did you think? Facebook is a billion dollar company. How do you think they’re making money? Well, we’re at the end of the conversation. I try to keep these episodes about a half hour long and this one has been a lot of fun talking about all sorts of different stuff.

Bryce: Yeah, I’ve had fun.

Joe: We’ve talked about where you’ve been, so I’ll end with the two questions I always end with. The first one is, what are your plans for the future?

Kind of trying to figure that out now. A big part of the last few months has been worrying about getting Engage, which is that e-mail automation product, live. Especially because it’s an add-on product. Before Metorik was just one price depending on how many orders you had, but now there’s two. It’s a little bit extra if you want to send unlimited emails through Engage.

And a lot people are using it now, and that’s been an amazing feeling just to ship that and get people paying for it. Now I’m starting to think OK, what do I do next? I don’t have an amazing answer for you. But in general I just have the longest to-do list in the world. Of ideas and feedback from customers, and things I need to work on.

Some of the things I’m going to be adding soon, specifically to Engage, like cart recovery e-mails. Everyone’s asking for. Some of those things. I want to keep leveraging the platform I’m building with Metorik. Well, I don’t know if you’d call it platform, but the product. And take it further and keep making it easier for my customers to run their stores and make money, and to grow their businesses.

Joe: That’s awesome. Actually, talking about Engage real quick. Let’s make that distinction. Because you said you have the Engage e-mails and then you’re working on the cart abandonment e-mails. Would Engage be more akin to the WooCommerce follow up e-mails plugin that’s there? Someone places an order, and they get the four e-mails that WooCommerce generates, but you can customize it and say, “Hey thanks for buying this product, this and that,” and stuff like that.

Bryce: Very similar to that one for sure. But more so in that it’s also just saying, “I want to contact customers when they haven’t ordered in a while,” or with subscriptions I want to e-mail them a week before their next payment. I’ve got a lot of customers that use WooCommerce subscriptions and charge their customers a large amount, a few hundred dollars, every year.

They don’t want to just charge them, they want to e-mail a week before. But also if the customer cancels they want to follow up and say, “Can we get you back? Here’s a coupon.” It’s hard to describe it as doing one thing because the approach I’ve taken with leveraging this infinite segmenting system I have is, people are coming up with these crazy approaches to using it.

And it’s making them money, and I didn’t even think of it. Some people will use it for getting reviews, but then some people are just using it for telling people about relevant products. I just had never really thought of all the ways that they would use it. And so for me, it’s really amazing to see what people come up with.

I’m trying to do more on my end to write about the different ways to use it, because I only know what I can think of, but customers are thinking of all these different ways. I want to make it easy for other customers to use those similar methods.

The cart stuff and anything else I do is all going to fall under this umbrella of Engage, where it’s about engaging with your customers or potential customers and trying to provide them with your product and your service and grow your store with them. I don’t think it’s just going to be that. There’s other things I want to do under that umbrella, for sure.

Joe: For those of you who are listening to this episode, but have not listened to the most recent Chris Lemmer episode on managed WooCommerce hosting. He talks about the importance of segmentation and connecting with your customers. The example he gives, which sounds like a perfect use case for Metorik’s Engage, is you have customers and you have repeat customers.

Maybe they buy a new product every 127 days. That’s the number that he gave. You don’t want to hit that customer with, “You bought this, why don’t you also buy this?” A week later, you’re using a dumb segmentation system there. But if you know that every 127 days they buy something, on day 127 or day 126 you could say, “Here’s a little coupon. Go buy something else in the store.”

Bryce: For sure. Or you can even take that and try to think, “If I know it’s 120 days, how can I lower that to 100?” Maybe I e-mail a month before, but give them a coupon. So now I’m giving up a little value with the small 10 percent or whatever discount, but they’re ordering more frequently. That’s how it started with Metorik because I had that data. It’s so easy to see, “What’s your average time between orders?”

And you can segment and say, “What’s my average time between orders just for customers from Pennsylvania?” Or, “For customers that bought this product as their first product?” So you can see all that, but then people kept saying, “I’ve got that data, but now I need to use it. I want to contact them.” That’s how Engage happened.

Joe: Very cool. Very cool and awesome. I took a cursory look at it, but now I actually understand. I have a lot more context of why it’s important for me. I’m definitely going to take a look at it now. And I want to ask you my favorite question which is, do you have any secrets for us?

Bryce: I’m figuring it out. I’ve got one bit of advice, I guess. And it’s just something I say having gone from starting this and being in that early stage, to now, I feel like I’ve built something I’m quite proud of and is successful in that sense. I’m always hesitant to give advice because I don’t like the idea of telling people what they should do, like I know better. But it’s just one thing I do see a lot, especially in the software as a service world.

