<?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[The Design Loop - Medium]]></title>
        <description><![CDATA[Experience design and research articles from KINTO Join Ltd UX team reflecting on the lessons learnt from the work we do building sustainability focused web based software and applications as part of the Toyota group. The infinity symbol is a shorthand for the design process. - Medium]]></description>
        <link>https://medium.com/the-design-loop?source=rss----135073075061---4</link>
        <image>
            <url>https://cdn-images-1.medium.com/proxy/1*TGH72Nnw24QL3iV9IOm4VA.png</url>
            <title>The Design Loop - Medium</title>
            <link>https://medium.com/the-design-loop?source=rss----135073075061---4</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 21:28:01 GMT</lastBuildDate>
        <atom:link href="https://medium.com/feed/the-design-loop" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Tai Chi and the art of volume control]]></title>
            <link>https://medium.com/the-design-loop/tai-chi-and-the-art-of-volume-control-b4549c76d16a?source=rss----135073075061---4</link>
            <guid isPermaLink="false">https://medium.com/p/b4549c76d16a</guid>
            <category><![CDATA[ux-design]]></category>
            <category><![CDATA[sonos-app]]></category>
            <category><![CDATA[product-design]]></category>
            <dc:creator><![CDATA[Guthrie Watson]]></dc:creator>
            <pubDate>Fri, 28 Aug 2026 11:31:01 GMT</pubDate>
            <atom:updated>2026-08-28T11:31:01.326Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*LXOLd2BZWmFycWH48GbWzQ.png" /></figure><h3>Why Sonos’s new volume control is more significant than it looks</h3><p>In May 2024, Sonos released a ground-up redesign of its app, not just a visual refresh, a complete rebuild of the technical foundation, replacing the older UPnP and SOAP-based system with a new, more cloud-dependent architecture. It was pitched as faster and better, and timed to launch alongside its first headphones, the Sonos Ace.[1][2] What followed was one of the more serious missteps in recent consumer tech. The new app shipped without long-standing features like sleep timers and alarms, broke accessibility support for VoiceOver and screen reader users, and left some longtime customers unable to control speakers they’d owned for years.[3][1] The impact was severe: Sonos delayed two major hardware launches, reportedly including a next-generation soundbar and a long-rumoured set-top box, to redirect engineering effort towards fixing the app. Revenue fell, around 100 staff were laid off, and by January 2025 the fallout had cost CEO Patrick Spence his job, with board member Tom Conrad stepping in as interim chief executive.[4][5][6]</p><p>It’s against that backdrop, over a year of apologies, patch cycles, and slowly rebuilt trust, that Sonos’s latest app redesign is worth paying attention to this new redesign.</p><p>I’ve a number of Sonos speakers and I remember this app relaunch, including the fact that they had done it as a big bang release, meaning they couldn’t easily roll back the release because they’d changed the underlying technical architecture. For months it was barely usable.</p><p>Each morning I open the Sonos app to set up music for my Tai Chi practice, the app is a key part of the process after the morning alarms it needs to be reset. I turn down the volume and pick a playlist, it needs to be a quiet but audible as an accompaniment to my daily routine. It’s a small, habitual task, one I’ve done hundreds of times. So when I opened the app recently and found the volume control had changed, nothing else, just that one small interaction, it caught my attention. Importantly the rest of the app hadn’t changed — no deleted features or connectivity issues.</p><figure><img alt="The Sonos App icon and a selection of Sonos product shots showing the connection between the app and the physical speakers." src="https://cdn-images-1.medium.com/max/1024/1*XVesObJfNVVejDvb3bZfSw.png" /></figure><h3>Why the app is the product</h3><p>Sonos speakers have almost no meaningful physical controls. No mute button. Nothing beyond basic on/off and volume on the hardware itself and some slightly tricky multi-room options (which involve long presses and don’t always work). Everything else, playing music, switching audio sources, setting alarms, configuring voice control, happens through the app. This isn’t a case of a “smart” add-on wrapped around a product that works fine without it, the way a smart water jug bolts an app onto something that was already a jug. With Sonos, the app is the control surface. If the app doesn’t work, the speakers effectively don’t either, and the hardware investment goes with it.</p><p>For a long time that’s the tension that defined the product. The hardware itself was always reviewed well: strong sound quality, easy home cinema setup, once configured they all worked well, including a robust multi-room experience that allowed you chain speakers seamlessly. Sonos understood the app was the only way to set up and really control their system, and that for many users it was something used daily. The problem it turned out was underinvestment, not neglect: they knew the app mattered but didn’t put in the work to match it, then rolled out sweeping changes without testing them properly. That gap barely touched people with one or two speakers, or anyone leaning on Bluetooth rather than the app. It hit hardest for the people running multi-room systems or home cinema setups, switching between speakers and music sources daily, exactly the use case where there’s no way around opening the app.</p><figure><img alt="The compact app control component showing volume and mute controls." src="https://cdn-images-1.medium.com/max/1024/1*M4jFNboQtKgWY9xgSoiq1A.png" /><figcaption>The compact app control component showing volume and mute controls.</figcaption></figure><h3>A different kind of rollout</h3><p>Sonos began rolling out the latest update on 14 July 2026.[7] The headline change is tabbed navigation at the bottom of the screen, replacing the stacked screens and gestures introduced in 2024 with a familiar Home, System and Search layout. Alongside it came the ability to sort and pin speakers and rooms, and a rebuilt volume interface on iOS, with Android to follow.[8][9] In the months before launch, interim CEO Tom Conrad had flagged all three as priorities: clearer navigation, a genuinely new volume interface, and more control over how players are listed and displayed.</p><p>What stands out isn’t only what changed, but how it arrived. The update was rolled out gradually, in phases, over roughly two weeks, rather than landing on every device at once. The new navigation and volume interface are opt-in, something you switch on deliberately rather than something imposed overnight. Set that against 2024, when the redesigned app arrived all at once, mandatory, barely tested, and stripped of features people relied on daily. This year’s approach clearly reflects lessons learned.