 I'm going to go full screen and then confirm someone on mute and then tell me that you can still see my screen when I do this. We're good? Yeah, we can still see it. Awesome. Do you guys see my little bar at the bottom or is that just my Mac artifacting? All right, cool. We'll go for it. It is today, a day in June, and this is the May 2017 support functional update brought to you by myself, Lee Matos. Let's jump in and go right into the accomplishments. Very, very excited. Cindy has been working incredibly hard, leveling up in every way, shape and form, and she has been promoted from a junior to a, oh my gosh, major typo. Cindy is a support engineer. Sometimes it slips my brain as well. So Cindy is a support engineer. I'll get the slide updated after this. Chen Jay has joined as a support engineering intern. And what we're going to do is have him work on a couple of big projects with us and show us what he's got and excited to see what he can do here at GitLab. Apart from that, other things that have been on the radar community advocacy has their own Zendesk instance now. This was something that affected us for a while because we kept stepping on each other's toes and the way that Zendesk does like triggering and filtering, it would affect things. So we've separated those, decoupled that, and it now has allowed both of us to have much simpler workflows, which is incredibly awesome for both teams. So I'm happy to announce that. We also have something that we've been direly in desperate need of, which is a sync between Salesforce and Zendesk. I'll give some shout outs later to the people involved with that. We had built our own custom tool for a little bit that we were using, but it was almost done. We were hitting some edge cases. We needed to maintain it. Now we figured out how to use Zendesk's new syncing tool, and we're making that work. I'll show you an example of what that gets us in a minute. Every time I do this update, I like to show this graph. It's important to me. I think about it often on runs, in showers, at dinners. I zoomed out on the graph a little bit to give more time. This is a graph of our backlog, which I've renamed the crush log. And what I want to talk about is effectively, the smaller the backlog is, the better that is. And we talked about last time, we've been shrinking the backlog, and we've been wanting to do bulk closing of pending tickets that are in our backlog and shouldn't be. So you can see that here. We did. We did a huge bulk close. We saw what that did to us. We're planning on doing another one, slightly smaller, but we still see that the overall trend is we're bringing that backlog down, which is incredible. That's really exciting from a support managed perspective. So I'm keeping my eye incredibly closely on this as something that is an important metric for support. So zoomed out, we can see that beautiful downward slope start to form, which is great. To switch gears in a different metric, this is our solved versus created week over week metric. And the higher this is, the better. So I like this phrase because it's very confusing. Our lows keep getting higher. If we look at this chart, this is over the past 25 weeks. We had a huge streak going where we were just completely crushing, getting closing more tickets in the week than were created in that week. And before that streak started, our high was 86%. And our new low is 88%. So we've been doing better than we've ever done before. We're almost as efficient as we can be. And this is something that I'm keeping my eye on as well because we did have a dip the week that we did the major close. And this gave me some insight that I want to build another view in the next support functional update. I'll have some more insights as to what was happening and trying to pinpoint when we have these dips, what happens. So this number is really important for us because this helps me to see that the team is just about the right size where it needs to be. Geodistribution may be something that we have to think about. But size-wise, we're able to handle the load we have. And if this number starts dipping, say, into 60% to 70% on average, then the team is too small and we need to hire more. That means we're not able to keep up with the load. So this is an important number that helps us see how we're doing with load. So that's very exciting. Now, this one, just hashtag goals, this is the clearest view we've ever had in Zendes before. And this is due to the Salesforce sync. We're getting data from Salesforce. Some of this is obfuscated, so you won't be able to see. But in this row over here is company name. And effectively, before three weeks ago, we could not tell you in support. Spoiler alert, we could not tell you in support what support level they had unless they were premium. And we were doing that all manually. Now it's happening automatically. We can see some that have empty value, which is something that we need to figure out why that's happening. We can also see their number of seats. So this is a snapshot of a table of the top 20 organizations by ticket created, their support level, and their number of seats. So that gives us some insight now that I'm going to start playing with data to see if we can start to identify what orgs are we talking to the most, what orgs are really quiet value, and starting to align with sales. So this is something that we've really, really wanted. And we are 90% there now. Just auditing the data is the last bit that we need to confirm. And I'm working with Francis on the Salesforce side to get that all in tip top shape, which is really exciting. So I want to take a step back and talk for a second with some new crazy graphs, some new crazy graphs. Before I joined GitLab, I did some forex trading, foreign exchange, and you can use some graphs to spot trends. So I've wanted this graph for a long time to