 Yes, they earned that Pearson Emily can you stand up once more because they've been Working for months without any complaints getting everything in order Bring us all here bring all the swag here organizing everything and I think we're gonna have an awesome awesome awesome week. So Once more, please give them a big hand Remember it's not a democracy here Here's the ones my phone now why does he want my phone? That makes a lot of sense But I trust her so I was prepared to hand over my own It's a nice one can we swap I'm I'm just very very glad to be here. I feel very fortunate that we're all here together It's amazing. My view is amazing It's amazing to see you all together and I know that some of you had to overcome use obstacles to be here Some of us are still overcoming use obstacles to be here. So this is my co-founder Dimitri But most of us are here and I'm very very glad to see that I want to start with the goal of the summit, which is to get to know people Outside of work We'll be working this week. So we heard last time that Working together helps you get to know people people complained that that was too little working time during the summit Which I think was a great problem to have But we plant more work time this year so you can get to know each other As a remote only company we should be able to coordinate effectively remotely so don't focus on your work focus on getting to know as many people as you can and Celebrate together how far we've come and where we are going Talking about how far we've come this was our first summit It was quite exclusive because it was with 14 members and two significant others So we got that tradition of being significant others started early Then Our time during my accommodator in Mountain View wasn't really a summit or you could view it as a three-month-long summit Our first big one was in Amsterdam Was nine months after YC and we got everyone together in a canal house, and I must admit it barely fit everyone Our last summit was in Austin in May of 2016 We were 70 people with significant others So since August of last year we doubled the size of our team So we're here with more than 200 people including more than 50 significant others and I've made it a tradition We'll go slower going forward So What do all these people do we have a mission our mission is to change all creative work from read only to read right so that everyone can continue and That is inspired by get get change software by making contributing to software from something You had to be invited for do something everyone that could get the source to do and We're now focused on making it easier to collaborate on software, but we'll not stop after that Right now many parts of our culture are not open to suggestions So our goal is to make every book every movie open to collaboration Because culture evolves much more rapidly if you can stand on the shoulders of what others have done before you and To realize our mission. We're working in a sequence of five steps We've already done the first thing become the most popular self-hosted get solution and now we're focused on number two Because we want to become the self-hosted get solution that generates the most revenue Because we need we'll need a lot of people and a lot of money to finance the rest of our journey and After we realize that we'll focus on step three and that is to have people host their private projects Don't get my calm so become the most popular sauce for private projects And we need the whole world to use that because in the end we also want people their public projects and get But there are huge network effects there So we want to convince them to use get a lot of comfort a private projects find out how good it is that it is better And then we can make the world switch And our big hairy odysse's goal is to make sure full creative work is made with it Right now we as a company have to focus on step two not get ahead of ourselves But because get matters open source Companies like pen flip and already already used get back to make collaboration on books easier And I hope that going forward many people will use get back to create our own solutions for other types of media so Why is it that we have this opportunity to grow so fast? It's because it is a generational shift in the software collaboration process Previous generation products find it very very hard to adapt to game so Basically, everyone needs to switch to a new product right now on the marketplace. That is github Atlassian and github And of those three we have a huge advantage in being the only integrated product so Github is not an integrated product It has lots of features that you need other tools for so you need to amend it with all kinds of other tools Atlassian has a full suite of products But they're not integrated We make sure that customers and users only have to install one package They only have to upgrade one package. We have one Configuration for example to integrate with Ella. We have one convention of what a happy path is We want interface one license that they need to purchase one data store We can find all the events and one code base for both server and comm So because of that github allows you to get from idea to production faster Because of one package, it's easier to install easier to upgrade because it's one configuration It's easy to set up Because there's one convention in the way things work. You don't have to customize it It makes a faster whole point if you switch teams. There's not a completely different workflow one interface makes for a lot better user experience and One license makes it easier to purchase Because we have one data store. We can do complete analytics We can do cycle analytics that show you how long it took you to get from idea to production It's very hard