Don’t take things too seriously at the start. Don’t worry so much about perfecting things that don’t matter. When you’re starting, at least the software as a service, or any kind of software product. Just ship it and get revenue, and then iterate and talk to customers. I see so many people that are working on products for a year before they’re selling it, and I’m sorry, that’s just way too long.

When I launched Metorik, of course it was far worse than it is now. It didn’t do like a quarter of the things it does now. It had bugs, it had all these problems. But I still got customers because they noticed and recognized the value in it, and a lot of them were just supporting me because it was of value to them but also they wanted to see what it would become.

So they were happy to support me in those early days. If I hadn’t launched maybe it would have all failed, so I can’t stress that enough. Just put something live and get feedback. You can always improve something but you can’t recover that time you’ve lost if you wait too long.

Joe: Absolutely. If you’re starting something out I think the thing that I always got stuck on was like, “Is it scalable? Can it scale to a million people?” I don’t have a million people yet.

Bryce: Doesn’t matter, exactly.

Joe: I mean, “Let’s scale it to 5 people and then 10 people.”

Bryce: For sure. And also you’ve got to think, “What’s relevant?” For your business you’re trying to build. And with Metorik I knew my average customer would be paying $50 to $100 US a month. There’s some a little bit lower and there’s some a lot higher. Some people pay me $600. So I knew, “If I’m making between $50 and $100, how many customers do I need to be at the point where I’m breaking even? How many customers do I need to have my old salary? How many customers do I need to be making more money than I know what to do with?”.

You can have those numbers and I promise you those numbers are going to be lower than you planned to build the app for. For me that number is anywhere from 50 to a couple hundred customers. Anywhere in that area is amazing. If I get to 1,000 or 2,000, that’s great.

But again it’s not millions of customers, so there’s no point of planning to scale to millions. Because if I have a million customers, that means I’m doing a billion dollars in revenue a year. I don’t think so. I’m not planning to get to that point. I don’t expect I will ever–

Joe: And if you do get to that point, you can afford to scale.

Bryce: Exactly, yeah. When you’ve got a billion dollars in recurring revenue I think you can figure something out.

Joe: Well, Bryce. Thanks so much for your time today. Where can people find you?

Bryce: My pleasure. Best is probably Twitter. Twitter.com/BryceAdams. And then check out Metorik. If you want to run a WooCommerce store and you haven’t heard of Metorik, please come and try it out. Because I want to know what you think, and I want to know how you survived this long without something like it. Because I really feel like it can make your life easier. Give me the chance to do that, I suppose.

Joe: The reason I had you on the show is because I’m such a big fan of the product. Definitely everybody will be linked to a bunch of stuff that we talked about, including Metorik, in the show notes. Which you can find over at HowIBuilt.it. Once again Bryce, thanks so much for your time. I appreciate it.

Bryce: My pleasure, thank you.

Outro: What a great guy to talk to, and the creator of a fantastic and very useful tool. I use Metorik every day to track sales, trends, and run reports. And Bryce’s advice about engagement and stats are top notch.

And Thanks again to our sponsors Pantheon and Creator Courses. Definitely check them out. Both are teaching you all about Gutenberg and WordPress 5.0.

For all of the show notes, head over to howibuilt.it/86/. If you like the show, head over to Apple Podcasts and leaving us a rating and review. It helps people discover us! You can also join the Facebook community over at howibuilt.it/facebook/. I want to build a strong community for this podcast, and Facebook is the place to do it.

Thanks for joining me. Before you get out there and build something, I’d love if you stuck around and listened to this preview of Creator Toolkit, a new podcast by me, starting Thurday, August 9^th.

Josh Koenig and Pantheon

February 19, 2018

Show Notes

View on separate page

Transcript

Intro: Hey everybody and welcome to another episode of How I Built It! In today’s episode, I get to talk to Josh Koenig, founder of Pantheon. Josh is energetic, passionate, and knowledgable in his field that’s a killer combination you love to see in company leaders. We talk all about the technologically cool stuff Pantheon is doing, so if you’re a developer, you don’t want to miss this one. But first, a word from our sponsors.

Sponsors: This season of How I Built It is brought to you by two fantastic sponsors. The first is Liquid Web. If you’re running a membership site, an online course, or even a real estate site on word press, you’ve likely already discovered many hosts that have optimized their platforms for a logged out experience, where they cash everything. Sites on their hardware are great for your sales and landing pages, but struggle when your users start logging in. At that point, your site is as slow as if you were on three dollar hosting. Liquid Web built their managed word press platform optimized for sites that want speed and performance, regardless of whether a customer is logged in or logged out. Trust me on this, I’ve tried it out and it’s fast, seriously fast. Now, with their single site plan, Liquid Web is a no-brainer for anyone whose site is actually part of their business, and not just a site promoting their business. Check out the rest of the features on their platform by visiting them at buildpodcast.net/liquid web. That’s buildpodcast.net/liquid web.

