 Okay, I guess we are ready to start So hello everybody. I'm very pleased to be here and more I'm happier that you are here because you are in my home city and I was actually born here and I just wanted to ask how many of you read my blog post about the practice because there are a lot of people Saying that it's good. So I really like that. Okay, cool so this session is about my story going through a kind of experience being a freelancer and then going into the enterprise and So my name is Marek Sodak, and I am a web developer for about 12 years. I am a Drupal developer for about past seven years and 47 weeks, which is almost eight years. So that's a long time I also do design research and user experience everything around user experience experience I don't have really a title that I can say like user experience ninja or something like that I'm also a founder of an inline manual which is kind of like my startup which focuses on Documentation for end users. So if you want to check it out come come over and I also founded a little company Drupal shop Back in London where I used to live for about three years Which is called atomic and limited and we are building websites using Drupal So the story So me coming from a freelance world well, I never really worked for a company only like in these past 12 years that was just like one year together being employed and There was one day I got a call from a recruiter and That recruiter just called me up and said well, we we are looking for someone in Switzerland to to belts to actually hire someone that is Drupal Themer basically and doing Drupal front end and So they called me up and I was in a position where we were actually moving I was here in Prague and we were moving to a new flat with a really nice terrace overlooking do the whole place The whole Prague. It was even nicer views than you have here in Corinthia Well there So I wasn't really sure if I want to go there. So I just said well, whatever what can you offer and You know, like I was really being mean and I just said, okay well I need these conditions and that question conditions and I said the price really high and Yeah, and then after like three days they came back and I had another interview Which was quite nice and they actually after these three days. They called me back again So just saying that we are in this position as a true Drupal people varies really high demand And you can really say anything that you want your conditions so My friends advised me like not to go there because it's a bank. It's a Swiss bank and They always told me like okay run away or don't go there. It's a rigid structure and I never experienced this kind of enterprise You know enterprise area So then I said to myself, okay. Well, what what will make me happy and you yeah That would be like this what I want to do just these simple bits And I was really specific what I want to do and that was just theming and you know theming is quite easy in Drupal So that was enough and I asked for high high salary and Then yeah, then I accept it I challenge accepted and I just went there so it took me Almost six months, I mean the contract was just for three months there the first three months I was just exploring stuff. That's the other three months when they extended it I was still exploring I was understanding what's happening there And the stuff you are working on was the support portal for about 100,000 employees there And it was a self-help support portal with that knowledge base so that's basically means that the users should be able to help themselves and you know if they have a problem they should search as for a solution and Yeah, fine without you know picking up a phone and calling support which is much more expensive and It also integrates with other systems property systems a big ones So my first observations were Exactly the same was that my friends told me so it there's a rig a rigid structure You can completely forget about being agile in this environment at least in my experience and What was really interesting for me being a freelancer and really switching jobs in between doing some contracts and stuff They there are people that actually work there like 15 years And you know they had these batches where they have their photo and they were really young there on the photo I was like, no, that's not you. It's like, okay. How long are we working here? I'll fight 15 years and or since I was 14 or something like that So that's that's really interesting and it says something about the culture And it also another place where the innovation happens and I will later explain why And mainly it is because the people because they are they're so long I think if you are working somewhere for 15 years and you are almost under still on the same position But you are just changing titles So all of them on on our level well not all of them But most of them are kind of like as I say directors But they are still on the same level they might be changing the roles of the jobs but it's kind of I would say a lot of similar to what they were doing for the past 15 years and the culture there is kind of like or Like that they are trying to please the people about like the managers everyone to keep them happy. So they don't really think about things Well, not that much We had also these monthly release cycles Which sometimes because of these proprietary systems ended up being like half a year cycles or even more Because you know, there are some dependencies and they didn't fulfill it So we were writing for them because we had some other feature and and so on it was also difficult to Decide on something or designs of You know get a sign-off for some features because the team itself The the four key people or three key people were in in three countries and there was another Indian guys who were doing testing. So it was all spread around around the world and Mainly the managers what I found out is that they don't know really Well, they thought that they know what the users needs what are the users needs and what you know, what? What they can offer and how to make it work Actually, that's what they were saying But they didn't and there were no analytics no research whatsoever, but they knew what what the user needs Then you know everyone now is kind of making some nice names. So I made one on myself It's a succinct success driven metrics based on percentage and it in practice is kind of like I had this one business analyst who came to me and That was after I did some research already and she knew that I'm doing some research and she asked me well I need to prove that this we had there like a forum communities and It wasn't it was fail basically because there were no people They