</p><h3>The detail that stood out: a volume slider worth noticing</h3><p>Of everything in the update, it was the volume control that made me stop and pay attention, not because it’s the biggest change on paper, but because of how it felt to use. Tapping the slider now shifts the interface into a focused volume-control state. The slider itself is thicker, with plus and minus buttons at either end for small, precise adjustments. Drag it and the phone gives a faint haptic pulse with each increment, like turning a real dial rather than smearing a finger across glass. A numeric display confirms the current level as you go, read-only, feedback rather than input. And tucked into the same interaction is a mute toggle, a capability the hardware doesn’t have. Previously the only option was to turn everything off, or slide the volume down and then back up again once the interruption had passed. Now a brief interruption, someone at the door or a phone call, is just one tap away.</p><p>None of it required heavy engineering. It’s a handful of modest, well-considered choices that, together, turn a routine task into something that feels considered. It’s a small, useful illustration of what the rest of the redesign is reaching for: interface decisions made with real daily use in mind, rather than a coat of visual polish.</p><figure><img alt="Thicker slider, plus and minus controls, numbered, and mute affordance." src="https://cdn-images-1.medium.com/max/1024/1*Ojs9TxHT39RRF8OlfKaDQA.png" /><figcaption>Thicker slider, plus and minus controls, numbered, and mute affordance.</figcaption></figure><h3>The bigger shift this signals</h3><p>For years, Sonos used the app as a shop window for Sonos Radio and its various premium audio and subscription features, pushed prominently into a space where people mostly just wanted to control what they were listening to. That’s a crowded market where Sonos has no real advantage over Spotify or Apple Music, and competing there always felt like a distraction from what the company actually does well.</p><p>This redesign, and the careful, phased, opt-in way it’s being delivered, reads as a quiet admission of that. It’s a pivot toward Sonos’s genuine value: being the best possible control layer for whatever music service you already use, rather than trying to be a streaming service itself. Nothing about tabbed navigation or a rebuilt volume slider needed a large engineering investment. It needed clear-eyed product thinking about what people actually touch every day, a decision to invest there instead of chasing streaming ambitions that were unlikely to pay off, and the discipline to ship it properly this time rather than rushing it out the door.</p><p>There’s a familiar shape to this kind of recovery. When Steve Jobs returned to Apple in 1997, the company was spread thin across a sprawling product line, chasing every market it could justify rather than doing a small number of things well. His fix wasn’t a bigger roadmap, it was a smaller one: cut the product line down to a handful of machines, and put the focus back on the everyday experience of using them. The turnaround came from restraint and clarity, not from new ambition. Sonos, in a much smaller way, looks to be making the same kind of move: less concerned with expanding into what it isn’t, more focused on doing the core thing, letting people control their music, exceptionally well.</p><p>If Sonos brings that same care to the rest of the app, the case for doing so is hard to argue with: low-cost, high-impact, good for the business, and good for the people who just want their speakers to do what they’re told, reliably, every time.</p><p>Tomorrow morning I’ll open the app again, set the volume for Tai Chi, quiet enough not to dominate the room, and probably won’t think about it at all. That’s the point. The best interface is the one you stop noticing, and for the first time in a long while, Sonos seems to understand that the small, unnoticed moments are exactly where the trust gets rebuilt.</p><h3>References</h3><ol><li>Faster to use, but missing many features, the new Sonos app is a work in progress. Digital Trends, May 2024. <a href="https://www.digitaltrends.com/home-theater/sonos-app-redesign-2024-work-in-progress/">digitaltrends.com/home-theater/sonos-app-redesign-2024-work-in-progress</a></li><li>Pennell, A. What happened to the Sonos app? A technical analysis. LinkedIn, July 2024. <a href="https://www.linkedin.com/pulse/what-happened-sonos-app-technical-analysis-andy-pennell-wigwc">linkedin.com/pulse/what-happened-sonos-app-technical-analysis-andy-pennell-wigwc</a></li><li>2024: The year Sonos slipped. TechRadar, December 2024. <a href="https://www.techradar.com/televisions/2024-the-year-sonos-slipped">techradar.com/televisions/2024-the-year-sonos-slipped</a></li><li>Sonos Announces CEO Transition. Sonos Newsroom, 13 January 2025. <a href="https://newsroom.sonos.com/254650-sonos-announces-ceo-transition">newsroom.sonos.com/254650-sonos-announces-ceo-transition</a></li><li>Fitzgerald, D. Sonos CEO Patrick Spence is leaving following bungled app update. TechCrunch, 13 January 2025. <a href="https://techcrunch.com/2025/01/13/sonos-ceo-patrick-spence-is-leaving-following-app-update-disaster/">techcrunch.com/2025/01/13/sonos-ceo-patrick-spence-is-leaving-following-app-update-disaster</a></li><li>Sonos CEO leaving after botched app revamp led to customer revolt. Fortune, 13 January 2025. <a href="https://fortune.com/2025/01/13/sonos-ceo-patrick-spence-out-tom-conrad-in-botched-app-revamp-customer-revolt/">fortune.com/2025/01/13/sonos-ceo-patrick-spence-out-tom-conrad-in-botched-app-revamp-customer-revolt</a></li><li>14th July 2026, New Sonos App and Player Updates Now Available. Sonos Community, 14 July 2026. <a href="https://en.community.sonos.com/product-updates/14th-july-2026-new-sonos-app-player-updates-now-available-6934332">en.community.sonos.com/product-updates/14th-july-2026-new-sonos-app-player-updates-now-available-6934332</a></li><li>Sonos Tries to Repair Its App With Tabs, Better Volume Controls and Speaker Sorting. <a href="http://ecoustics.com/">ecoustics.com</a>, July 2026. <a href="https://www.ecoustics.com/news/sonos-app-update-2026/">ecoustics.com/news/sonos-app-update-2026</a></li><li>‘I’ve been waiting for this for ages’: The Sonos app’s new navigation, volume controls and speaker ordering are finally rolling out. TechRadar, July 2026. <a href="https://www.techradar.com/audio/ive-been-waiting-for-this-for-ages-the-sonos-apps-new-navigation-volume-controls-and-speaker-ordering-are-finally-rolling-out-though-you-might-not-get-them-right-away">techradar.com/audio/ive-been-waiting-for-this-for-ages-the-sonos-apps-new-navigation-volume-controls-and-speaker-ordering-are-finally-rolling-out-though-you-might-not-get-them-right-away</a></li></ol><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b4549c76d16a" width="1" height="1" alt=""><hr><p><a href="https://medium.com/the-design-loop/tai-chi-and-the-art-of-volume-control-b4549c76d16a">Tai Chi and the art of volume control</a> was originally published in <a href="https://medium.com/the-design-loop">The Design Loop</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Measuring Feature Success Without Losing Sight of the User]]></title>
            <link>https://medium.com/the-design-loop/measuring-feature-success-without-losing-sight-of-the-user-fb22440f5b0a?source=rss----135073075061---4</link>
            <guid isPermaLink="false">https://medium.com/p/fb22440f5b0a</guid>
            <category><![CDATA[product-analytics]]></category>
            <category><![CDATA[ux]]></category>
            <category><![CDATA[data]]></category>
            <category><![CDATA[ux-research]]></category>