to do if you use all kinds of products And because we have only one code base that we maintain we can ship and a much faster rate Our rate of improvement is much faster than the competition So Those benefits are recognized in the marketplace and we've been able to exceed most of our goals for 2016 We triple our annual recurring revenue We were able to front-based 20 million at great terms in a tough fundraising market You cannot buy enterprise edition per single user so and Enterprise edition products Are now available at a premium subscription of $199 And last but certainly least we built a complete idea to production license so The tripling of revenue to start with that that is great that we exceeded that goal because It has this helps us to get to IPO in 2020 our goal And it also means that next year we have to achieve the same thing again In order to stay on that track We're really thankful for all new customers that believe in us. This is a selection of the 426 customers that We gained in the last quarter But I think even more important is that existing customers are happy with get loud We can see that they are adding a lot of users to their installation So if you take a group of customers and look at them a year later You normally see a significant drop in the number and the dollar amount At good luck. We see that 98% of our customers are still with us And I like to believe that that 2% switch to the community edition and our therefore And they spend 70 by 5% more money with us So this is very important because it shows that as get lab We are helping them to get to idea to production faster And that they're happy with the product So what kind of products do they use so right now we have three products to run get lab yourself We have give up to be an idiot. It's our open source edition and We want to be a good steward of the open source project It's very Hard to do right. It's a tricky balance, but we've proven again and again We can get it right and we have our ears open for any time we make a mistake be corrected We also use it to lockout competition We want our open source edition to be better than the great product of the competition And we use it to generate the month for offerings. It's helping us to create awareness for get lab Then we have enterprise edition started It's $39 per user per year and it started to markets of teams of more than a hundred people in the company Then we just introduced and a price edition premium It started market is teams of more than 750 people And then in the future we might introduce enterprise edition ultimate Next year it's gonna be $999 per user per year and it targets organizations with more than 5,000 people But we're very careful that when we introduce higher price offerings. We never take features away from the existing offerings People that use that can expect to get more features every month without paying a cent more So this is important because introducing higher price points can get us to a hundred million in ARR we need to grow We need to grow our average revenue per user So we're an integrated tool. We're replacing multiple paid products for the customer So large organizations need help from our marketing department our sales department our customer success teams To replace those legacy products by replacing them we can help them to ship their applications faster and If we can transform your two organizations into being a software company, they are prepared to pay more per user So that's the same case and they all need to become one If whether they're a car manufacturer now or whether they're full manufacturer whether they're a bank or a movie studio Software is eating the world and every company in the world Will have to become a software company with good luck So apart from our self-hosted products. We have two products that we host ourselves There's good luck hosted It's for users that want a single tenant solution They want their own private installation, but they don't want the burden of maintaining it We do the backups we do the upgrades We're growing that slowly and steadily We're gonna focus more to guide users to the enterprise edition variant of users and the community edition variant of it will make that a more specialized offering for basically teams that are Insisting that all of the software they use is open source Then we run the GitLab.com There's a couple of goals with GitLab.com We want to create awareness for GitLab. So we want a place where you can easily try it out We want to test our own code and get our data on our own platform because it helps us to iterate faster and prevents us from shipping boats that affect the installation of our self-hosted customers and We want to upgrade people that use it to pay GitLab.com subscriptions At last but certainly not least it's an entry point to move people to enterprise edition and GitLab hosted as they grow so you can export your products and You can export your projects from GitLab.com and import it in a self-hosted installation So GitLab.com is an important part of our future Right now we have a couple of subscriptions plans So we have the free edition it comes with limited bill in it and it's aimed at organizations that have have less than 100 users We have the bronze edition We're gonna change that price to 199 per user per month and it's gonna have more bill in it It's aimed at organizations up to 750 users Then in the future will likely Release two more subscriptions. There will be a five dollar producer for month edition that will contain the features of enterprise edition premium So that's for up to 5,000 users in your organization and we'll have a poll subscription it's going to be $49.99 per user per month and It's aimed at organizations that run on GitLab.com with more than 5,000 users Right