It’s also brought to you by Jilt. Jilt is the easiest way to recover abandoned shopping carts on woo commerce, easy digital downloads and Shopify. Your e-commerce clients could be leaving literally thousands of dollars on the table and here’s why. 70% of all shopping carts are abandoned prior to checkout. Yes, you heard that right, 70% of shoppers never make it to checkout. That’s why you need to introduce your clients to Jilt. Jilt uses proven recovery tactics to rescue that lost revenue. It’s an easy win that let’s you boost your clients revenue by as much as 15% and it only takes 15 minutes of your time to set up. Jilt fully integrates with woo commerce, EDD and Shopify. You can completely customize the recovery emails that Jilt sends, to match your clients branding using it’s powerful dragon drop editor. Or by digging into the HTML and CSS. Even better, Jilt’s fair pricing means your clients pay only for the customers they actually engage. You get to earn a cut of that through Jilt’s partner program. Whether you have clients that process one sale per month or 10,000 sales per month, be the hero and help them supercharge their revenue with Jilt. Check them out at builtpodcast.net/jilt. That’s builtpodcast.net/J-I-L-T.

And now…on with the show!

Joe: Hey everybody, welcome to another episode of How I Built It, the podcast that asks, “How did you build that?” Today, my guest is Josh Koenig of Pantheon. Josh, how are you today?

Josh: I’m doing great, how are you Joe?

Joe: I am fantastic. Thank you so much for joining me, I’m really excited to talk to you about your new global CDN tool that you rolled out, I guess a couple of months ago at this point, or at least at the point where it’s being released, this episode is being released.

So, why don’t we start at the beginning though. Why don’t you tell us a little bit about who you are, a little bit about Pantheon, for those of you who don’t know but Pantheon was a sponsor in season two, and then the global CDN tool?

Josh: Sure. So, my background is I’ve been working on the web in one way shape or form for now 20 years, which makes me feel old, maybe I am getting a little old. I was lucky enough to get turned on to this technology in the very early and even though I have a background in fine arts from an education standpoint I always was sort of put myself through school and then made my living with my internet skills.

Over the years that sort of developed more and more into a real career. I worked in the Drupal community for a number of years in the aughts as that technology was becoming more and more popular and taking off, and started a consultancy based in San Francisco that’s still around and still doing great work for clients serving their needs to move their cause, or their business, or their brand forward on the internet.

And then started Pantheon with a couple of my co-founders from the consultancy and a friend of ours we had made through the community, and it was really based on our observation that in our consulting work we were doing some things that were totally different from project to project. You know, design, content strategy, user flows, business logic in your website or in your application. Those are all very different depending on what your website is, what it’s trying to accomplish in the world, who you are as an organization or a business.

But we were also doing this infrastructure work. Setting up development workflows, especially as we would on larger and larger projects there would often be developers who worked for the client, or they would have hired multiple agencies, or they’d want to accelerate things by adding contractors. So, having infrastructure for doing website development that really supported a larger team using version control, and merging workflows, and so forth. And then actually launching some of these sites that are on the bigger side that maybe are going to get a lot more traffic in such a way that they don’t fall over when you put them online.

And although the projects we were doing were super different from like nonprofits, to businesses, to even some government work, and media stuff, these infrastructure components of how we built the site collaboratively and then how we launched it successfully were kind of the same. We were kind of learning as we went, but it wasn’t like they were different based on the requirements of the project.

So, we thought to ourselves “Hey, this stuff is potentially could be turned from pretty high end consulting work into a service since it’s repeatable, and we could deliver it to everybody so that we have more successful projects in the open source web community.” Because we could see some of our peers in this space, you hear sort of stories in the hallway track at a conference or something about less successful projects, and often times the thing that fell apart was either “We couldn’t work together as a team. We were under pressure to meet our deadlines. We tried to add more developers and it made everything worse.” Like that’s a common story.

Josh: And some of that’s just Mythical Man-Month thinking and you can’t just add more people to solve something, but you really can’t add people if you don’t have a structured way to collaborate because you just start stepping on each others toes and it makes things worse. Or “We managed to hit our deadline for our client but then we launched the site and it crashed, and we spent the next two weeks in a really uncomfortable fire fighting mode just to get the website stable and online.” And those are experiences that don’t leave clients happy, and lead to burnout on the teams that work on them.