            <category><![CDATA[carpooling]]></category>
            <dc:creator><![CDATA[Sima Ercevic]]></dc:creator>
            <pubDate>Fri, 31 Jul 2026 11:16:01 GMT</pubDate>
            <atom:updated>2026-07-31T11:16:01.317Z</atom:updated>
            <content:encoded><![CDATA[<p><em>When a feature ships, the real question isn’t whether we built it. It’s whether it’s quietly working for the people we built it for, and whether we’d even notice if it wasn’t.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*TeYjjzz9WpF_kGFn5a16aQ.png" /></figure><p>There’s a moment every product team knows well.</p><p>A feature ships. The demo goes smoothly. Everyone claps, and everyone moves on to the next thing on the roadmap. And then, weeks later, someone asks a deceptively simple question in a standup:</p><p><em>“Is anyone actually using it?”</em></p><p>More often than we’d like to admit, the honest answer is <em>we’re not sure.</em></p><p>We can usually say how many people opened the screen. We can pull a chart of taps. But those numbers rarely tell us the thing we actually care about: whether the feature is solving the problem it was meant to solve, for the person it was meant to help. Without that, we’re quietly designing the next iteration on instinct rather than evidence.</p><p>This is a story about closing that gap. Not with a heavyweight analytics programme or a dedicated data team, but with something surprisingly lightweight. And it’s about a realisation I didn’t expect: that measuring a feature’s success, done well, brings you <em>closer</em> to your users rather than reducing them to a row in a spreadsheet.</p><p>I’ll use a real example throughout: the <strong>Booked Carpool</strong> feature in <strong>KINTO Join</strong>, our mobile app for sustainable commuting. It lets a driver plan a carpool in advance and then, on the day of the trip, record the journey so it counts. Simple enough, but exactly the kind of feature where “did people use it?” and “did people <em>actually</em> use it the way we intended?” turn out to be very different questions.</p><h3>Start From the Need, Not the Feature</h3><p>The easiest numbers to collect are almost always the least useful. Page views. Total taps. Session length. The UX community has a well-earned name for these: <em>vanity metrics.</em> The simplest test is whether you ever actually act on them. App downloads always go up. Pageviews always climb. Neither tells you if a single person got what they came for.</p><p>The critique isn’t new. In his 2023 paper <em>“</em><a href="https://uxdesign.cc/so-how-can-we-measure-ux-93fad8a8ddc9"><em>So, How Can We Measure UX?</em></a><em>”</em>, Maximilian Speicher argued that none of the metrics readily available from standard analytics can reliably measure experience on their own. They don’t tell you where someone got confused, what they were trying to do, or why they gave up.</p><p>So before measuring anything, there’s a better question to sit with: <em>what did this feature promise the user?</em></p><p>Not what it does. What it <em>promises</em>. A booking feature doesn’t promise “a form.” It promises <em>certainty</em>: I’ve secured what I need, and I don’t have to worry about it. A journey-recording feature doesn’t promise “a button.” It promises <em>effortless proof</em>, where my contribution is captured without me having to think about it.</p><p>Frame it that way and the thing to measure becomes obvious. You’re no longer asking “how many people tapped the button?” You’re asking “did the person get the certainty we promised them?”</p><p>This is also, quietly, where user needs and business needs stop being in tension. A reservation that’s made but never used doesn’t serve the user, because they didn’t get what they came for. And it doesn’t serve the business either, because the value the feature was meant to create never materialised. When you measure whether the user’s need was genuinely met, you’re very often measuring the business outcome at the same time. They turn out to be the same question, asked from two directions.</p><h3>Three Lenses on the Same Experience</h3><p>Here’s the mental model that made all of this click. Good feature measurement isn’t one tool or one number. It’s three lenses pointed at the same human experience, each answering a different question.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*6va0cT0VKHHRXWJOoyktDg.png" /></figure><p>A <strong>funnel</strong> shows you <em>where</em> people fall away in a flow. A <strong>session replay</strong> shows you <em>why</em> they hesitated or gave up. A <strong>heatmap</strong> shows you <em>what</em> drew their attention, and what they missed entirely.</p><p>None of these on its own is enough. A funnel without a replay is a number without a reason. A replay without a funnel is a single anecdote with no sense of scale. Layered together, they stop being analytics and start becoming <em>observation</em>.</p><p>The tool I used for all three was <a href="https://uxcam.com/">UXCam</a>, a mobile-first experience analytics platform that keeps funnels, session replays, and heatmaps in one place, and lets you jump straight from a drop-off point in a funnel into the recordings of the exact people who dropped there. What made it click for me wasn’t any single feature but that connection: the numbers and the human behaviour behind them live one click apart.</p><h3>The Funnel: Where People Fall Away</h3><p>A funnel is the sequence of steps a person moves through to get what they came for, and it shows you where they fall away. Not <em>that</em> people drop off, but <em>where</em>. That distinction is everything, because “the feature isn’t working” and “step three isn’t working” lead to completely different responses. One sends you back to the drawing board. The other sends you to fix one screen.</p><p>For Booked Carpool I set up a simple three-step funnel: <em>created the reservation → opened it on the departure day → started recording the journey.</em> Here’s one month of real drivers.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*SXsaU1sSZUKQZ3Oz2aS7iA.png" /></figure><p>Read as a single completion number, around 64% of drivers who created a reservation went on to record the journey. It’s easy to shrug at. But the funnel showed something I wouldn’t have guessed from that one figure.</p><p>Almost everyone who <em>opened</em> the booking on the day went on to record: about 91% of them. The final step worked. The real loss was happening <em>earlier</em>. Nearly a third of drivers never came back to the app on the day at all.</p><p>There’s a second detail in that funnel worth pausing on: the average time to conversion was over sixteen hours. This flow doesn’t happen in one sitting. Someone books on Monday evening and records on Tuesday morning, which means the measurement has to span sessions and days to see the whole thing. Get that setup wrong and the feature looks broken when it isn’t.</p><p>That drop-off is a completely different problem than the one I’d have assumed from a low completion number. It’s not a broken screen. It might be a missing reminder, a notification that never lands, or friction in getting back into the app. The funnel didn’t just hand me a percentage. It pointed me at the <em>right question</em>, and at a specific group of people to go and understand.