<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Nikola Arnaudov on Medium]]></title>
        <description><![CDATA[Stories by Nikola Arnaudov on Medium]]></description>
        <link>https://medium.com/@nikolaarnaudov?source=rss-23752a9646f6------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*gLiXw-WmSw6mR3uY6cb4UA.png</url>
            <title>Stories by Nikola Arnaudov on Medium</title>
            <link>https://medium.com/@nikolaarnaudov?source=rss-23752a9646f6------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sat, 21 Nov 2026 22:39:55 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@nikolaarnaudov/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Research in a “don’t-do-reasearch” environment]]></title>
            <link>https://medium.com/design-bootcamp/research-in-a-dont-do-reasearch-environment-1a681f6faed9?source=rss-23752a9646f6------2</link>
            <guid isPermaLink="false">https://medium.com/p/1a681f6faed9</guid>
            <category><![CDATA[corporate-culture]]></category>
            <category><![CDATA[ux-research]]></category>
            <category><![CDATA[enterprise-ux]]></category>
            <category><![CDATA[tips-and-tricks]]></category>
            <dc:creator><![CDATA[Nikola Arnaudov]]></dc:creator>
            <pubDate>Fri, 21 Feb 2025 20:42:47 GMT</pubDate>
            <atom:updated>2025-02-21T20:42:47.671Z</atom:updated>
            <content:encoded><![CDATA[<p><em>Product design cannot exist without various data about your users. Here are some non-standard ways to get it.</em></p><figure><img alt="A UX designer sitting alone on an island, looking for user data" src="https://cdn-images-1.medium.com/max/800/1*fDX6bcpspcuHGhwg_fHLNQ.png" /></figure><p>Research should be a non-controversial topic. Especially in a billion dollar enterprise creating complex software in an even more complex domain (in my case: IT infrastructure, devops, SRE, self-service, and automation). Then suddenly, you have to deliver stuff in 2 weeks, every 2 weeks. And you get a process that has n levels of approvals. And no way to share future product updates without an NDA that gets approved by the US president. And zero-ish research budget. You might have a small dedicated research team, but you quickly discover they are swamped with requests from all over the company, and not familiar with the multiple domains they need to work on, meaning you won’t be getting any of the extra info you got when asking clarifying questions based on your product and competitors knowledge. At this point, you are f… inally ready to explore unconventional solutions.</p><figure><img alt="A knocked down boxer on a ring. His opponent is towering over him." src="https://cdn-images-1.medium.com/max/800/1*1VeMoZgsuS_1vdLbYuAT0g.png" /></figure><h3>Option 1: Roll with the punches &amp; ditch this whole research thing (turning into a glorified UI designer in the process)</h3><p>Design based on Nielsen’s 10 heuristics, memories from the HCI courses you did, and feedback from PMs who have talked to CTOs, who have talked to ITs, who have kind of explained what bugs them.</p><p>The pros of this approach are that you can deliver quickly, and if you keep in mind basic UX rules, your screens/system will be usable. Throw in a cognitive walkthrough with stakeholders to catch some problems that developers will know about (“our backend doesn’t support push alerts”) and get bonus points for stakeholder buy in. Cons: your pretty mockups have a high chance to not reflect any of the following: the user mental model, the actual task they want to achieve, the actual users who are doing the task, where your flow sits in the user journey. Also, you have 0 arguments supporting your design. You want auto update on status? “Let’s wait for user feedback on this”. Are we sure we can’t auto-populate these values in the UI? “Let’s wait for the next release if somebody actually needs it. I have a hunch people will mostly be using this feature via API”.</p><figure><img alt="A hermit meditating on top of a mountain surrounded by all kinds of information representations." src="https://cdn-images-1.medium.com/max/800/1*-5zApAUvRaDM5jq-Jl5IQw.png" /></figure><h3>Option 2: The answer has always been inside you</h3><p>If this was a movie, in your desperate search for direction, you would climb a high mountain peak and seek wisdom from a hermit who was doing human factors research during WW2. These would be his exact words: “You will now reap the fruit of your diligent research in the good years of your company. Dig inside your notes and recordings: you probably have some background info of use cases, JTBD, probably personas. If you were deeply involved, you know how your product works inside and out, and already have a list of pain points across the user journeys, as well as how they differs across major client types. Use this info to judge how new features might integrate painlessly. Now let me tell you about this booklet I wrote about knob shapes in airplanes in 1940…”</p><figure><img alt="2 kids taking an exam. One of the kids is looking at the test page of the other." src="https://cdn-images-1.medium.com/max/800/1*oAvrktCKldV5IJczT4-gZQ.png" /></figure><h3>Option 3: Hit the books.