 Good morning, good afternoon and good evening. Welcome to the UX research functional group update. My name is Sarah and I'm your Quiz House today. So a quick recap on today's rules. I'm going to ask you a series of questions. I'd like to guess what the answer is by answering your choice A, B or C into the chat window. And you need to do that before the countdown ends. You can only guess once per question or you will be disqualified. Keep a note of how many answers you get right as unless you're supposed to go in the chat window a little later on. If you managed to correctly answer two or more out of the three questions I asked today, you will be entered into the prize for an August to win a GetLab water bottle. So question number one, in my last update I explained that analytics tracking showed that a lot of people were searching for the term SSH in our documentation. And we wanted to understand why. To find out we conducted a round of usability testing with five people to see how easy it is to add an SSH key to GetLab account. Out of the five people that we tested with, how many do you think were able to successfully do this? A lot of people are thinking B so far. A good question. I promise it's not. Okay, three, two, one, time's up. If you said B, three, you were being correct. The answer is C, all five people. Now you might be thinking, well, Sarah, was that study a waste of your time? Well, actually no it wasn't. Because even when users successfully complete a task, it doesn't necessarily mean that the process can't be improved. So for example, there were a couple of things which showed you slowed users down, like the fact that they didn't know that commands could be copied and pasted. So some users were manually typing them and they made mistakes when they did that. And it also wasn't obvious to them what part of the command might need to be able to suit their personal circumstances. So for example, in the documentation, we might reference a file, path or name, and they were actually copying that command exactly. So they were searching for something that perhaps didn't exist or was in the wrong location. So whilst these might seem like really small usability problems, when you're a user that keeps experiencing these issues one after another, you soon become frustrated. And as adding an SSH key is crucial to getting started with GateLab, it's not a good first impression. The approach should only take a minute or two and for some users, it was a lot longer than that. So question number two. So we also recently conducted some usability testing on auto DevOps. We wanted to know if users knew how to enable all stages of auto DevOps for their project. We tested the existing process with six users. As some of you may be aware, there is a banner within the interface under the auto DevOps settings, which I currently located under CI CD settings, which says auto review apps and auto deploy need a domain name and a Kubernetes cluster to work. What I'd like to know from you is out of the six users, how many overlooked this banner? Make sure B and C so far. Three, two, one, time's up. If you said B, four, you were being correct. The answer is C. Unfortunately, all six users overlooked the banner. So subsequently, no users instinctively set up a Kubernetes cluster. They actually will have to be prompted by me on how to do that and where to go to do that. So of course, we want users to know why they need to add a cluster, but also how to add a cluster without any support. Now, luckily, we weren't just testing the existing experience. Tori had actually created a prototype of what a step-by-step process for auto DevOps might look like. So the prototype proved to be a much better experience for users, but just how good was it? So to help you answer question three, we uncovered 22 usability problems with the existing auto DevOps experience. So how many of these usability problems did Tori's prototype resolve? It's up to you whether you trust Tori or not. A lot of B and C, I think a lot of people have a lot of faith in Tori. That's good. Okay, three, two, one, time's up. If you said B, 10, you would be correct. So this is a fantastic result. We've almost half a number of usability issues in just one iteration. And as an added bonus, all of the users that tested the prototype instinctively set up a Kubernetes cluster without any help from me, showing them that the step-by-step process makes it easy for them to understand why they need a cluster, and it also guides them through the process of creating a cluster so they're more successful. So that's actually all my questions for today. So please now total up your scores and post them into the chat window. Whilst you're adding up your scores, I'm going to share a little bit about what myself and Catherine are currently working on. So you may remember that I started investigating what areas of GateLab.com are areas of friction for new users. So we wanted to try and find out why paid users typically churn. We created a survey to send out to people who have left GateLab.com, but unfortunately the study was put on hold whilst we reviewed whether we actually have permission to contact these users. So I'm pleased to report that we do. So I'll be continuing with this study this month. And Catherine is currently investigating permissions within GateLab. So she's also looking into how we can improve the structure of the handbook. So well done to all those who scored two or more. I'm happy to see that a couple of people got two. Do we have any people who got three out of three? Who won last time? The prize last time went to Bob. He got himself a pair of GateLab sunglasses. Okay, cool. So I'm happy to take any questions about the research we've undertaken or what we're currently working on, if anyone has any. Hey, this is Leslie. I have a question around how do you go about getting the people who are going to be in the study or the test? Yeah, sure. So we have a research panel at GateLab. So we've currently got about a thousand users on that panel. And these are people who have opted in to kind of take any kind of studies with also like surveys and usability tests and things like that. So that's normally our first point of call. And then if we need people that are perhaps like, we do have some non-GitLab users on the panel who, you know, might be using a competitor to like get pulled or a bit bucket. But generally, if we need to recruit those types of users, then we may go to a third party recruitment service. The one that we typically use is respondent. And do we capture anything about these people, like in terms of where they come from, the type of company? Oh yeah, definitely. We always screen every user before they take part in a study because it's important that, for example, like people who were adding an SSH key to their account, we couldn't use like seasoned GitLab users because they would have gone through that process. So we needed people who were either very new to GitLab. So, you know, we would see when they'd signed up to GitLab. We actually had one user who has previously been a GitLab user but didn't really do anything with his account. And the other four users were all like completely new to GitLab. They would get help users or a bit bucket user. So every person goes through a screening process, basically. Thanks. Any other questions? Okay, thank you very much everyone for taking part. Enjoy the rest of your day.