</p><p>This lines up with a principle running through most serious writing on UX measurement, from <a href="https://www.heartframework.com/">Google’s HEART framework</a> onwards: measure the <em>critical</em> flows, not everything. A small number of key moments explain most of what you need to know. Chasing every possible metric buries the signal under noise.</p><h3>The Session Replay: Understanding the Why</h3><p>Here’s where measurement stops being cold arithmetic and starts becoming empathy at a distance.</p><p>A funnel tells you <em>where</em> someone dropped off. Session replays let you ask <em>why</em>. You can watch the actual session of a real person moving through the flow. Not their identity, nothing invasive, just the interaction. Where they hesitated. What they tapped. Where they went instead.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*hFsCGI3ITnnMkDtpgTeFMA.png" /></figure><p>Look at the timeline on the right. This driver spent fourteen seconds on the bookings list, opened the ride details, swiped up, tapped the start button, and the detail event fired <em>again</em> before the journey finally started at twenty-five seconds. That small stutter, a swipe and a repeat, is invisible in any funnel. It’s someone checking themselves, making sure they’ve got the right trip before committing.</p><p>A percentage on a chart is abstract. Watching someone hesitate, swipe, pause, and check again is not abstract at all. It’s a person interacting with something we made, in a way we didn’t fully anticipate. You can’t un-see it, and you can’t argue with it in a review meeting the way you can argue with a number.</p><p>This is what “friction-first” thinking is really about. Some of the strongest signals about an experience aren’t in surveys. They’re in behaviour. Repeated taps. Backtracking. Rage taps. Abandoning halfway. These are things people simply <em>wouldn’t do</em> if the product were doing its job perfectly, and exactly the things a survey would never catch, because the person might not consciously remember the friction at all.</p><h3>The Heatmap: What Gets Noticed, and What Doesn’t</h3><p>The third lens answers a question the other two can’t: on a single screen, what is actually pulling people’s attention?</p><p>A heatmap aggregates where people tap, scroll, and linger, and paints it over the screen itself. I pulled up the one for the bookings screen, where a driver lands when they open the app on their departure day.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*WTFf9p0jyhrUe66Me0fNkA.png" /></figure><p>Two things jumped out. The first is the average visit duration: barely a few seconds. People are not browsing this screen, they’re glancing at it and acting. The second is where the heat actually lands. It clusters on the booking cards themselves rather than spreading evenly, which tells me people arrive knowing which trip they want and go straight for it.</p><p>The reassuring part is what <em>isn’t</em> there. Rage taps appear in under 1% of visits, and swipes make up only a small share of gestures, which is normal scanning behaviour rather than frustrated hunting. Combined with the funnel result, this reinforces the same conclusion from a different angle: the screen isn’t the problem. Once a driver is here, the experience works. The loss happens before they ever open the app.</p><p>Put the three lenses together and you have a complete picture. The funnel tells you <em>where</em> people are lost. The heatmap tells you <em>what</em> they’re drawn to when they arrive. The replay lets you watch it happen and understand <em>why</em>.</p><h3>Why This Belongs on Every Feature</h3><p>Setting this up was remarkably lightweight: a small handful of key moments marked in the flow, being the point of intent, the point of action, and the point of completion. Once those are defined, the funnel essentially builds itself, and I could go from “I wonder if this works” to “here’s exactly where it doesn’t” in the same afternoon.</p><p>That speed changes the relationship between design and evidence. When answers are slow and expensive, you ration your questions and fall back on assumption. When they’re fast and cheap, curiosity becomes routine.</p><p>Which is why I’d like to see measurement stop being a special project reserved for flagship launches, and become a quiet habit attached to nearly everything we ship. If defining a few key moments is cheap, the question stops being “is this feature important enough to measure?” It becomes “why would we ship anything <em>without</em> knowing whether it worked?”</p><p>Metrics are also most useful tracked <em>over time</em>. Your own trajectory, such as a flow climbing from 61% to 74% over two releases, tells you far more about your product’s health than any external benchmark. But that only works if measuring is something you do consistently, feature after feature.</p><p>One last principle worth borrowing: Google’s HEART framework deliberately pairs <em>task success</em> with <em>happiness</em>. Behaviour and feeling, side by side. The tools I’ve described are strongest on the “what they do” side. The most complete practices pair them with the occasional lightweight check on how people <em>feel</em>, whether that’s a short in-context survey, a usability session, or a conversation. The number tells you there’s friction. The person tells you what it costs them.</p><h3>Final Thoughts</h3><p>It would be easy to read all of this as an argument for more data. It isn’t.</p><p>It’s an argument for staying close to the people we design for <em>after</em> launch, precisely when it’s most tempting to look away and move on. Measurement, done thoughtfully, isn’t the opposite of empathy. It’s a way of <em>extending</em> it past the moment the feature goes live.</p><p>A funnel is a way of noticing that someone is struggling. A session recording is a way of watching what they’re struggling <em>with</em>. A heatmap is a way of seeing the screen through their eyes.</p><p>The numbers, on their own, don’t understand anyone. But used well, as a starting point rather than a conclusion, they point us back toward the person. They tell us where to look, then get out of the way and let us watch, and understand, and improve.</p><p>And when the user’s need is genuinely met, the business outcome tends to follow. Not because we chased the metric, but because we stayed close to the person the metric was really about.</p><h3>References and Further Reading</h3><p>Kerry Rodden, Hilary Hutchinson &amp; Xin Fu, <a href="https://research.google.com/pubs/archive/36299.pdf"><em>Measuring the User Experience on a Large Scale: User-Centered Metrics for Web Applications</em></a> — the original Google paper introducing the HEART framework and the Goals–Signals–Metrics process (ACM CHI, 2010).</p><p><a href="https://uxdesign.cc/so-how-can-we-measure-ux-93fad8a8ddc9">Maximilian Speicher, <em>So, How Can We Measure UX?</em></a> (2023) — on the limits of standard analytics metrics as proxies for experience.</p><p><a href="https://uxcam.com/">UXCam</a> — the mobile experience analytics platform used for the funnels, session replays, and heatmaps in this article.</p><p><em>If you’ve been measuring your features mostly through taps and pageviews, try this on your next release: pick the one moment that represents the user actually getting what they came for, and measure the path to it. Then go watch a few of the people who didn’t make it.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=fb22440f5b0a" width="1" height="1" alt=""><hr><p><a href="https://medium.com/the-design-loop/measuring-feature-success-without-losing-sight-of-the-user-fb22440f5b0a">Measuring Feature Success Without Losing Sight of the User</a> was originally published in <a href="https://medium.com/the-design-loop">The Design Loop</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[From new joiner to designing for a complex ESG product in one month]]></title>