</h3><p>Remember the boring “desk research”? Well, suddenly you sign up for all your product’s direct competitors. Most probably you already know about them, as some client has asked at some point “Can we do it like GCP/Azure?” or your VP has said something along the lines of “Have a look at how simple AWS is”, or you showed n8n to your dev team and said “Hey, look at how pretty their canvas is.” If not sure, just check the Gartner quadrant in your field and quickly sign up for everything in the top right corner. Most probably these guys had time and budget for research and have the user journeys and flows that match the 90% chunk of customers.</p><p><strong>Protip 1</strong>: do your subscriptions one by one. The trials are often time-limited, so by the time you have explored product A, your untouched trial for product B might have expired</p><p><strong>Protip 2</strong>: know your use cases. Every UX-er with 1+ years of experience can handle basic information architecture and UI tasks, so even if it is tempting to spend a lot of time looking at ways to improve your forms layout, probably you can use your 14 days better. Instead: note down the whole user journey which probably maps to client feedback and processes they have access to (and you don’t). Focus on things like how they onboard and get the needed info, can they actually achieve the use case of the feature you are designing, what are the steps in that specific flow, is this actually a feature targeted at the same persona you are assuming, what part of the user journey that feature is.</p><p><strong>Protip 3</strong>: Watch tutorials: they can map 1:1 with how actual users will do stuff.</p><figure><img alt="A group of colleagues are piled on top of each other trying to complete a task together in a slightly exaggerated way." src="https://cdn-images-1.medium.com/max/800/1*zanYYDTbXBTULA1jS2n4Tw.png" /></figure><h3>Option 4: the power of friendship</h3><p>You know this trope in manga and anime where the protagonist is on the verge of defeat by the final enemy who’s had a zillion years of training and an army of minions but suddenly remembers about their friends and overcomes all? Well, that’s colleagues can do for you in a corporate environment. There are a few places where you can get actual and valuable test-subjects for anything from a quick feedback session to proper user testing.</p><p><strong>Technical account managers/ technical marketing teams</strong> — whatever they are called in your company, these are either the people that help clients implement your solutions or the ones that do demos for clients. They tend to know everything about the current generation of the product, how it integrates with the rest of your company’s ecosystem and other products used by clients, what competitors are doing. They really, really want to know if the next iteration of your product will break everything their assigned clients have built so far. Also, such teams are company employees, so definitely not under the “You need an NDA signed with blood” category. Once you survive the 3 pages of complaints and feature requests technical managers/ technical marketing people have kept in their back pocket for a couple of years, you can do anything: proper user testing with a group of experienced users, actually valuable cognitive walkthroughs where you get feedback based on multiple client needs in one go, and a bunch of happy colleagues who will be available to ping over slack/teams/gchat and get quick feedback on a specific change.</p><p><strong>Dev teams (replace with whateve team is using your product) inside your company — </strong>find out if you do some form of dogfooding: if your devops team is using your Infra-as-Code solution and configuration tool daily, call them. They will 100% have a backlog of features they want, bugs and features that… bug them. And they can definitely help with questions about the holistic workflow of your product and how it integrates in their tasks. Real life example: I was part of a team building a new way to self-service and manage all types of infra based on K8s concepts. We (dev, PM, UX) were all ready to swear that if we follow kubernetes paradigms our flows would require zero learning from anyone who has worked in the devops space. However, when we demoed to a guy from our CI/CD team, he had multiple questions about the way the VMware concepts and K8s concepts work together and it turned out our “very obvious” infra nesting was anything but. I noted every place where I had to explain our model and together with the team turned them into a first time experience leading user through a simple setup till they deploy a VM in a similar way to kubevirt. Once we added it as first step in the product, not only did the infra model questions never come up again, but some users were actually reusing the startup experience as a shortcut to deploy their day-to-day stuff.