weren't answering helping each other. I guess it's also because it's in the bank and She came to me and asked well, I need to make this forum the particle forum as success Can you give me some metrics? Which will prove it because she saw okay There's another forum which has just let's say to replace replies and another forum has like 20 replies So, you know obviously that was success and she wanted it in percentage. So that would be like 200 percent or yeah or more success and But if you looked into the forum where were like 20 replies you would found out that Let's say 18 replies of these were moderators That's what they didn't see Also, it wasn't really clear who is the like leader or in the project So there's no I mean this is supposed to be the managers, but that wasn't the case and Everyone was even looking up to them So they were waiting for the for the action direction or show them the right direction And if I ask someone well, where is this project heading? They didn't know they were just waiting from one release to another So after these six months I said to myself to myself well the contract ends in three months I haven't seen much progress because I was doing just theming and then development and other other things and So I thought Well, I thought to myself well if I'm going to spend or basically waste more time here What should I do? I should try to change something not be part of these Of this group where there are like 15 years there and they are just listening afterwards so Thinking should I stay or should I leave because I didn't really care if I lose the job because you know be triple developers have jobs everywhere and So my first attempt was that I was walking around the office and From time to time I was just asking these people that were there so long I was asking them like just discussing what are you doing and How you are approaching things and what is your role in the project and everyone was kind of happy because nobody asked them before You know and you have to talk to people and I learned some some interesting things So I was trying to get to know people around I was trying to bring ideas and innovation during the meetings and use my use my previous experience but It kind of failed because it was overall always by the managers whatever managers required or put in like the requirements and So I didn't even get any real feedback from from anyone just like okay. This is a nice idea So I was frustrated again, and it was like after three weeks or so The second attempt I thought okay. Well, it was the period where There were holidays So there was less pressure on the on our team on the bill team and as I said I was originally hired for as a developer theme or theme or primarily and They didn't want or they didn't thought that I will be doing this kind of work like research and stuff and So I thought okay, let's let's give it some structure. Let's let's do some kind of approach something more more Yeah convenient so I started a research and I base it on a Principles around use UCD, which is the user centric design How many of you know about the user centric design? Okay, not that many but I guess you will find out that you that's something you are doing every day, but this is the definition Basically you are focusing on the user needs wants and limitations while they are using your product and so we are really trying to implement what these Aspects in in every process every design process during when you are building the product And then if you follow that you should be able to design You know a product that will be based on or in it will support the intended users existing beliefs attitudes and behaviors and so on and This means that you will kind of build a user friendly usable Usable product which might later increase sales reach business goals in our case lower the support goals just Briefly you can use these user perspectives Which are needs and wants goes motivation and triggers optics obstacles and limitations during your research? I mean you can see this kind of less user perspectives can be long and it not always will fit for your needs so for example geography and language won't fit for every project and stuff like that so use your common sense and You see the is not just that you have to If it's called user-centric design It doesn't mean that you really need to do everything that user asks you and there is a one Check entrepreneur or boss Tomasz Batia. I don't know if you some of you might know him or at least the logo He's he was a shoemaker and this brand still exists and he always said our customer is our master And I'm sure he didn't mean it exactly or you know like Like it's written so if everyone will come back and say okay I don't like let's say this color of the shoe then they won't immediately create a whole collections of new shoes and There's also this brand quote, which says people don't want to buy a quarter inch drill They want to buy a quarter inch help each hole. So for example, if you have a If you have a guy coming to the to the shop and asking the salesman, I need a drill and the The user the salesman will just give him a drill without any questions then you know the guy will just get the drill go home and Probably make some holes, but he doesn't know you know what the drill does But if the salesman actually asked him like okay, well, what do you need to drill for and he would say well I need to put there like have the how it's called let's say the nail or something and Then if he asks more and more then he will find out that this guy doesn't need a Drill, but he needs a hole actually There are several disciplines again. It's just this is just a Yeah, one of these but you can put in anything you want or use anything you want the process It's quite simple. You just do the research then the concept which could be like wireframes Design you just create the prototypes Like low-fidelity prototypes and then you evaluate that means like you test with the users you Yeah, whatever is convenient for the project. I would add to this to other things which is deploy and learn and Question how many of you actually do something after you launch a website. Do you do any research afterwards how it's being used? Okay, but not much. I think I think that's kind of like a key thing for us for like web designers and Developers because still now if you if you have let's say some research some or if you don't use research before It's all your assumptions and your skills that you had before and After you launch something that that is the point where you can start learning new things because that's something you implement it And something you believe that would work So if you are not doing this try it once