            <link>https://medium.com/the-design-loop/from-new-joiner-to-designing-for-a-complex-esg-product-in-one-month-1bf2a51b3d0c?source=rss----135073075061---4</link>
            <guid isPermaLink="false">https://medium.com/p/1bf2a51b3d0c</guid>
            <category><![CDATA[ux-design]]></category>
            <category><![CDATA[product-design]]></category>
            <category><![CDATA[newjoiner]]></category>
            <category><![CDATA[esg-reporting]]></category>
            <category><![CDATA[ai-in-esg-reporting]]></category>
            <dc:creator><![CDATA[Vaggos Angelis]]></dc:creator>
            <pubDate>Tue, 07 Jul 2026 10:31:01 GMT</pubDate>
            <atom:updated>2026-07-07T10:31:01.214Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*cIZuZ4KrC0ko-0PgkNnHIQ.png" /><figcaption>Image showing a team collaborating around a futuristic ESG reporting surface.</figcaption></figure><p>How structured onboarding, product mapping, documentation, and collaboration helped me become productive quickly in a complex ESG reporting product. 📈</p><p>Three months ago, I joined <a href="https://www.kinto-join.co.uk/">KINTO Join</a> as a Senior UX/UI Designer, working on <a href="https://kintozero.com/">ZERO by KINTO</a>.</p><p>Like any new role, there was a lot to learn: new people, new processes, new terminology, new ways of working, and most importantly, a new product.</p><p>But this was not just any product.</p><p>ZERO by KINTO operates in the world of ESG reporting, a space that combines carbon emissions, reporting frameworks, business processes, data collection, governance, and multiple user roles.</p><blockquote><em>This is not a product you understand through casual clicking alone.<br>It needs structured exploration, domain context, and a clear understanding of how workflows connect.</em></blockquote><p><strong><em>How do you become useful quickly in a product that takes time to understand properly? </em></strong>🎯</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*NI3RKNvadmvrm-SDF0sL0Q.png" /><figcaption>Image showing a designer working between design tickets and a design tool.</figcaption></figure><p>For me, the answer was structure.</p><p>I did not try to learn everything at once. I focused on mapping the product, completing realistic user tasks, asking better questions, and learning from the people who already held the context.</p><h3>The complexity behind ESG</h3><p>B2B SaaS products are already challenging to onboard into because they often come with specialised users, business-specific workflows, and technical terminology. ESG reporting adds another layer.</p><p>At first, the goal sounds simple:</p><blockquote><em>Help organisations measure and report their carbon emissions.</em></blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*HOtmrvgIw3ZmukGGmvyIDw.png" /><figcaption>Image showing a complex carbon reporting dashboard connected to multiple ESG data sources.</figcaption></figure><p>But behind that simple goal sits a lot of complexity.</p><p>Carbon reporting involves different scopes, categories, methodologies, data requirements, calculation logic, and reporting responsibilities. A simple example is transport-related emissions.</p><p>Scope 1 Mobile Combustion and some Scope 3 categories can all involve vehicles, fuel, travel, or transport activity. To someone new to the space, they can sound almost the same, but they are not.</p><ul><li><strong>Scope 1 Mobile Combustion</strong> usually relates to company-controlled vehicles, such as company-owned, leased, or company-issued vehicles.</li><li><strong>Scope 3 transport-related categories</strong> may involve employee-owned vehicles, contractors, suppliers, or third-party transport providers.</li></ul><p>That difference matters because it can change the user journey, the data required, the evidence needed, and the calculation pathway.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*p-8SdFsK9vZkA1BoV3O-xg.png" /><figcaption>Image showing two similar dashboards with different reporting logic.</figcaption></figure><p>That is the design challenge.</p><p>You are not just designing screens. You are helping users make the right reporting decisions in a complex environment.</p><h3>AI adds another layer</h3><p>When I joined, ZERO by KINTO was actively building a carbon calculation tool. In some ways, this helped. The product goal was clear.</p><blockquote><em>Help organisations calculate, understand, and report their carbon emissions.</em></blockquote><p>But a clear goal does not always mean a simple product.</p><p>A carbon calculator still needs to support different user roles, different data inputs, different reporting paths, and different levels of user knowledge.</p><p>On top of that, the product is also exploring how AI and automation can reduce repetitive admin and make carbon reporting less manual.</p><p>For a designer, that means asking two questions at once:</p><ol><li>How does this process work today?</li><li>How could it become simpler tomorrow?</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*m1WDQveOz0EW-mQQd8Kpqg.png" /><figcaption>Image showing an AI-assisted workflow helping a user organise information across multiple tools.</figcaption></figure><h3>What helped me move fast</h3><p>The biggest reason I was able to start contributing within my first month was the onboarding.</p><p>Previous experience helped me recognise the type of complexity I was dealing with, but it was the onboarding process, documentation, product mapping, realistic user tasks, and collaboration that turned that foundation into contribution.</p><p>Good onboarding does not remove the complexity of a product. It reduces the friction around learning it.</p><p>I had access to documentation, product walkthroughs, internal knowledge, and colleagues who were willing to explain the context behind decisions.</p><p>That combination helped me move faster from:</p><blockquote><em>“I am learning the product.”</em></blockquote><p>to:</p><blockquote><em>“I can start contributing to it.”</em></blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3Fol63LiwFx7PBWppoJEEQ.png" /><figcaption>Screenshot showing design tickets assigned to me in Jira.</figcaption></figure><h3>Mapping the product</h3><p>One exercise helped me more than anything else.</p><blockquote>→ Mapping the product journeys end-to-end.</blockquote><p>I explored the platform with structure, taking screenshots of key pages, workflows, and states, then placing them into a large visual canvas.</p><p>The goal was not just to click around. ❌</p><p>The goal was to build a mental model. ✅</p><p>I wanted to understand:</p><ol><li>how different areas of the product connect</li><li>where information flows</li><li>which workflows depend on each other</li><li>where complexity starts to appear</li></ol><p>By turning the product into a visual map, I could understand it faster and refer back to it when new design tasks came in.