</p><p><strong>The QA people</strong> — you will never find anyone more motivated to give you feedback, as making sure the company ships a good product is their literal job and talking to the UX team is one more lever to get all these QA jiras fixed rather than rejected. QA folks come in 2 flavors: manual testers who try to actually perform what your product is supposed to do based on a scenario, and the ones who build automated tests doing pretty much the same thing. While they are mostly trying to make sure your buttons and fields work as intended, they go through your flow repeatedly several times a day and due to the nature of their work tend to hit all kinds of edge cases noone ever thought of. Also, once someone has done a process 20 times, they start to find ways to optimise or wonder why something is not optimised — anything from “why things I do first are at the bottom of the navigation”, to “why don’t we autofill this, it’s part of the API response in step X”. In many organisations, QA team is not focused on a single feature but whole products, so they can actually help you with identifying user journey and integraion problems, too.</p><p>While QAs often work on a very detailed fixed scenario, which rules them out as full observation targets, having at least one such session is still worth it — you can catch UI kinks that you didn’t think through — someone having to jump up and down in the UI, changes in other parts of the product that affect your slice of it, even accessibility problems like tab order.</p><figure><img alt="A small town market selling all types of hardware and software." src="https://cdn-images-1.medium.com/max/800/1*8CkyrxM0tk2tY48X-mp1_Q.png" /></figure><h3>Option 5: Events</h3><p>If your research budget was cut, so was your travel budget (probably). But if you have a company/client event on your continent, you might just hop on whatever low cost company works near you and do some semblance of field research.</p><p><strong>PM/client sessions.</strong> These could give a treasure trove of info even if you just sit on them. Your PMs will be privately demoing products for big customers, and their teams will be asking questions. “Hey, we want to do X” — now you have a use case; “How can I do Y?” — if Y is there, you have some usability feedback, “We have separate teams for Z and Z1, do we have proper roles to manage access?” — that’s some nice persona info. Do a few of these, and you can actually do your work based on actual client insights.</p><p><strong>Demo sessions of your product </strong>— if the company is letting visitors try your products at the company booth, you could probably watch over their shoulder noticing if they get stuck, and listen to questions they ask to company representatives to get similar info to the private sessions, but in a noisier environment.</p><p><strong>Creative solutions</strong> — we once managed to get a tiny table at one of our user group conferences, put a laptop on it with the product feature we wanted to test, and one of us (one of the friendliest designers I have worked with, and also the guy in charge of the whole product design team) wrote on a piece of paper: “We are the Orchestrator UX team. Come shout at us” and walked around the hall to recruit participants.</p><h3>Instead of conclusion</h3><p>Mix and match any of these to apply to your organisation. And remember that research gives you power: “5 of our major clients couldn’t do this” has more weight in meetings than “this flow is not intuitive/this is cluttered”. But it also gives you a reality check: “this is cluttered” is not how a 777 pilot views their cockpit.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1a681f6faed9" width="1" height="1" alt=""><hr><p><a href="https://medium.com/design-bootcamp/research-in-a-dont-do-reasearch-environment-1a681f6faed9">Research in a “don’t-do-reasearch” environment</a> was originally published in <a href="https://medium.com/design-bootcamp">Bootcamp</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[UX patents in enterprise IT]]></title>
            <link>https://medium.com/design-bootcamp/ux-patents-in-enterprise-it-1d4a1e03fca1?source=rss-23752a9646f6------2</link>
            <guid isPermaLink="false">https://medium.com/p/1d4a1e03fca1</guid>
            <category><![CDATA[ux]]></category>
            <category><![CDATA[enterprise-software]]></category>
            <category><![CDATA[product-design]]></category>
            <category><![CDATA[patents]]></category>
            <dc:creator><![CDATA[Nikola Arnaudov]]></dc:creator>
            <pubDate>Wed, 08 May 2024 09:59:33 GMT</pubDate>
            <atom:updated>2024-05-10T15:25:11.320Z</atom:updated>
