 Live from San Francisco, it's theCUBE. Covering Google Cloud Next 2018. Brought to you by Google Cloud and its ecosystem partners. Hey, welcome back everyone. We're here live in San Francisco for Google Cloud, Google Next 2018. I'm John Furrier, my co-host Dave Vellante here. Next guest is Mike Fisher, who's the CTO of Etsy. Welcome to theCUBE, thanks for joining us. Thank you, thanks for having me. So, love the company Etsy. What a great success story. You know, it's been around for a while, started in a Brooklyn apartment, became this great marketplace, very community-oriented, very tech-savvy. You've been there doing a lot of work at Etsy from the beginning now, the CTO. Yes. What's going on with Etsy? You guys have announced a migration from having a data center to Google Cloud, looked at a big bake-off in other clouds. How's that going? Yeah, yeah, that's exactly right. So, when I arrived there full-time a year ago, one of the things that Josh asked me to look at was, should we be in our own data centers and manage our own infrastructure? As you mentioned, Etsy has a very strong engineering culture and for the 13 years, they've really been innovators. And when they first started, they were some of the first people doing the continuous integration, continuous deployment, the monitor everything, but back then, there weren't all of these tools and companies that could help them. And fast-forward 13 years, there's now companies that can help us do some of that stuff. And so, one of those that they can help us with is infrastructure. And so, we did a big bake-off with lots of providers and ultimately announced in December that we were going to partner with Google and migrate to the cloud. So, we started in January with our migration efforts. We're looking at about two years to get it completely done, but almost right away, we were doing stuff like serving all of our images for the marketplace from Google Cloud. How do you do a bake-off on the cloud? I mean, you don't just put a box out and say, okay, run the app and put that to stopwatch. How do you do a bake-off? Yeah, so great question. So, we wanted to make sure that we weren't biased in terms of somebody's perceptions or maybe a cloud provider that someone had worked with previously. Skill sets, yeah. Skill sets, exactly. So, what we did was we took this one big project of a migration and broke it down into multiple sub-projects. And from each of those, we asked them to re-architect or come up with an architecture that would work in the cloud. And that gave us a set of requirements. We had about 1,400 requirements. And from that, we could then experiment with each of the providers and do architectural discussions with them to rank them. And we ended up with scores around the 50,000 mark for each of the providers. And ultimately, the provider that won was Google and they won by over 10%, but by doing that, really focused quantifiable decision, we feel that we made the right decision for our needs. So, Mike, I got to ask Diane Green this off camera. I didn't have time in the interview, but I always bring it up to events. Dave knows where I'm going with this. So, you go to these events, AWS re-invent, Google Cloud, they throw up the magic quadrants. We're in three of the resilient magic quadrants, but Etsy's been a DevOps culture from the beginning. You mentioned they've been building their own tools, but heavy DevOps culture. So when you look at the most successful companies that have been cloud, whether they're cloud-native, whether they're a born internet company, born in the cloud, or a company in the cloud, and cloud providers, the best companies are so horizontally scalable, so elastic, so agile and versatile, use data across databases and data strategic. That profile actually wouldn't make them qualify for any magic quadrant. Because magic quadrants were built in the stove pipe days. So, how do you evaluate, I mean, that's just me ranting on the magic quadrant, and rightfully so, I just don't think it's relevant. Some pocket areas, okay, might be, applications, what not, but cloud, no. Everyone's trying to figure out, how do you determine who's a good cloud? Because it's hard, there's no scoreboard. What's your advice? I mean, how do you squint through that? You might have your own reasons, I'm sure, machine learning and what not, but how does a general purpose enterprise go on? Who's better, AWS or Google? I do think it depends for the company and what they're really trying to get out of it. For us, we started with very high-level functional requirements, and we said, the things that are important to us, besides the cost, are things like the culture fit. You mentioned, we've been heavy dev ops, or no ops, or very agile from the very beginning, and so we wanted a culture of someone we could partner with that really could meet us at the table and talk our language, and really, the other thing is, we have a mission to keep commerce human, and so part of that is, we need to do everything we have sustainably, and so another really high-level requirement from us was, someone needed to meet us at the table on sustainability and to do it the way we believe in renewable energy sources and things like that, and so we actually, as we did the bake-off, those requirements drove that decision to find a partner who culturally fit with us and would acknowledge our sustainability goals and help us meet them. But the technical requirements, essentially, you said, came within 10% of each other, so it was these other factors that weighed in and pushed Google over the edge. That's exactly right. We found that Google would come into, even before we had selected them, that they would come into the meetings and willing to really pull their chairs up and meet us halfway and say, well, we don't do that, but let's figure out a way we can do that for you. Awesome. So on the migration, how's it going? What are some of the criteria? What are these services that attracted to Google? Obviously, we see a Google Cloud that put on the big guns. They got critical services that they've been using internally, but not everyone wants to be Google or is like Google, but might want the benefits of what Google got out of those services. Spanner, big table, big query, the list goes on and on, TensorFlow, what are some of the things that got you excited technically about some of the goodness? That's obviously the open source is key culturally, that alignment there. What was the big thing? Yeah, so the Etsy marketplace, I describe as kind of an iceberg and the marketplace is actually the tip of the iceberg. Behind that, what you really don't see is the big data machine learning. So we process over a billion events a day off of the marketplace. When a user comes, when a buyer comes there and they click on things or they add to cart and ultimately purchase, we take all that data and process it and then feed our machine learning engines to then make better recommendations, better ranking, better search results. And so the things that really got us excited about Google is the machine learning, the big data. So things like big query and their support of a lot of the machine learning APIs that they're opening up. And what's your criteria? The marketplace is actually low latency. You saw images, what are some of the things that you need at table stakes for you? Yeah, absolutely. So we are moving from our own data centers that we manage, we have extremely low latency. So that's a big concern of ours. And as you know, in e-commerce, that really matters, that page load time matters. So things like super low latency or that we can scan out into different regions that will help improve that latency. Things like that are just the basic table stakes for us. Great, and destination, time-wise, a couple of years of doing this, how long? So we're looking, we announced in December, we're planning on a two-year migration. As I mentioned, we almost started right away with some of the smaller services, but we're looking at about 18 more months to get all of our systems and services over there. Well, Mike, I know you got to go. Thanks for coming on theCUBE. I really appreciate Sharon, the update on Etsy, great company, great culture, and Chad Dickerson over there did a great job. Remember when he was there, shout out to Chad and those guys out there, congratulations. Thank you for sharing. Thank you so much. theCUBE coverage here at Google Cloud, live in San Francisco. I'm John Furrier, Dave Vellante. Stay with us, more great coverage, day two of three days of wall-to-wall coverage. We'll be right back.