And so we really felt there was this need if you could just solve both of those problems with snapping your fingers to turn on the service and that would help so many more people successfully develop web projects, ambitious web projects using this open source technology.

So, that’s where Pantheon came from, and its been a wild ride since then. We started out, you know, very much anchored in our roots in the Drupal community but started immediately seeing interest from the WordPress world and added WordPress support after a couple of years, and we’ve just been kind of going and going since then. Its been awesome.

Joe: Man, that’s fantastic. And using Pantheon I can tell that there’s a lot of developer focused stuff right? When I got set up with an account I didn’t really have to do anything to get push, or get deployed to work, it just kind of worked, which is not my experience on any other hosting platform so I really enjoyed that. And you’re absolutely, from a large teams perspective I’ve done agency work for Fortune 100 companies, big teams need to scale well in order for a project to be successful and any tools that could help that really, really help that. Like it’s very noticeable. So, that sounds fantastic. I’m really glad that Pantheon exists in the space for that reason.

Josh: Thanks. And I mean, there’s always more for us to do. We’re always looking at other aspects of this world, the operations aspects of websites. So, whether that’s how you manage your collaboration, manage access, how you think about reusable components across multiple websites, how you think about things like the data about how the website is performing, including my background as a server side engineer I think about performance tuning a lot and I look at new relic graphs. I enjoy doing that sort of thing, but also there’s the aspect of performance like “Is it actually working? Are you getting the results in the real world you want from your website?”

And so I kind of conceive of this world of hosting as a critical component of this bigger world of website operations right? Like if you don’t have a website online nothing else really matters, but once you have that core infrastructure problem solved all these other challenges emerge almost immediately. Like, “Okay, so your websites online but can you actually make changes to it? Do you feel like you have a development workflow that lets you try out new things? Do you have metrics on your website that tell you whether the things you’re changing are making it better or worse?”

It’s interesting because I think a lot of people in this world are still … I think there’s a lot of people for whom they just want to have a simple website online that represents them so they can be found there, and that’s great and we’re getting better and better tools for achieving that high quality but base level web presence, but then you have a lot of people who are actually trying to make something happen with their website. They want have an effect, they want to have impact, but a lot of the thinking in the industry is still stuck in the world of, I guess like on the agency side there’s still the waterfall patterns of like Mad Men style agency sales, and on the engineering side you still have patterns that are from the era shrink wrap development.

So, it’s kind of like everyone thinks “We’re going to do a big redesign and we’re going to spend two years redesigning and reimplementing our website and it’ll be this huge project, and at the end we’ll launch it and then everything will be solved.” And actually it’s problematic because I’ve been on the implementation side of those sorts of projects. And yes, there are times when you need to re-platform, there are times you need to redesign, there are times when you need to just take a fresh look at everything. That is absolutely true, but the notion that you should burn through 90% of your budget just to get that first iteration live is setting yourself up for failure.

If you don’t have a way to actually continue to make changes to the site, whether that’s content changes, which people should, you know, with WordPress and so forth you can make content, that’s the beauty of it, you can make content changes at the snap of a finger. You can respond within an hour.

But also changes to your functionality, to your layouts, to your design. All those sorts of things that aren’t just tweaking the words in a post, you want to be able to iterate on those things too, and then measure the results, and that’s how really great websites get really great because people do this.

But I think that’s still something that a lot of people, they feel like it’s out of reach for them and it’s often because they don’t have ability to do this kind of collaborative work. And that’s like to me the bigger picture problem we’re trying to solve isn’t just this core infrastructure kind of problem, although I’m going to talk about our CDN, which is very much a core infrastructure kind of capability with the How We Built It because it’s a really interesting story.

But that core infrastructure problem is just the base problem. The next level up is “Okay, how do we work on this thing and continually make it better and learn while we’re doing it?” Because that’s what you have to do to make a really great product.

Joe: Yeah absolutely, and we certainly do have the tools today to do things like that. I mean, you guys are at the forefront of kind of adding some of those tools, and as well as some of the other bigger hosting companies too, and like the developers who are creating things like Vue.js and stuff like that.

You mentioned that you started on the web 20 years ago, I’m not that much behind you. I started about 15 years ago. So, around the time we both started making a website was pretty simple. We had HTML, we had CSS. CSS was probably in line. We had maybe some graphics, everything was table based layout, and we would just FTP stuff up to a server right?