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*0kW-PfRs7e8dc9M6e7ID5w.png" /><figcaption>Screenshot showing the end-to-end mapping exercise.</figcaption></figure><h3>Learning through real tasks</h3><p>After mapping the product, I wanted to go one level deeper. <a href="https://kintozero.com/">ZERO by KINTO</a> supports different user roles and permissions, so I wanted to understand the product through realistic user behaviour.</p><p>I asked two colleagues to give me real tasks:</p><ul><li>three from the Product Designer I work with</li><li>three from the Product Owner</li></ul><p>That changed the exercise. I was no longer asking:</p><blockquote><em>What does this screen do?</em></blockquote><p>I was asking:</p><blockquote><em>What is the user trying to achieve, and where might they struggle?</em></blockquote><p>This helped me understand real workflows faster and made the platform feel less like a collection of screens and more like a set of responsibilities, decisions, and user goals.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*t742eXeMxfVUvxG8fBc62g.png" /><figcaption>Screenshot showing the realistic scenarios exercise.</figcaption></figure><h3>Docs helped. People helped more.</h3><p>The design wiki and internal documentation were extremely useful. They gave me access to product knowledge, design decisions, terminology, and explanations I could return to whenever I needed them.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*k3svMBCZpCsPFZWh7R7-Jg.png" /><figcaption>Screenshot showing our onboarding wiki.</figcaption></figure><p>But in a complex domain like ESG reporting, documentation alone is not enough. Some knowledge lives in conversations.</p><p>Sometimes you need someone to explain:</p><ul><li>why a decision was made</li><li>how a customer thinks about a workflow</li><li>where the edge cases are</li><li>what the documentation does not fully capture</li></ul><p>That is where collaboration became the real shortcut. 🤝</p><p>Having access to people across Product, Design, Engineering, and ESG helped me connect the dots faster.</p><blockquote><em>Good onboarding is not just about giving someone information.<br>It is about giving them access to the people who hold the context.</em></blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*p0kgMP3feR_alAI1auXF-w.png" /><figcaption>Screenshot showing the ZERO product flows.</figcaption></figure><h3>What I’d tell other designers</h3><p>Picking up design tickets within one month was not the result of one thing. It was the result of strong onboarding, useful documentation, structured product exploration, realistic user tasks, and good collaboration.</p><p>For any Product Designer joining a complex B2B SaaS product, my advice would be simple:</p><ol><li>Don’t just read about the product.</li><li>Explore it with structure.</li><li>Map it visually.</li><li>Complete real user tasks.</li><li>Ask questions constantly.</li><li>Learn from the people behind it.</li></ol><p>Because the fastest way to understand a product is not only to learn how it works. It is to understand why it exists, who it serves, and who builds it.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1bf2a51b3d0c" width="1" height="1" alt=""><hr><p><a href="https://medium.com/the-design-loop/from-new-joiner-to-designing-for-a-complex-esg-product-in-one-month-1bf2a51b3d0c">From new joiner to designing for a complex ESG product in one month</a> was originally published in <a href="https://medium.com/the-design-loop">The Design Loop</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[What EV Parking application work taught me as a designer (and driver)]]></title>
            <link>https://medium.com/the-design-loop/what-ev-parking-application-work-taught-me-as-a-designer-and-driver-cbfdaedf7eb9?source=rss----135073075061---4</link>
            <guid isPermaLink="false">https://medium.com/p/cbfdaedf7eb9</guid>
            <category><![CDATA[mobile-apps]]></category>
            <category><![CDATA[mobile]]></category>
            <category><![CDATA[user-experience]]></category>
            <category><![CDATA[ev]]></category>
            <category><![CDATA[parking]]></category>
            <dc:creator><![CDATA[Renata Waskiewicz]]></dc:creator>
            <pubDate>Wed, 24 Jun 2026 10:31:01 GMT</pubDate>
            <atom:updated>2026-06-24T10:31:01.027Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*VV_2m4ETTOwx6zHF" /><figcaption>Photo by <a href="https://unsplash.com/@whykei?utm_source=medium&amp;utm_medium=referral">why kei</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><blockquote><em>Booking a parking space is rarely something people look forward to.</em></blockquote><p>I drive a hybrid car myself, though not a plug-in hybrid. It’s the kind with the battery built into the vehicle, so charging has never been part of my day-to-day driving experience. Before working on an EV parking product, I had very little understanding of what EV drivers actually go through, from opening an app and finding a charger to completing a charging session and driving away.</p><p>What I quickly realised was that EV parking introduces an entirely different layer of complexity.</p><p>Unlike traditional parking, drivers aren’t simply looking for somewhere to leave their vehicle. They’re also thinking about charging availability, connector compatibility, battery range, charging speed, pricing, and whether the charger they’re reserving will actually be available when they arrive.</p><p>While working on a parking management feature, I found myself thinking less about parking and more about confidence.</p><p>Because ultimately, that’s what users are looking for. Not a charger. Not a parking space. <strong>Confidence</strong> that everything will work as expected.</p><p>They want reassurance that the charger is compatible, the bay will be available, charging will start without issues, and their journey will go to plan.</p><p><strong>That makes EV parking a fascinating UX challenge.</strong></p><p>Users are often on the move, under time pressure, navigating unfamiliar locations, or making decisions with limited battery remaining. Sometimes they’re dealing with the worst-case scenario: arriving somewhere only to discover there are no available charging spaces.</p><p>In UX terms, it’s a high-friction, low-attention environment where uncertainty carries a much greater cost.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*N9RvH0cUT7wwWUGd" /><figcaption>Photo by <a href="https://unsplash.com/@chuttersnap?utm_source=medium&amp;utm_medium=referral">CHUTTERSNAP</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><h3><strong>EV Parking Solves Two Problems, Not One</strong></h3><p>Traditional parking products usually focus on a single goal: finding somewhere to park.</p><p>EV parking experiences need to solve two connected problems at the same time:</p><ul><li>Finding a place to park</li><li>Finding a place to charge</li></ul><p>The challenge is that both requirements must align perfectly. A parking space without a compatible charger is effectively useless. A charger without an available parking bay is equally frustrating.</p><p>While reviewing parking journeys, I noticed that many user concerns appeared long before the booking process even started.</p><p><strong>Questions such as:</strong></p><blockquote>Is the charger compatible with my vehicle?