even on a small thing and you will see that you can you can improve there so I started a research and I was questioning myself. Okay, what to research There was a nice talk from Eric Reese actually and he was asking this why and if you weren't there Didn't look it up on the on the web afterwards But I started with kind of micro approach because I knew or I have seen that in the in the company itself everyone was chipping in everyone was kind of Developing their own ideas without without any research anything just based on the assumption and So I took just one thing just simple thing or I wanted to pick one thing and Analyze that and I didn't have anything like you know like you have Google analytics because it's a bank It's a close bank. You can't have that the out that the outside or or hosted somewhere else So I choose that just the simplest thing that we had and that was the access lock, which is quite geeky I would say and That's something that let's say in truly can see in the reporting and these kind of like refers and stuff That's all coming from access locks so I started with that and You know if I would go to because of course there are some reportings done by the reporting team but You know asking them it would just throw in more obstacles because they would be asking what do you want to research and yeah That would be a strange exchange. Okay, I think So I did this. I don't know if you can see it. I hope you can This is just a basic You know graph or a chart that shows What I did actually is I took the whole access lock and based on the path Like you have the clean path. So let's say within Drupal you have node slash node ID So I took this and created it kind of like a tree So that's just running through the one week of locks. I did this and immediately we saw Which parts are being used of the website? They didn't have this before or Something really in a basic shape and Then I did another one Which is like showing the user flow. This is like a dendogram It's using the D3 library in JavaScript and it was quite complex to write but you can see these the Brown Lines that are going around are actually that users are coming back all over and you can see the flow there So this was quite helpful because this is as well interactive so we can click on one one path And it will show you where the users are going and I mean they they had really a lot of fun when they saw this So I chose to use the search as my Item for researching Because the the search is kind of kind of like the most valuable functionality of the self-supporter so so let's say if I'm looking for a for a solution for a help the search needs to work and Sometimes like there was this discussion in previous session as well that we take as granted that the things will work out of the Box or whatever we take. Let's say with Drupal Good case and we were just discussing this that I'll say you have a module calendar module and you install it You just turn on the module and you take that this user interface is working because everyone is using it So you don't touch it at all and that could be from a simple smallest things like a search box How many of you actually like, you know looked into the search whether it's working for your users as it should Yeah, just for you So we tend to be really lazy because we get something for free which is there or it kind of works Okay, so this is this is the It can be seen. This is the initial Search box that I was working on because I chose just this particular part that was on the home page And it was like in a one column In the bigger column so you can see here like there's a search and find title There's a pretty face of a girl corporate style. You have their search placeholder. There's even the How it's called autocomplete. So when they were typing, you know, it gave them suggestions and there is the search button and like magnifying glass So when doing such a research you need to set your well These were my personal goals for me to keep, you know, doing these things and what I wanted to do is to show the team that we don't know much about the user experience and We don't know really what the users are doing And I also wanted to showcase some of the research approaches so they can, you know, do it themselves and understand that there is something more behind and as well make the search to be used more often and Make the user centric design one of the processes which is kind of like high level Goal that will take probably ten years From my experience now And there needed to be an output. So the output would be an analysis of how the search works and it should be delivered as a presentation because people there like presentations and It needed to be sent out to all the members and most importantly as well managers so they can notice that oops Okay, that was a wrong key. Sorry Right. So the search analyzes so we wanted to increase the use of the self-help portal itself Which will immediately it showed at least decrease the race ticket so users actually find the right solution for their problem and We assume that the search is not working based on the feedback that we got from the there There's kind of like a rating on the portal so you users can actually rate the service and they weren't very happy with the search so that's the only kind of input we got from the users and the team itself already had a lot of improvements based on these assumptions and One note about assumptions is that as I was saying about these managers that they really wanted or they really Always knew what the what the user knows It was all based just on their assumptions Which is not always the the best thing if they don't have especially if they don't if they don't have the experience So then I created a simple scenario. It's kind of like a user story What would be the ideal simplified workflow for a really good, you know User experience that they can search something as simple as that like find a search field enter the search Term click the available action to search if it's not found and refine the search read the solution Optionally perform some action like increase the rating of the solution and then exit Peacefully and tell everyone how happy they were with the service. So that was the goal And obviously it wasn't like that. So I started with Kind of researching what users are searching for And here you can see there's a blank keyword or link the blank actually is a blank that they type in blank the link keyword nick password mind and so forth and First of all You see they're like blank and what is blank? Why do they type in blank to search for something something? So I had to ask about this. What is blank keyword? And why do users are searching for that