now we're focused on saving on hosting costs So we're looking to partner with one of the major cloud providers and to optimize our use of Resources to produce our hosting bill We're very much focused on improving the latency right now. Getland.com is too slow Great, we have different initiatives to change that one of them is Githy to make the get access faster The other one is probably implementing site is to be charting the database Get access in the database are the only like choke points for GitLab.com And if we're able to solve that we can make the whole installation faster For more information, you can look at infrastructure issue 947 So we used to be a product company we ship software our customers run it We'll now have to become good at two things products and services offering a service like GitLab.com Is a fundamentally different thing and that's going to be challenged as we grow our Organization to be able to become good at two things. That's not easy We thought about hosting our own infrastructure But we're very lucky that we can stay in the cloud So we don't also have to be good at hosting because that will be a first thing and three things are pretty hard So but it's gonna you're gonna find that as we go along that will be changes to the organization to make sure we can offer Service and that we we offer that effectively and it's comes so We get lab We want to ship the whole idea to production life site and that consists of ten steps First you create you talk about an idea in chat We ship with matter most and we also have a great slack integration for that You plan an issue on our issue tracker you plan that issue with our issue boards and I think we've done a great job of making sure our issue board is completely integrated with our issue tracker But it's still easy to use so we don't need to acquire another Company we can make our issue board better Then we ship with a web terminal to get quick access to machines and to start coding in an online environment We have repositories you can commit that code to we have a CI and container registry to help with testing We allow you to review your code we have much request discussions and we have the review apps The review apps are amazing because anytime you make a proposal people can quickly test how that feels how that works then we allow you to Deploy those things to state we give up a point in order to deploy and to get it into production by deploying from chat With chat ops and last but certainly not least we give you feedback So we ship with cycle analytics that will tell you how long it took you to get from idea to production at what parts you were sucked and you need to improve that time And we're going to give feedback with monitoring So right now when you're deploying you really don't know how you've affected your metrics and we're Prometheus We're going to do a much better job there So in June during fundraising We presented shipping this to investors as our goal for 2016 at least three investors said that we couldn't ship the complete Idea to production life cycle this year. They didn't believe us. So I'm very very proud We shipped this year six months after announcing it in 8.15. I'm very very proud of all of us I want to mention Camille specifically for going above and beyond to make this happen Looking back Is very strange that this life cycle used to require so many different tools before get like integrated It was not one of our goals It's not our goal to grow an organization We rather stay small, but we have so many things to do so What happened during the year is that we also become the largest Organization where everyone works from a different location and just a different location from the location They prefer to work from And that forces us to focus on results instead of ours works It focused us to be it forces us to be very disciplined in how we communicate I am a broken record of can you write that down can you make an issue for that? Can you update the handbook? Do we really need that meeting? it also forces us to organize informal communications and We have team calls. We have virtual coffee breaks But what we're also seeing is people traveling the world to visit colleagues Yesterday Frága mentioned that housewall. So I think it's great that Informal communication doesn't isn't an accident But it's something we can organize and we can make it a better and a more fun experience I want to I'm very thankful for all the significant others that organized their life So they could be here and I want them to feel very welcome. So please give them a hand Organization has many advantages. We can hire the best people in the world No matter where they live and it allows us to be effective by having less disruptions active lab Only this week people are tapping you on the shoulder It's gonna be fun this week But we're gonna be a lot more effective after we after we get back home And we're we can focus on the things that we think are important and that will get results and Last but certainly not least. It's easy to combine work with the rest of your life if you're not start to fix ours and I Want everyone to experience the freedom to work the hours that they want to work and to to take Days off after noons off to go shopping during the day and everything like that Another thing that I think went well in 2016 is that we start to our values our values are results transparency efficiency frugality collaboration directness kindness diversity boring solutions and crookedness and I think a great example of our values and sticking with them was our walk post how we knew it was time to leave the cloud It was transparent. It was a conversation that normally happens internally and we put it out there It was efficient because we got the opinion of hundreds of experts