</blockquote><blockquote>Is it currently available?</blockquote><blockquote>How fast is it?</blockquote><blockquote>Can I reserve it in advance?</blockquote><blockquote>How long can I stay?</blockquote><blockquote>What will it cost?</blockquote><blockquote>What happens if I arrive and the charger is occupied?</blockquote><p>Each unanswered question adds uncertainty. Good UX doesn’t simply provide information. It reduces uncertainty at the moments when users need confidence most.</p><h3><strong>Understanding the User Mindset</strong></h3><p>One of the biggest lessons from working on parking management was recognising that users aren’t looking for feature-rich experiences.</p><p><em>Users looking for </em><strong><em>reassurance</em></strong><em>.</em></p><p>Most users want four things:</p><ul><li>Speed</li><li>Clarity</li><li>Trust</li><li>Confidence</li></ul><p>That changes how interfaces should behave.</p><p>Someone planning a journey tomorrow may be happy comparing multiple locations and charging options.</p><p>Someone arriving in a city centre with 10% battery remaining wants the fastest route to a reliable charger.</p><p>The same person can have completely different needs depending on where they are in their journey.</p><p>The lower the battery level, the less patience they have for exploration.</p><p><strong>This made me think carefully about context.</strong></p><p>Rather than presenting every available feature at all times, <em>the experience should adapt to what users are trying to achieve in that specific moment.</em></p><p>It’s easier said than done.</p><p>Many EV parking apps still struggle here. They often present charger specifications, tariffs, loyalty schemes, account settings, parking rules and promotional offers all at once.</p><p>The result is cognitive overload at precisely the moment users need simplicity.</p><p>Coming from years of driving petrol and diesel cars, I realised how many extra decisions EV drivers are expected to make. Refuelling isn’t necessarily effortless, but charging introduces several additional layers of decision-making that many non-EV drivers rarely consider.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*jGoBEUPuT3qPr_c-KWxCWw.png" /><figcaption>ev battery level dashboard</figcaption></figure><h3><strong>Battery Anxiety Changes Everything</strong></h3><p>One thing I hadn’t fully appreciated before working on EV parking was how much battery level influences behaviour.</p><p><strong>A user with 70% battery is exploring options.</strong></p><p><strong>A user with 10% battery is solving an urgent problem.</strong></p><p><strong>In those situations, priorities change dramatically.</strong></p><p><strong>Distance becomes more important than price.</strong></p><p><strong>Availability becomes more important than amenities.</strong></p><p><strong>Reliability becomes more important than choice.</strong></p><p><em>Designing for those different mindsets became one of the most interesting challenges of the project.</em></p><h3><strong>Why Map-First Discovery Matters</strong></h3><p>One pattern that consistently felt right was a map-first experience. Users naturally think spatially when searching for parking.</p><p>Their first question is rarely:</p><p><em>“Which parking spaces are available?”</em></p><p><strong>It’s usually:</strong></p><p><em>“What’s available near me right now?”</em></p><p><strong>Maps answer that question immediately.</strong></p><p><em>However</em>, maps can quickly become overwhelming if every charger type, price point, availability status and parking rule is displayed at once.</p><p>One idea that repeatedly surfaced during the design process was reducing manual effort wherever possible.</p><p>If vehicle information could be retrieved automatically through registration plate recognition or VIN lookup, we could remove lengthy forms and eliminate the need for users to manually select vehicle types and charger compatibility settings.</p><p>A principle I kept returning to was progressive disclosure.</p><blockquote><strong>At the search stage, users often need only:</strong></blockquote><blockquote>Availability</blockquote><blockquote>Distance</blockquote><blockquote>Charging speed</blockquote><blockquote>Price</blockquote><blockquote>Connector type</blockquote><p>Everything else can appear later once they’ve selected a location. Showing less information initially often helps people make decisions faster.</p><h3><strong>The Trust Problem: Real-Time Availability</strong></h3><p>If there was one area that felt particularly important, it was availability.</p><p>Users will tolerate a longer walk. They’ll tolerate slightly higher pricing. <em>What they won’t tolerate </em>is arriving at a charger that was shown as available but isn’t.</p><p>Inaccurate availability damages trust far beyond a single booking. It creates doubt about the entire platform. Because of this, I started viewing availability indicators as trust-building features rather than operational ones.</p><p>Good experiences communicate status clearly through:</p><ul><li><strong><em>Live occupancy indicators</em></strong></li><li><strong><em>Recent update timestamps</em></strong></li><li><strong><em>Reservation confirmations</em></strong></li><li><strong><em>Clear visual states</em></strong></li></ul><p>Sometimes it’s actually better to communicate uncertainty honestly.</p><p>“<strong><em>Available — updated 2 minutes ago</em></strong>” is often more trustworthy than simply showing “<em>Available</em>”.</p><p>Users can handle uncertainty. What they struggle with is inaccurate certainty.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/760/1*cBVktiJAPwxyNS6kJegc8w.png" /><figcaption><a href="https://unsplash.com/photos/feet-on-asphalt-between-two-directional-arrows-PXB7yEM5LVs">https://unsplash.com/photos/feet-on-asphalt-between-two-directional-arrows-PXB7yEM5LVs</a></figcaption></figure><h3><strong>The Last 100 Metres Problem</strong></h3><p>One insight that stood out to me was that booking success doesn’t necessarily guarantee arrival success.</p><p><strong>Drivers often reach their destination and still find themselves asking:</strong></p><ul><li><em>Which entrance should I use?</em></li><li><em>Which level is the charger on?</em></li><li><em>Is it behind a barrier?</em></li><li><em>Do I need to scan a QR code?</em></li><li><em>How do I identify my reserved bay?</em></li></ul><p>The booking may be complete, but the experience still feels broken.</p><p>In many cases, a photo, visual landmark or clear arrival instructions can be more valuable than another search filter.</p><h3><strong>Designing Filters for EV Drivers</strong></h3><p>One challenge unique to EV parking is that not all charging spaces are equally useful. A traditional driver can use almost any available parking space. An EV driver often can’t.</p><p>Drivers may encounter multiple connector types and charging speeds.</p><blockquote>Common connector types include:</blockquote><blockquote>Type 2</blockquote><blockquote>Type 1</blockquote><blockquote>CCS</blockquote><blockquote>CHAdeMO</blockquote><blockquote>Standard 3-pin charging</blockquote><p><strong>Charging speeds also vary significantly:</strong></p><ul><li><em>Slow charging (up to 3kW)</em></li><li><em>Fast charging (7kW–22kW)</em></li><li><em>Rapid and ultra-rapid charging (50kW–350kW)</em></li></ul><p>While reviewing the experience, I found myself questioning whether charger compatibility should even be treated as a filter.