So, as you build out all of these tools to help with collaboration, especially as we have remote teams and other tools to iterate quickly not just your content but your functionality, what kind of research are you doing and who are you talking to stay on top of current trends right? Because a common conversation that a lot of developers have is “Hey, I see you’re using React, have you heard about Vue? Hey I see you’re using Vue, have you heard of a new JavaScript framework that’s going to be outdated by the time this episode comes out?” So, how do you stay on top of that stuff?

Josh: Well honestly our business has an overt focus on the agency world in a developer focus, as you called out initially, and then on a business level we have an agency sort of focus because that’s where developers, the rubber meets the road, and agencies are generally pushing things forward a little bit ahead of where an in house developer or an IT person might just because the demands on them are to constantly be innovative. That’s how they differentiate.

So, I’m really lucky in that we have this big network of great agencies who we’re targeting our collaboration tools towards and we get to talk to them. There’s a little bit of interesting observational work that we can do just by seeing what’s launching on Pantheon. We have a wall board in the office that’s just the sites that went live in the past 24 hours and it’s awesome because it’s just this stream of the beautiful randomness that is the internet, and so you can kind of see some trends there.

Especially just literally from a design standpoint I remember when we started to see the move towards not just mobile response design but really mobile first design and the just big center hero kind of presentation on a homepage became the norm rather than the exception.

But really in terms of the question you’re asking it’s all about keeping up with the agencies and just actually having conversations with them. I’m a huge fan of the method that is generally known as customer development. And customer development if you’re not familiar with it is basically just a widely structured process of finding your ideal type customer, and that means someone who is actually using your product successfully, so you have to have a minimally viable product to get there, but hopefully your intuition can take you to that point, and then you just get in their head of while they use your product. The best interviews are almost like “Narrate for me the documentary of you using my product and tell me how it makes you feel, and where you run into trouble, and where you wish it did something that it didn’t currently do.”

And it’s interesting because you’re not just … It’s a little different than doing observational user experience research where you’re just watching people and intentionally not engaging them because you don’t want to influence the data you’re collecting, because what you get from customer development is that you actually get, if you do it well, you get into the thought processes of the people who are using your product, and that can provide really amazing insights.

So, a really pat example is we always wanted to have this ability in the core of Pantheon for there to be a hand off right? Like “I built the site but it’s for a client and we’re getting close to the launch so I need to bring the client in and I want to hand off ownership to them because I feel like the right thing to do is for the client to own the infrastructure for the website, not for me to own it and then have to build them in some other way.”

And so, in talking through that with some people who were using a very early version of Pantheon we came about this notion of invite to pay, and the process of being able to input someone’s email address, have that send them to, generate a link for them that brings them to a simple page where it’s saying “So and so is inviting you to pay for this site, they recommend this package.” You can change that package if you want because ultimately you’re the one who’s paying and “Here’s where you put in your credit card information.”

And that, frankly right now we’re working on redesigning a bunch of that stuff because it hasn’t changed very much since the four and a half years ago when we implemented it, and now it looks actually out of date. It’s not up even to our style guide, but that insight of it’s actually this invitation to pay and you want to have this exchange with your client was very much a customer development driven insight.

Joe: Nice. That’s fantastic, and it totally reinforces another interview that I had for season four with Anthony Katz of Hyperice. It’s so funny, I always find the theme of the season emerges in the first couple of episodes, and he talks about this same concept of customer development right?

So, they made this sort of restorative ice pack that you can strap to your body, and he worked with professional athletes to really get it up to snuff. Like he gave it to them and just got their feedback. These are the people who always have like ice packs or ice buckets on them.

So, it’s really great to hear you reinforce that because I think a lot of developers, myself included, will build something to our perception and then be a little resistive to feedback. You get maybe a little bit attached to the thing you’re building, you built it in a way that you think is good, but everybody works differently and you need to have that target, that perfect customer as you said, in mind to really build something great that they’re going to use.

So, I really love that and I love that it’s coming again in two episodes. So, very cool.

Josh: I can’t say enough about that, and the one thing that’s interesting to me, you mentioned you get attached to the things you build, and I am definitely guilty of that as I think any honest developer would say they can empathize with being in that situation, and it’s a tough line to work because the great products are not just the result of having people tell you what they want and then doing that. It’s the Henry Ford, like “If I just listened to my customers they would have built a faster horse.” Right?

Joe: Right.

Josh: So, the great products are a combination of your own expertise, or some kind of, whether that’s a technology expertise or something else, and then the ability to have an insight into the needs of the customer that they might not necessarily directly vocalize themselves.

And that’s what customer development can give you because you’re figuring out how to put yourself enough in their shoes that you could have a light bulb moment, and then you have to obviously market test it to make sure you’re not just off on a wild goose chase, but it’s that ability to have an insight into the world of your customer in addition to also listening to what pleases them or what displeases them.