and it turned out it's a it's a feature of the portal Because if you click only on the icon it would search for the blank keyword and if there is a blank keyword It would list all the solutions in on the portal a bit of pagination, but they are like at 10,000 solutions and Then they can filter out because they have like filters on the on the side So this is kind of interesting approach that I haven't I haven't seen before anywhere So I wanted to prove whether the filters are actually being used and you can see that I'm doing really just small steps researching just one functionality step-by-step just looking what it is doing So there are these URLs addresses which you can From these like as I said, I was using the access locks And for me it was easier to dig into the code and I found out that for example The first starting page is with filters undefined and if they used filters they there would be something in the content part So it was easy to find out which ones were undefined and which were which ones were content or were holding content so I was I really wanted to know how people get to this actually blank page and I was expecting that As it was implemented, I will get many refers back and that's just because they you know Click just the button went there and then they just filtered out and Again, I found out that From home, it's like 1100 searches. Let's say from 5000 searches and From the search page actually 510 refer 110 and so on and then there were like a small part And refer is like where where where they were coming from to the search to the actual search So I thought okay home why they are coming from home. I think that there there must be some sort of like a problem Why would they just you know click the search button? That's probably not working as it should so probably the users were actually you know Used to just clicking the home. I think Yeah So that was kind of like the first problem They shouldn't go there. They should if they you know approach the website. They should start searching and You shouldn't get to the blank page Search again Why if you searched first for a keyword, why would you? you know remove the remove the keyword that you entered before and search again with the blank space you So that was another problem and I found out that's actually an issue because if you if you in that implementation if you type in let's say a Keyword you search for the first time then you search for the second one second time from the search page again It would automatically Becomes default to blank. So there is a problem with that So they weren't able to refine their search keyword And then the referral which was pointing to the to the blank page It's yeah, there was like a wrong link on the internet so if you remember we were trying to find these whether the filters are working and I found out it from these like five thousand felt searches only twenty times it was used and That's a pretty low number. So the conclusion from that was that the page The blank page the blank behavior wasn't really working as it was designed and So you can see like from these little steps. We also find we also found these The problems like the the buck with the search we found that from the home It shouldn't be like that and as well this internet broken link and with this just a simple research You can achieve these things that were always there, but nobody nobody reported them and Another conclusion was that the basic search user experience seems frustrating because it didn't really You know apply to this simple workflow that I defined before and Most importantly after doing this research these are kind of Not assumptions, you know, we have the numbers there. We have you know the What actually the users are Searching for it's not that we we think okay, we have let's say that number of hits for this solution So people are probably searching that and that and that so these are not assumptions So I did one created this presentation. I send it to send it to team members and managers and there was some feedback but Nothing happened. So That was kind of like interesting moment where one of the colleagues came to me and she said well We need to do something with that. We know right now. That's well. She said we know what the users are doing So let's do some change Because she was already tired of you know being there 15 years and nothing changed and she she went She saw like this sparkling that something might change and make something Usable and beautiful and you know something that will work actually for the user So we pushed a change below the rudder and that means that we didn't go through this whole rated process where We had to have like it, you know design sign off of from four managers or three managers and going through all these kind of Calls and and meet things which would be another 50 hours or maybe more and we delivered this this thing and Of course, there was like a loads of blame on our hat and a lot of problems And because we yeah, we didn't really go by the rules but I think that was the best thing I did in in in the bank so far and Three weeks after I published a paper with all the learnings and findings Which was which I will go through You don't have to read all the stuff there. It's just like a kind of like a template if you like to take the slides afterwards You can see how it was structured So basically I created like this what was before so the initial version and then the reasons why the search why this research has been created or you know why I started this research So we can see as well the the old version This is actually the new version and you can see that There I defined like the goals and the goals were to increase the visibility of the search books to make it more prominent that would fix the That it's not being queues that that often make users type in more than just one word to get more relevant search results and Make users to type even sentences and questions and this was achieved like you can see this input type which before there was just search and afterwards was there like a ask a question or enter a search term here and There was quite interesting feedback from few people that they were Because this change took us about I think two hours or so and Then people were asking a question Oh, it must have been really hard to make something that will talk our language and will understand our questions right and The only thing we did is that we asked the user to ask actually questions because all the solutions or most of the Solutions always have something like how do I do something? So giving the user the option because they didn't realize that they can type in a question they can type in like how do I change