that chimed in with the best knowledge in the world It was frugal because we our input was the day of writing and we got a huge amount of input It was a great collaboration. We were responding on the blog post people were Tweeting people were sending me lots of emails to my private address as in I can't comment publicly But I really want you to know you should use IBM power PC It It was I think a great example of directness because being direct to other people is easy What is hard to use to accept the directness? Someone gives you feedback giving feedback is Easy it's although it's on the center to make it to make sure it's structured Well, but it's hard to accept that you're wrong. So we were wrong and we changed our opinion we're not leaving the cloud and I think we learned from that that there was a more boring solution. So I was very excited with the idea of using Seth and That wasn't a boring solution and we figured that out with the help of all the people that commented and now we're gonna do the boring thing and stay Not everything went great in 2016 We the most important things like lots of things went wrong. I make lots of mistakes every day I hope that everyone makes lots of mistakes every day. Otherwise, you're probably not doing a lot and not taking a lot of decisions But the biggest problems were we had too few sales opportunities too few sql's and too few new customers Both were below what we planned for the year. It's great that despite that. We still met a revenue target but in 2017 we're doing things different to remedy the off points So we're gonna focus our marketing on enterprise adoption of the lab We're gonna generate more sales opportunities in marketing. We're gonna grow the SDR team was formerly known as the album BDR team The SDR team is now reporting into sales to make sure that they can collaborate closely with the account executives We're gonna grow a customer success team to help with getting more sales from existing customers by helping them adopt more of the product And we're gonna grow our sales team Now with all the changes that are happening and as the company is growing. It's hard to stay alive It's hard to know What are the three things I should do? What are the three things my boss should do? What are three things my boss's boss should do? So we're introducing something called OCRs. It stands for objectives and key results Everyone will have about three objectives and every objective has about three key results and The idea is that the key results of your manager relate to your own objectives It's not always a one-on-one mapping, but they should be closely aligned and that aligns the whole company from top to bottom and from bottom to top and That also ensures that everyone's three priorities are clear to them and the rest of the organization even when we grow so OCRs are not what we expect to do or plan to do what we aim to do. So it's in business So we expect to achieve only 70% of the things we put in the OCRs If you're achieving more than that, I did it really well You've got lucky where you aim to go and the thing we don't want to do is aiming to low Now our OCRs will be public Transparency is one of our values. You can already find the link on our website that show our entire goal tree for the whole company What will not be public is how you're achieving on those OCRs Because we want performance feedback to be between a boss and a report not to involve the rest of the company So what are the OCRs? so our company goal is We want to with a great team. We want to create a next-generation product used by everyone And I'm gonna Still that down. It has three objectives One have a great team Want to build the next generation product? It should be used by everyone so How do we know we have a great team? There's three elements that I had as my key objectives or key results We want to be a great company to work for So we want you to be happy and a metric will use for that is E-NPS M-P-E net promoter score. We'll ask you how likely are you to recommend the lab to a friend You'll score us. My aim is to make sure that score exceeds 4.5 on a scale of 1 to 5 And I think one of the most important things we can do is to provide you with like leadership Because people leave jobs because their boss is bad So one of one of the things I'll be focusing on is improving leadership and good luck as we grow so fast And we're promoting people fast. We have to do a great job of training those people and backing them up with support And the second thing is great interviews. I Heard how many candidates we have every week, but I couldn't believe the number and I forgot to double-check it This session not yet know how many applicants we have every week or every month 100 applicants a week It's uncountable That's a lot of people although we grow really fast. We're not growing that fast. So We have to decline the vast vast vast majority of people. It's like a single-digit percentage pulling low single digits that actually gets higher and I've applied to companies and it's a very stressful process and it's very It confronts you with yourself. So we should do a great job of following up with those people and Making sure that even though they might be rejected. It is a great experience So I think that process has improved a lot. That's also true too Only the most important part such an idea who's who've been doing a great job helping everyone there And we want to do even better So we're gonna send all those people that have declined needs of you like how likely are you? To recommend applying to get lab to a friend These are people that were just declined So we expect a half point drop in a promoter score But we're gonna strive to be Direct in our communication because one of the most annoying things I've ever