compare a 30-day moving average against the 60-day moving average to see kind of how they're operating, what the dance is, if there's any cycles happening. And if you're not familiar with comparing graphs like this, it's simple that I can explain right now. If the 30-day, which is the red line, is above the 60-day, that means that in the past 30 days, the average was higher than the past 60 days, which means we're trending up. If it crosses below, that means you're trending down. You could see some inflection points here where here was a huge trend up. We've started trending down and we're kind of dancing around, but they're pretty close, which is good. And we could start to see this plateau happening, which is helping support right now because we're not getting more ticket volume, which raises questions, which I want to find answers to. Is this due to lack of sales? Is this due to some new efficient dev process that's happening? What's causing tickets created to kind of plateau? Hopefully this number is not directly correlated to sales. If we find that correlation, then this graph mirrors the sales graph. If we don't find that correlation or it's loosely correlated, that's great. This is good news. So this is something that I'm keeping an eye on. The other graph, the counterpart to this graph is the solved graph. And we can see here that the 30-day, which is green versus the 60-day, we've been consistently solving more tickets than we've ever solved before, and we're trending way up. So if we overlay both of these together into the graph that I call the big Papa graph, because there's a lot of stuff going on, it's a little noisy, but I deal with it for now. We could see we've outpaced and we're starting to crush. I'm starting to get into that backlog, which is exactly where we want to be. So this is something that's going to help us see as these lines start to move where we're at, how things feel, comparing emotions to data, and doing support, because that's what we do. So last time that I did a functional update, Sid asked the question, how can development reduce tickets? So I spent some time thinking about this and I've seen some stuff from Jim in sales and he's brought up some points and we need to figure out a way to make sure that we're weighing paying customer feedback higher. Right now that's a hard thing to do and I saw Mark P was talking about maybe we can have some kind of Chrome extension that lets us see on gitlab.com, who's a paying customer or something like that. But yeah, we have to start figuring out how to know when we're interacting with paying customers and making sure that things are tagged and kind of bubbling up is something that I think will help reduce tickets by keeping paying customers happy. Another thing that's on my mind is turnaround time. In a couple of our places on about.gitlab.com we talk about our dev process and how we're pretty quick so we can move quickly and we talk about a one month is about as long as it takes to get things in, but it seems to be taking two to three and customers are saying this to us. Customers are saying it's taking a little bit longer. How long is it gonna take? And this comes into the way that we do scheduling and things like that, which I think works but I think is difficult and customers can feel sometimes that we're not moving fast enough. So that's a challenging thing to think about. We move very fast and we know it but sometimes customers may even feel like we're not moving fast enough and that's just figuring out how to identify and solve their needs. So this is something that's on our brain but we need to work with development to work this out as well. I wanna shout out developers who've been crushing regressions. It has been helping a ton. We feel it. Regressions are getting solved faster than ever so that's great. We just gotta start on the feedback side focusing there and another piece that I've noticed a small trend we sometimes deprecate things or in our moving fast we'll get rid of things and it ends up affecting customers and that's something that we have to also figure out how we can communicate that. This is the eternal dev struggle. I know that this is something that's hard. How, when do we communicate at what fidelity? But I do wanna acknowledge that that is playing a part. We have not beat that struggle yet. Looking towards the hashtag future goals there are things that are on my mind and on the support team's mind things that we wanna figure out. We wanna build a process for prospects and leads to get support. Right now it's kind of fuzzy. It's kind of throw it to us in Zendesk forward us the email and that's the way that it works which works but I want a little bit more formalized process. As we develop these Zendesk and Salesforce bridge we're gonna start tightening down and start adjusting SLAs for paying customers but we want prospects and leads to also be able to get support. They're not paying us yet. We hope they're gonna pay us. So we need to make sure that we build proper channels that they can get support and that we can set up proper SLAs and things like that. So shout out to Molly because she's been talking with me a lot about this and this is something that we want to work on. So this is on my brain. Another thing that's on the radar is a metric to track the number of tickets from unknown organizations. That's gonna help us see the flaws in our Salesforce to Zendesk bridge and that number should start dropping as we know more orgs and as we start to fix our deficiencies. So that's a metric that Stan and I are gonna work on to monitor and make sure that we're adopting this bridge and it's working and we're solving the problem. Something else that's huge, huge, huge on our brain that I'm really excited about. Shout out to the build team for getting Omnibus HA, for getting PG Bouncer into Omnibus. This is something