something which gives more keywords better results so that improved pilot with just a keyword with just a copywriting there and We also changed the search button to be more prominent and So I describe all these all these changes in a document I also put in like functionality development time effort and I wanted to show them that it's really possible to do something in just Like these micro researches or micro actions to to just push something a little bit farther and Release more often than just one one month release cycles Then of course follows the learning from the search box And then again, you need to know like what are your goals? So search from home page it increased quite a lot There is the blue one is like the previous week the week before the change The week after and second week after you can see that the second week after is already quite above the lines I don't even understand right now. What was the formula because I see strange person there probably I got into this success-driven metrics and So this is the search from home page blank keywords only we thought that it will decrease the number but There wasn't a big change in that behavior. So it needs improving and How do I search is that improved a little bit? So people were actually searching more the questions afterwards And there as well one of the goals was this single keywords and multiple keywords because users users before were typing Let's say if they had a problem with outlook they would type in just outlook that wouldn't give them any results So or really broad results. So we are looking to make their users type in multiple keywords and that improved a little bit but again we can do better and Part of that was also the proposal to for a new version of the box where we learn from these statistics we learn from what can be improved and we put in some ideas and and Then propose some sort of like a life cycle of the search box of the next release cycles because we needed or we wanted to make it fit into this Design process that that was there this rigged one and we didn't want to go below the radar because they wouldn't allow us and Back then yeah, or still I don't care if they if they fire me, but the others of course they do so then This is the this is the life of the of the search box that we did and what was really interesting is it that I think I'll Get to that that right after we push this this change below the rudder everyone was freaking out like the managers because they they really wanted to make it like a you know like a Success story and they they came to me and they never asked whether it's it's failing or it's is there a success They just wanted to get the blue numbers and that's kind of interesting That's why the innovation can't happen there because everyone expects you that you will be always successful and Anyways, the the version 1.2 didn't saw the light or didn't see the light Which is a little bit sad but I'm working on everyone because my contract is ending soon and The the conclusion is that you shouldn't be afraid pushing things forward And for example like in the bank it's something different and if you have people working for you know a company that works there like 15 years They have their security there But it was nice to see that these people as well kind of liked like they got into the you know new things and they they try to Try to innovate or try to push ideas. Maybe a little too much Because they thought that we already have some sort of statistics loads of statistics But we only had these two charts and they thought we have like a we know what the users are doing and it's really just what I showed you it's just like tip of the iceberg and Where you have like the one name above and the rest is below so if you I would say you should start thinking now about your projects and what what can be improved there to to make the users much more happier and as well like to let's say drive the business and Think about what can be measured and define your goals Deploy it learn it and make a report and that means you or try to make it that you can fail So you can fail fast and fell often and from that you can learn. It's not good. Well if you fail It's it's fine But if you fell on something that you felt before that's not okay So you should always kind of iterate and never repeat your fails again. That's why you learn And I think that's that's it for me Any questions? Yes Okay, so the question was if this was if doing this research was of my full-time work during being Employee there it wasn't I was doing it in between of the build regular build work So I was spending let's say like I don't know one hour a day just you know analyzing the the logs and stuff But it took it took some time Yeah Any other questions? Yes, I have a question. Yep I'm working for a bit different structure and I'm working for Drupal shop We have clients who pay us money. We build a website and I really like this idea and the improvement that we can gain from this approach can be really huge but in This business when we have fixed budgets how we can approach this kind of clients When we have smaller projects, it's not like 15 years and it's not huge banks How we can put this into the process, maybe you have some experience on maybe smaller scale projects, right? Well, it depends on the What are you are going to gain from that? I mean it's easy to to measure that for like e-commerce site because you know with e-commerce site you always can measure your sales Even it's a small project So it depends I would say it depends on the goals what you can set and what kind of benefit it will give you Whether it's a worth it or not, but it's always good Yeah, you have to sell it as a like Doing their users research should be really part of the web designing or at least some sort of testing That's the problem. The thing is that we don't know and results at all It's not something that is predictable, especially when you launch the website, right? You cannot say that Okay, we'll launch this website, but in half a year. We're going to do a research for one month This will cost this amount of money and we will increase some number of visitors or some other metrics for this amount if we could say that that will be the way go forward, but We fail. I mean the aim of research and tries is to fail Sometimes and we can fail five times and on sixth attempt. We will be way better But we don't know and results before we started How we can go for them? Yep, I can That makes sense No, not really. Okay Any other questions? Okay, I guess not so thank you and if you have any other questions you can come directly to me Or if you have even questions about Prague, I'd be happy to answer that Thanks