Joe: Nice. That’s fantastic. So, as we’re talking about this sort of stuff, and you work with a lot of developers, and we’re a little more through half way through the show right now, but there’s been a big focus on performance and so why don’t we talk about this interesting story that you have about how you’ve built this global CDN tool. So, maybe you can start with exactly what that is, and for anybody listening who don’t know what a CDN is if you could define it, and then talk about how you built it.

Josh: Sure, so CDN is an acronym that stands for content distribution network, and basically the core concept of a CDN since the earliest days of the internet has been get the stuff required for a user to see a website closer to the user. So, the idea is that we have our phones, or our desktops, or our laptops connected to the internet, I type in a URL, I hit enter, that actually begins this very complex dance of TCP, IP, and HDP protocol transactions that require the device that I’m on to talk to a server somewhere right? Usually some internet end point and that could be halfway across the country, or internationally halfway around the world.

And while packets move at almost the speed of light, not quite because actually in fiber optic cables you’re not really moving at light speed because stuff’s bouncing around. While packets almost at the speed of light, the number of back and forth pings that have to happen just to establish a connection, to then negotiate HTTPS is you’re working securely, which everyone should be going forward, that’s now the norm, and then to start to request the assets and receive the assets that could actually be as many as five or six round trips between your device and whatever the end point that its talking to is before the first bit of real data begins to flow.

And so if you can say, “Hey, let’s move the assets, or let’s move that endpoint from halfway across the globe to maybe halfway across the State that I live in, or perhaps halfway across town.” Then that’s going to result in a much, much faster and more pleasant experience for customers.

So, CDN’s have been around since the beginning of the web and they were originally very much about just distributing the images, because they’re static assets, they don’t change that often if at all, and it is a little bit like that FTPing mindset. We’re just going to bulk load this information off to all of these different points of presence around the world and then websites, people who are loading websites will be able to access the data from those locations and stuff will load faster. That’s a core concept.

In a modern, and actually just to go back to your “When we first started in this business the web was real simple.” You had some HTML, you had some images, you had some CSS, you would FTP them up to a thing. If that’s how you built your website then CDNing is pretty straight forward because you just distribute the HTML file to all of the different points of presence as well. That’s another acronym in CDN land is point of presence, or PoP.

So, in the modern world where we don’t just have static content, where we’re actually using dynamic publishing systems it becomes a lot more interesting because in order to get the experience that really, really will knock peoples socks off, that will really delight your end users, your customers, you have to figure out how to get the full response … You want to get the end point connection to be close to the customer, you want to be able to negotiate your HTTPS close to the customer, and you want to be able to deliver the full page response, if possible, from a location that is as close to the customer as can be.

And we, for our background as a technology we had this interesting capability in Pantheon. We’ve always been kind of associated with high performance and speed, and that really matters because pick your metric, any business metric that you could care to apply to your website will be positively by a faster website, or negatively influenced by a slower website. There’s a great website speed tester that Google uses called Think with Google, and it’s particularly focused on the mobile use case, but it will actually give you an estimate of how much traffic you’re losing because people don’t even wait for your site to load. In some cases it could be 30 to 40% of your traffic is never even seeing your page because they’re not waiting for it to load.

Likewise HubSpot has a ton of metrics around conversion rates and so forth, and how significantly those can be influenced by how fast the web experience is for the customer. You know, how much more likely people are to abandon a shopping cart, how much more likely people are to disengage with the content if they’re having to wait.

So, everybody recognizes that the website being fast is a real business value, but when you’re not just shoving HTML files out onto a CDN how do you do that? So, Pantheon has originally had a varnish layer as part of the platform, and that’s a technology that’s used, again, to accelerate website delivery. It’s kind of like from a WordPress perspective it’s kind of analogous to pick your favorite page caching system, just like much faster and much more scalable because it’s a technology that was built from the ground up just to do that rather than being a WordPress plugin.

So, it’s orders of magnitude faster at delivering cached pages and can deliver thousands of them a second without breaking a sweat. So, it’s a really cool technology that we have learned how to use and it’s one of the ways that when we were doing our consulting work we would launch websites and never have to worry about a big PR hit crashing the site. So, we just built that into the platform, but it was built into the platform in the primary data center where we had.

So, what we did when we started to think about the global CDN and kind of the next generation of our infrastructure was we thought “What if we could take this varnish layer that does super fast page responses and actually push it out so that it was distributed world wide?” And we worked with a partner, a technology partner, a company called Fastly that is sort of a CDN/edge computing platform, and we used their network and their points of presence to essentially reimplement what had been built into Pantheon from the beginning with our varnish cache layer on top of Fastly’s network.