            <content:encoded><![CDATA[<h4>A.k.a. How to patent a flowchart</h4><figure><img alt="Abstract flowchart image" src="https://cdn-images-1.medium.com/max/1024/1*Zgg9HknY8CMz18CQBIUwDQ.png" /></figure><p>If you are a product designer and you’ve ever tried to patent something in a large company (that is not Google or Apple), you would quickly notice that your employer is not very keen on paying a law company and USPTO tens of thousands of dollars to protect your great and very scalable tree pattern. The reason — this is not their core business, and probably won’t affect whatever competition they have. However, your employer will happily patent a meaningful experience innovation that has the potential to improve the workflow of your (and your competitors’) customers. In this case, your patent is giving them an edge and is worth every penny. To get there, you will need 3 things: good understanding of the actual user tasks beyond interactions; good understanding of your product’s capabilities; and a cross functional team to work through them and come to a great and viable solution. Viable in this case means something that will work and can be implemented at a reasonable cost.</p><p>This might sound abstract, but luckily five years after it was filed, my first patent out of 5 filed for VMware (now Broadcom) is finally approved by USPTO and I am free to walk you through it as an example. And just as any useful innovation, it solves a real world problem, and goes beyond a simple pattern in the UI. It is the fruit of cooperation between dev, PM, and UX.</p><h4>TL;DR</h4><p>For all of you who don’t like the food blogger format:</p><ol><li>Base your patent idea on real customer need to get support from your company. Probably pick a problem that gets mentioned often in user groups. It could even be found somewhere in your product’s JIRA.</li><li>Work with your X-functional team. You need a developer to help you with both implementation and implementation details for the patent filing. You need a PM to get all the use cases (you talk to customers sometime, they talk to customers all the time).</li><li>Get buy in: working with a team means iterations — maybe you will make some sacrifices, maybe your counterparts will, but you need to get out of this with a single common solution. After that, you can be sure that your partners will give all they have to implement it (thank the sunken cost bias later:))</li><li>Make sure you patent a solution, not a pattern (unless you work on the next iOS, you lucky you).</li><li>Go through whatever flow your company has for internal patent approval — you can usually get helpful tips from your developer colleagues who already have patents filed.</li><li>Be ready for multiple meetings with external legal company. Probably they specialise in tech patents, but still you will need to do a lot of explaining about the use case &amp; differentiators of your solution.</li></ol><h4>The setting</h4><p>Back when I was joining VMware, we had just decommissioned a large piece of software (vRealize Automation 7) and released Automation 8.0 — prettier, more capable, and more flexible version of our product, taking advantage of tags and capabilities to make sure your workloads were landing on the right infrastructure without human intervention beyond the initial setup. Users could say “I want my RabbitMQ cluster to run in Sofia” and depending on who you are, we would automatically find the right compute resources from all available, we’ll make sure that the networking you need is there, we would find a storage supporting whatever your machines needed, setup your security, and present you with the working thing neatly integrated in the company IT infra.</p><h4>The problem(s)</h4><figure><img alt="Panicked clients in front of laptops" src="https://cdn-images-1.medium.com/max/1024/1*H2PqePC9Sd0yCjCHFSc0Vg.png" /></figure><p>As we released the product , it became apparent that our high-tech customers were struggling.</p><p>Problem one: No user had visibility into the system process: all configuration elements (compute, storage, networking, policies) were detached from each other for maximum flexibility and Automation mixed and matched them until it found a possible solution.</p><p>Problem two: Deployed resources landed on all sorts of places or failed, even when our clients were sure that they had all the right infrastructure capabilities and had placed all the right tags on their configuration files.</p><p>Problem three: There was no way to troubleshoot beyond the error you got. Even if “We couldn’t find an Ubuntu image” was the specific problem encountered, our users might have Ubuntu somewhere, just not on the datacenter where we found the right type of storage, for example.</p><h4>The horror</h4><p>You can imagine our dedicated support slack channel (a great way to get customer feedback and identify pain points very quickly) was flooded with questions why something failed, why something was provisioned where it wasn’t supposed to, how our new logic worked. Log bundles were tossed left and right, engineers were crying, the world as we knew it was about to collapse (OK, the last 2 were an exaggeration, but the words “not moving to production” were heard, which is pretty close to a calamity).