that we in support wanna start really jumping on, getting our Omnibus HA up and running, starting to get HA stronger, better, faster. And if I can step back for a second, this opens up a question, this opens up kind of a dialogue that I've had a lot of conversations about this idea that GitLab is an open core product and GitLab EE is our product that we sell and we are the support team that supports GitLab EE. So there's kind of sometimes this word play that happens where we may use the word like support in the sense that GitLab, the product supports Apache. GitLab, the product supports HA. But really the thing that we wanna say is GitLab is compatible with these things. And then we need to define what GitLab, the company, GitLab Inc. and the GitLab support team actually helps customers support. And Omnibus HA is kind of one of these first steps into that where we wanna start staking a claim and saying we know Omnibus HA and we will build and maintain and own. And if you deploy Omnibus HA, you can absolutely expect our support team to support it. You can make GitLab HA in other ways, but we want to own Omnibus HA. So that's something that we're thinking about, this question of compatibility versus supported. And so that's something that Omnibus HA is where we're starting to dive in and really want to own what we support. The other thing that getting the data from Salesforce gets us really exciting stuff where I wanna start doing account segmentation based on number of seats and then seeing number of tickets open per month, per quarter, per year, seeing if we can come up with correlations to find the sweet spot because there might be a support sweet spot of a customer that is the customers of a certain size create a certain number of tickets and generate a certain amount of value and these are our dream customers. There's a chance that that will form. May not, but I would love to get that. So that's something we're working on. Mark Bell from Sales has talked to me about getting Zendesk data into Salesforce. This is something that's on my brain as well. I wanna make sure that Sales has all of the data they need to be actionable. So this is part of that bridge and making sure that we can push that data back into Salesforce. So Francis is gonna be helping me with that as well. And the last thing, which is something that I know has been on SIDS mine, Stan Hughes mine, my mind, support teams mine, sales mind. So keep this, this is a public presentation but keep this close to your chest. Sales, I know you're chomping at the bit for this to market it this way. We want to start being able to market our support as 24-5 coverage and we are testing that right now. I will absolutely announce on the team call when that is ready, expect about a month or so before we are 100% firm, solid, confident that we are ready to deliver that in that way. We're very close, but I wanna make sure that we have all the things set up so that it'll be an avenue for success versus a set up for failure. So that's something that's in our radar and that we're really excited about because that helps us market our service better and it is closer to what the reality is. So that's really exciting. My last slide, my favorite slide, always shout outs. A special shout out goes out to Francis Paul, Chad and Stan because you guys were hugely instrumental in getting Salesforce synced with Zendesk. So that was great. That has given us leaps and bounds. The community advocacy team, you have been so helpful in getting the transition. This was the smoothest transition I've ever done in any type of transitionary way to get your Zendesk instance set up, to get things off, to move things to shut down your access and that was just so smooth. So I have to thank that team. And I wanna give a shout out to the support team. Pairing sessions have been going, popping off like crazy and people have been pairing all the time and I've seen a ton, a ton of growth and challenging tickets and things getting resolved faster than ever and it shows in the metrics. So it's great when the metrics align and it's beautiful. So shout out to everyone there. So I will pop down now to the last slide and I will look in the chat to see if anybody has any questions for me. I see 24, the chat was popping off. So let's see what's going on. I got a congrats for Cindy, Chenje, Sean McGowan will show me what you've got. Jim Ash says, how good is reporting in Zendesk? It's kind of a dark art and it's one of these things where I'll be 100% honest with you. Like one of my secret skills is like the dark art of Zendesk reporting. They use a tool called Good Data which has a query language that's almost like SQL except really annoying. So the reporting is okay but there is an opportunity if we build out Karu, if I'm pronouncing that correctly, if we get a data warehouse setup that if we get a data warehouse we can translate our queries right into that and move off of Zendesk reporting but it gets things done for now. Jim asks, are all of those tickets from paying customers? Right now Jim, most of them are in the future I will have a 100% confident answer of you of how many tickets are coming from paying customers, how many tickets are coming from prospects, how many tickets came in that shouldn't have come in and we'll figure those out. Let's see what else we got. Plus one compatibility, Zendesk reporting. This board is crushing, cool. Cool, so I don't see any other questions in chat if anybody wants to vocalize questions I'm happy to answer them with what I can. If not, I'll see y'all as on a team call very shortly. All right, cool. Thank you team. I hope that you have a great day. I am gonna have the best day I've ever had because I've decided that right now and I will see you all in a team call, peace. Hi Paul and Francis, can you both push leave meeting on the FGU so the team call can start?