So, instead of having one point of presence with a super fast cache in Chicago, we have 40 points of presence around the world that all have their fast cache’s. And that’s actually, I did some benchmarks of us against us, like old us versus new us, and it like cuts down page response times by sometimes as much as 40 or 50%. It’s awesome.

Joe: Wow. And that’s without really the developer having to do anything right? Because I’m a frontend developer and I know that I can optimize my site by using SVG instead of like giant JPEGs, I can load nonessential JavaScript and CSS in the footer so it’s not like a header blocking stuff and things like that so I can speed up my site, but it sounds like I might just be able to upload reasonably optimized site, not a fully optimized site and still see returns because of this global CDN. Is that right?

Josh: Yeah. That’s the idea. I don’t want to discount the virtue of frontend development – because you actually really do need both because even the fastest infrastructure in the world with a poorly optimized site will still probably not deliver. Like you might get the first parts of the page to load and maybe the user will see them depending on how the page is structured, but you can totally screw it up right? It’s still possible. You know, if you’ve got like weird slow loading ad blocks that are causing all of the content to wait to render and so forth. So, it’s only half the battle in my mind, but the great part about it is that you just get it. You don’t have to configure anything. You don’t have to worry about setting stuff up. You don’t have to worry about weird domain names. You don’t have to worry, and you get and you get it with HTTPS by default out of the box from the beginning, it’s not like a thing that you have to worry about including later. It’s just built into the platform for every site.

And so it just becomes this new standard of “Well, now that have like a plus, plus infrastructure that can deliver this type of 100 millisecond, or sometimes even faster page response time. That’s the engine that I’ve got, now the site is kind of like, okay I’ve got to make sure that I have really, really well defined, a really aerodynamically streamed body chassis, and I need really good steering, and I’ve got to have the best tires, and get high octane guys.” You’ve got to have all of those things to win a race, but what we do is we just give you a world class engine basically without having to do any work.

Joe: Nice. Well that’s a perfect analogy and I love that. So, you guys rolled this out over the Summer, is that right?

Josh: That’s right.

Joe: Cool. So, what are your plans for this? Is it out now? Is it kind of just doing its own thing or do you have more plans for it in the coming months?

Josh: Oh yeah. We have big plans. So, we rolled it out, and we rolled it out first by saying “Okay, all new sites will be on this by default.” Because that’s the goal is to make it this kind of built in experience, and we identified a good 40% of our existing sites that we could just switch over because they were using a CNAME for their DNS and that’s something we could actually redirect, and we did that, but we’re still in the long tail of encouraging customers to upgrade. You know, it gives them HTTPS for free, which used to cost extra on Pantheon, and they get faster page performance, but they do have to make a DNS change and sometimes people want to test things very thoroughly before they go live.

So, it’s not much work, but it’s not zero work for existing customers and so we’ve been spending the past couple of months really working with customers who had questions, or concerns, a few people who had use cases where they had very, very legacy systems that were integrating with their website. One of the things that we did to deliver the HTTPS for free was we’re using Let’s Encrypt, and we’re using a really modern version of the TLS protocol, that’s the successor protocol to SSL, and that lets us do HTTPS really economically and with the highest possible performance, but if you have really, really old Java apps, I mean very, very few people have these, but if you’ve got a Windows XP user base, or an old school Blackberry user base some of those devices aren’t able to speak those new protocols so we’re having a customers who have concerns around that.

So, we’re working around all of these edge cases right now, but looking forward there’s a number of things that I’m super stoked about.

One is that we’re going to have an ability to offer some really amazing benefits for websites that are super sensitive around their uptime. Like, nobody wants their website to be down, but then there’s some websites where they want to measure their potential downtime in single digit minutes per month with like a three nine’s type SLA, and we’ll be able to use the global CDN to meet some of those use cases, which is going to be super cool. I mean, I think we’ll be able to bring some benefits to all customers, but we’ll be able to really, for customers who demand that super high level of service we’ll be able to offer that.

We’ll also be able to offer a bunch of neat features around potentially doing some stuff with auto-optimizing images. Kind of similar and maybe a little bit better than what Photon does where you can have images be auto-cropped, auto-scaled, et cetera by the CDN and not by WordPress, which can be advantageous in a number of cases.

And then also being able to add some of the cool developer features for doing things like detecting mobile traffic, serving alternate pages based on geography automatically, being able to get in this notion of personalization as a part of your website strategy. That is the marriage of really good content strategy with really good frontend development, but if you don’t have the infrastructure to support it it can be very kludgy to put together. So, well have some stuff to offer there.