</p><h4>Finding the right problem and solution</h4><p>I gave all this background to drive home a single point: if you want your company to back up your patent, you need to root it into an actual consumer problem. Just as importantly, it should also be the right problem, because in the enterprise IT world it’s often the case that a problem is actually a feature and you need to dig further for the root cause of all the client frustration. This is why we actually reached out to the customers that complained and talked.</p><p>It turned out the obvious UX solution — coupling different configurations more tightly and setting them up together was the wrong solution to a wrong problem. For our clients, being able to create separate sets of configurations and match them at run time meant more reusability, and ultimately less configurations. It also meant we matched our clients’ organisation — often one team dealt with storage, one with networks, etc.</p><p>Showing our clients how our system works and how the items are used together in the context of their own environment started to look like a better idea. Add an easy way to fix misconfigurations from this same context and you have a patentable winner.</p><h4>Flowcharts — the heart of every IT organisation and our solution of choice</h4><figure><img alt="Happy office workers around a flowchart" src="https://cdn-images-1.medium.com/max/1024/1*sInhxryCKuy8IW0QWDIcZA.png" /></figure><p>Have you ever wondered why every project kicks off with a bunch of flowcharts? They are easy to understand without specific knowledge — you can be a developer, an architect, a PM, a product designer, an IT admin — you will always understand what a flowchart means. They provide at-a-glance overview of compelx processes. They can integrate objects, and not just actions. You can quickly modify them. You can do comparisons between related processes — flow for the backend, flow for the front end, user journey… Flowchart is the language all our clients speak and we would speak it, too. Jakob Nielsen would have been proud how literally we follow the “Match between system and real world” heuristic.</p><p>We would build a flowchart that matches how our provisioning engine worked, which would explore all the available infra, provide info why something was chosen, and why not and we will link out to all configurations or resources, so the user would click on a problematic box and be able to do the fixes right away with no context change.</p><h4>Go alone to go quick, go together to go far</h4><p>We had come to the solution as a team. The engineering people were on board, the PM was on board, our leads were on board. This meant many restrictions that designers face doing work on an already released product were removed. We could rework APIs and our developers were coming up with solutions for backwards compatibility. We could reach out to other teams for assistance.</p><p>We built a comprehensive list of use cases that should be addressed by the provisioning diagram (~20).</p><p>We became subject matter experts on how Automation worked to make sure we listed all cases and our diagram reflected the process.</p><p>You can see a lot of “we” here, and there is a reason for this — a useful innovation (patented or not) is created by a team. A full understanding of the business impact, the possible cause and solutions, the actual client problem being solved, and shared knowledge of the problem and ownership of the solution basically remove any friction on the way to delivery.</p><h4>The actual feature</h4><figure><img alt="Flowchart in an enterprise IT product" src="https://cdn-images-1.medium.com/max/1024/1*Xj3op4YuxKHMLi4kwxM6Uw.png" /></figure><p>You can see the patent filing <a href="https://patents.justia.com/patent/11514073">here</a>, and you will notice it is pretty technical. However it is a plain flowchart that goes through the same steps as our provisioning engine and is linking out to all the places where these steps are configured.</p><h4>Success Metrics</h4><p>A successful feature, patented or not, is bound to make your customers’ lives easier. But you build your authority inside your team if this success can be easily measured. Remember the slack channel with customer feedback? It literally went quiet overnight once we released. No questions “What is this view?” “How do we use it?” Only requests to expand it to additional areas of VMware’s offerings.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1d4a1e03fca1" width="1" height="1" alt=""><hr><p><a href="https://medium.com/design-bootcamp/ux-patents-in-enterprise-it-1d4a1e03fca1">UX patents in enterprise IT</a> was originally published in <a href="https://medium.com/design-bootcamp">Bootcamp</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>