found is that no one gives you Great feedback on your interview process Companies don't do that because there's a legal risk for that. We're gonna take that legal risk We're gonna strive to do our very best to give them proper feedback that helps them improve themselves We should be thinking about all the other ways in which we can help those people to become better to Hoop them up with whatever they need to become better So we're gonna be thinking about them. We're gonna find out what they think about give them up because we can either Every week create a hundred people that are going down about give them or people that are a fan of get one We'll be better and apply again in two years Then the last goal is to have great hires As we grow we're getting amazing applicants and I think you should always try to hire people that are better than yourself and I think We're doing that, but we want to keep that part of our very high Right now. We're getting better known. So we're getting more applications So that will allow us to to to put that bar high and to make sure that we only hire the people that Align with our value and that will Get the results that are needed and other in the drop description The second thing is used by everyone and that has three key results. We want to see revenue growth I've already said we need to grow at least three times this year No, they are should aim higher. So our Oka RA is not to travel but to critical in animal recurring revenue in 2017 The second thing is we want to grow the number of leads So the draft goal there to create a thousand sales opportunities in the first quarter of this year And last but certainly not least We want to grow the awareness for good luck Right now. We're number three. There's github It bucket and get one our goal in the first quarter is to overtake the bucket in Google Trends last objective Be a next-generation product few things are needed for that. First of all, we gotta make github of calm fast and reliable So 99% of every time someone tries to load a piece of your lap It should blow within a second If it's down, we're coming down as not loading within a second Second thing is want to ship awesome features that make people excited about github. So every time they release We'll try to make it to ship so great features and explain them so well that people will retweet our release tweet more than 250 times and Obviously, it's not just about the tweets that the tweets are a way to to measure how excited people are and Object isn't key results the key results to be quantifiable. So I'm open to our essence But I think this is a great way to measure The last thing is to become a cloud native developmental we'll measure that by asking developers to name The cloud native application developmental and we hope that more than half of the time they'll talk of us so I'll go into cloud native more on the following slides But I want to say why are we why are this our goals? So the first one Is because if get my comment slow people think you'd love is slow Good luck is not slow. If you've run it yourself It's more than ample fast enough and our existing customers are really happy with it But if we give the impression that get a lot of slow that hurts that that will hurt us for years to come So we've got to make sure that wherever people try to get loud. It's always fast and the second thing We basically went from followers to leaders We used to play catch-up to a competition and now we're lead Features we release such as protected branches our issue board our merger press conflict resolution All the deploy we are shipping faster than the competition and competition. It has to play catch up to us So and they're they're they are awake now and they're trying to do that. So there's only one solution We should keep shifting faster and they can copy us The last thing the cloud native application development tool I want to go into that because the world is changing Cloud native means using containers as infrastructure Infrastructure first used to be bare metal machines then it was using virtual seats and now it's moving to containers So where it used to be the mware AWS hero cool Now things are moving to container schedulers and it's specific who we need us so the once in a generation Shift to version control would get gave us an opportunity to build the company as it is today But now we have a once in a generation Opportunity infrastructure that presents an opportunity to become the dominant player here It is really really hard to deploy your app on a container scheduler today There are about 200,000 developers trying to work with that Using containers for development, but using it in production is not easy So if we can make that easier, we can become the default application Developmental for cloud native applications So get lab It's already doing a lot of things to make container schedulers as simple as using her over We have other scaling CI. We got chat ops and we got review apps Those are all 8.14 We can now auto deploy any application Where it's groovy or Java or node We sit that husband and 8 up 15 Now we're working on other things integrate monitoring deploy boards feature flanks And we'll need the help of our new proletives team to build this If we are the default for using containers Then people will move to get lab ensuring were used by everyone The containers today are too hard and we found that our ourselves we were working on other deploy And we had a very hard time wrapping our head around auto deploying how it should work And we started building buttons and features and everything Turned out in the end. We didn't need anything new in get lab We just needed to add a get labs here in the general template called auto deploy that was it okay, and some of Camille's magic in making