And then some really neat features for people who run like large portfolio’s of websites. So, the ability to share some assets across all of those sites, to be able to do what we call domain masking where you might have like “Mybigcompany.com/blog” is a specific WordPress instance, and then “Mybigcompany.com/community” is actually a Drupal instance, and “Mybigcompany.com/store” is a Woo Commerce instance, and actually run those as separate applications so they can all have their own development pipelines and their own deployment pipelines, but actually mask them all under the main domain without having to just manage a janky server somewhere with like three subdirectories with three instances of the app in them.

So, being able to support that type of architecture on Pantheon.

Joe: Wow, that’s really cool. So, there’s a lot of really neat things coming down the pike. I’ve talked to a couple of people recently about the custom content kind of experience. It sounds like a lot of people are thinking through this problem and solving on as low level as possible. It’s, like you said, advantageous to the users and the developers as well.

So now we are, well we’re slightly over time but I really enjoyed this conversation, and always I can’t end the show without asking my favorite question which is do you have any trade secrets for us?

Josh: Any trade secret? So, a trade secret of Pantheon is the company runs specifically on the Monster Rehab energy drink. It’s a lighter weight energy drink, which is low calorie and comes in these various tea based flavors, and going back to the very beginning of “Okay, we’ve got to hunker down and build this crazy infrastructure” to the day to day now of people having four engineering squads cranking away on different aspects of the product, that is by far and away the most popular beverage of the energy drink section in the office.

Joe: Nice. Very nice. I love that. I lived on Full Throttle in grad school and I don’t recommend that for a million reasons. One is just you’re getting all of the sugar that you need for an entire week in one can.

Josh: Yeah, totally.

Joe: So, awesome. Well I’ll have to check out this Monster Rehab. I’m sure it’ll come in handy for me since I have a young baby at home who is teething and so she keeps me up at night.

Josh: Woo. Yeah. Good luck with that.

Joe: Yeah. Thank you. It’s very sad from the parent point of view because she’s in pain and there’s literally nothing you can do about it except give her Tylenol.

Josh: Yeah.

Joe: Now I do want to ask you one more question, and this is purely for me, have you ever been told … First of all have you ever seen Scrubs? The TV show scrubs?

Josh: I have seen it, but I am not a watcher of it.

Joe: Gotcha. There is this one character, he’s a recurring character, his name is Sam, and you look and sound like Sam. So, I just wanted to put that in front of you and ask if you’ve ever been told that before?

Josh: So, that is … No, that’s a new one for me, but I’m going to have to check that out. I used to get one of the guys from The Office, people would say that I looked like him, and then when I was younger and in a little bit better shape and had short hair people said that I look like Brendan Frasier.

Joe: Nice.

Josh: But yeah, I’ll check that out. Sam from Scrubs?

Joe: Yeah. Now, I don’t want you to take this wrong way because in the show he’s kind of like a drug addict, and I’m not saying that about you, you just look and sound like him. Fun fact, here’s how we’ll end the show, Brendan Frasier was also a recurring character on Scrubs. So, with that Josh thank you so much for joining me today.

Josh: Hey man, thanks Joe. I really enjoyed it.

Outro: What a great conversation – thanks again to Josh for taking the time out of his busy day to talk to us. And seriously. Check out Sam from Scrubs. It’s uncanny!

And Thanks again to our sponsors – make sure to check out Liquid Web for managed WordPress hosting. I use them on all of my important sites – they are that good! They are at buildpodcast.net/liquid. They’ll give you 50% off your first 2 months just for being a listener! If you want to save your clients (or yourself) money through recovering abandoned carts, check out jilt. They are over at buildpodcast.net/jilt. And finally, if you want to build incredible websites at a fraction of the time and cost, check out Beaver Builder. I use it and I love it. They are over at buildpodcast.net/builder/

For all of the show notes, head over to howibuilt.it/66/. If you like the show, head over to Apple Podcasts and leaving us a rating and review. It helps people discover us! Finally, last week I published my brand-new Patreon page. It offers a lot better rewards, and great goals, and I’m really doubling down on it. So if you like the show and what to support it directly, head over to patreon.com/howibuiltit/. You can support the show for as little as $1/month.

One more note before we close out the show: next week we’re starting a little miniseries on building your own business. It will be 6 episodes and we’ll cover niching down, marketing and content strategy, legal info, and more. We’re kicking it off with Dr. Sherry Walling and self-assessment. How do you know you’re ready to start your own business? Find out next week!

So until then, get out there and build something!

Bryce Adams and Metorik

Sponsored by:

Show Notes

Transcript

Josh Koenig and Pantheon

Sponsored by:

Show Notes

Transcript