a great container but There was no no change to get lab as an application needed It is just very hard to get started with to wrap your head around it We gotta do that work. So people that use get lab and focus on shipping their application so I want to Welcome our new Prometheus team Ben and Julia's future, please stand up. Please say hi to him in the front. What's Julius a Team member and there's two other team members that couldn't attend the summit given and Brian So then we'll also present first day at 4 p.m. So I encourage you to check out that presentation and if you want to know more there's a post recently released I encourage you to watch that video of Mark presenting about Prometheus in kid lab and there's a Infrastructure issue about all the different monitoring things that are going on so to make it concrete or we want to achieve in the first quarter of 2017 I want to present some slight where a preview of the flow that we want to achieve after the quarter ended So we want people to go to Google Cloud and create an account And then start a community disclaim cluster and install kid lab on it Why Google Cloud because they have something called Google Container Engine abbreviated as key key and That's a great way to host the Kubernetes cluster because they meet Kubernetes and they are created for you So after you install kid lab on it You create a project you create an issue and you click a button called rally it rally What rally does it rallies around the issue great it relays you around the issue What happens is a chat channel is created and All the followers of the issue are invited to that chat channel There's an idea by John made during a last summit and I think it's it's it's amazing because you should be able to switch between Working with more of a delay in an issue and contemplating things and it's switching to put more real time Communication in chat and then allow you to switch back But preserving the history and the law so that so that you did you don't miss out on anything? so you rally around the issue and you figure out you need to create a new issue, so We create an issue from chat and would be nice to include like the last three messages That you talked about You plan the issue with our issue board and And Making the issue board I think was an idea by your during our last summit and I'm very pleased to say it costs a lot less than Then you create a branch from the issue So it will show a clear branch issue But even when there is no repop Allows users to select the other deployed at lab CI template and initialize the repo with that Then you can start a web terminal make and push a simple rails out and We're thinking about the ability to create a merger press right from the web terminal So I want to make that link that you currently get make that clickable in the web terminal do you Have the merger press review apps should launch and you should be able to check that out And then if you like what you're seeing you use setups to deploy it to production to mask And then you review the monitoring So we want to show the metrics that most changed in the merge request This is a hard one because this will require a lot of work This means we have to ship from each is forget lab Not only to measure get lab itself how get lab is performing But also to measure all the applications that are deployed with get lab only when you measure those things You can be sure a change you made is improving things. So we want to give people feedback on The what change in the systems like your CV you love what change in the application like number of errors But also your business metrics. Are you making more money now that you deployed this? and this is Really hard to detect because you if you measure the whole application Over time everything changes So you need kind of a cannery deployment. We're only part of your applications Application servers are running the new version and compare that with what is happening on the other servers So we need to ship cannery deployments ship from from each is with get lab Tie that all up together so that you can see those metrics right there in the merger press Just that feedback is essential what happened after I deploy Then you can review the psychoanalytics see how well it took you and See what you can do to make it happen faster the next time What is the roadblock to get faster from idea to production? so We want to show a Denial of 10 minutes where we do what I just show because Those 200,000 developers are spending a lot of time making all of this work today and Today if you want to do this and you set it up with tools other than get lab You're spending a week of integration making that all happen. So if we can do that in 10 minutes They just have to spend only 1% of that time We already did it for another deploy and I think we can make this happen too and This will make us the cloud native application development tool and ensure our continued growth So I have a challenge If you are the first one to present the idea to production then all working on Google container engine on stage during the summit I will dance that sit shuffle from ice age our pure happiness And thanks Chad for reminding me that I have a dance So I want to conclude with two things first I want to warn you that our road ahead won't be easy. We're startup and We'll have our share of setbacks and problems We have to keep outshipping multi-billion dollar companies so It's going to be about me, right? but I think our work and mission is meaningful and I think that everyone will remember this what we're doing and I'm very certain we can achieve what we set out to We have an integrated product We'll make it a cloud native product We listen to our users and our customers and we have the best team in the world. Go get lab