</p><p>If a charger isn’t compatible with a user’s vehicle, it isn’t really an option.</p><p>In many cases, compatibility feels closer to a primary requirement than an advanced filter hidden several taps away.</p><h3><strong>Booking Should Feel Effortless</strong></h3><p>One of the easiest ways to create friction is by making users work too hard before securing a booking.</p><p>The strongest booking journeys tend to be remarkably simple:</p><ol><li><strong><em>Select a location</em></strong></li><li><strong><em>Choose a time</em></strong></li><li><strong><em>Confirm</em></strong></li><li><strong><em>Get directions</em></strong></li></ol><p>Every additional step introduces risk.</p><p>Lengthy registration processes, unnecessary forms and excessive confirmation screens often create frustration, particularly when users are already travelling.</p><p><em>One principle I kept coming back to was simple:</em></p><p><strong>The best form is often no form at all.</strong></p><p>If information can be inferred automatically, it should be. In mobile experiences like this, momentum matters. The booking process should feel almost invisible.</p><h3><strong>Designing for Failure</strong></h3><p>Most booking journeys assume everything goes according to plan.</p><p>Reality rarely does.</p><p>Throughout the project, I found myself asking questions such as:</p><ul><li><strong><em>What if the charger is broken?</em></strong></li><li><strong><em>What if another vehicle occupies the bay?</em></strong></li><li><strong><em>What if charging stops unexpectedly?</em></strong></li><li><strong><em>What if the user arrives late?</em></strong></li><li><strong><em>What if their battery is lower than expected?</em></strong></li></ul><p>Great experiences aren’t only designed for success. <em>They’re designed for recovery.</em></p><p><strong>Users often remember how a product behaves when something goes wrong more than when everything works perfectly.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*iiNx3qVyMxdJcc9_" /><figcaption>Photo by <a href="https://unsplash.com/@michaelfousert?utm_source=medium&amp;utm_medium=referral">Michael Fousert</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><h3><strong>The Experience Doesn’t End After Reservation</strong></h3><p>One thing that surprised me was how much attention is typically given to search and booking, while the actual parking experience receives less focus.</p><p>Yet once a booking has been confirmed, user priorities change completely.</p><p>They’re no longer searching. They’re managing their journey.</p><p><strong>This raised an interesting design question:</strong></p><p><strong><em>Should the app continue behaving like a marketplace, or should it become a parking management tool?</em></strong></p><p>In many cases, the second approach felt more useful.</p><p><strong>After reserving a space, users often care most about:</strong></p><ul><li>Arrival instructions</li><li>Bay identification</li><li>Charging progress</li><li>Remaining parking time</li><li>Session costs</li><li>Extending a booking</li><li>Exit guidance</li></ul><p>The interface should evolve alongside the user’s goals.</p><h3><strong>Designing for One-Handed Use</strong></h3><p>EV parking apps are frequently used while walking, travelling or multitasking. That makes interaction design especially important.</p><p>Small touch targets and dense interfaces may look efficient in design files, but they often create friction in real-world situations.</p><p>The experiences that work best tend to prioritise:</p><ul><li>Large tap targets</li><li>Clear hierarchy</li><li>Strong contrast</li><li>Readable typography</li><li>Minimal text entry</li></ul><p>Users should be able to understand what they’re seeing within seconds.</p><h3><strong>Common UX Mistakes I Notice</strong></h3><p>Looking across the EV parking space, a few recurring patterns stand out.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*jNinCjSnQQDo8Kmx" /><figcaption>Photo by <a href="https://unsplash.com/@amadorloureiro?utm_source=medium&amp;utm_medium=referral">Amador Loureiro</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p><strong>The first is information overload.</strong></p><p>I’m ok with it but most users are not. Many products attempt to surface every technical detail simultaneously, resulting in cluttered interfaces filled with unfamiliar terminology and competing priorities.</p><p><strong>The second is inconsistent language.</strong></p><p>Users shouldn’t need to decipher the difference between:</p><ul><li><em>Charging session</em></li><li><em>Connector occupancy</em></li><li><em>Parking reservation</em></li><li><em>Charging allocation</em></li></ul><p>Clear language reduces cognitive effort.</p><p><strong>The third is hidden information.</strong></p><p>Unexpected fees, charging restrictions or parking limitations discovered late in the process quickly damage trust.</p><p>Transparency is almost always the better design choice.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*4gc__P19HbGTvAlVQ75_ow.jpeg" /><figcaption><a href="https://unsplash.com/photos/a-woman-is-pumping-gas-into-her-car-9166Worfn_Q">https://unsplash.com/photos/a-woman-is-pumping-gas-into-her-car-9166Worfn_Q</a><br><a href="http://twitter.com/zaptec">@zaptec</a></figcaption></figure><h3><strong>What a Strong EV Parking Journey Looks Like</strong></h3><p>The most effective experiences generally follow a simple sequence.</p><p><strong>1. Search Nearby</strong></p><p>Relevant options appear immediately based on location and charging needs.</p><p><strong>2. Compare Options</strong></p><p>Users can quickly evaluate availability, charging speed, distance and pricing.</p><p><strong>3. Reserve</strong></p><p>Booking takes only a few taps.</p><p><strong>4. Get Directions</strong></p><p>Navigation and arrival guidance feel seamless.</p><p><strong>5. Check In</strong></p><p>Parking and charging start with minimal effort.</p><p><strong>6. Pay and Exit</strong></p><p>Costs are transparent and payment feels frictionless.</p><p>At every stage, the experience should reduce uncertainty rather than introduce it.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*HomXAwKP17Ov_ImU" /><figcaption>Photo by <a href="https://unsplash.com/@gilleslambert?utm_source=medium&amp;utm_medium=referral">Gilles Lambert</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><h3><strong>Final Thoughts</strong></h3><p>Working on EV parking taught me that the real product isn’t parking, charging or reservations.</p><p>It’s confidence.</p><p>Every screen, notification and status update either reduces uncertainty or adds to it.</p><p>As designers, we often focus on flows, components and features.</p><p><em>Drivers rarely remember those things.</em></p><p>What they remember is whether the experience helped them reach their destination without stress. Looking back, the most valuable design principle wasn’t speed, efficiency or even convenience.</p><p>It was helping users feel certain about what happens next.</p><p>When an interface removes doubt at every step, the experience feels effortless. And that’s often what great UX is really doing behind the scenes.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=cbfdaedf7eb9" width="1" height="1" alt=""><hr><p><a href="https://medium.com/the-design-loop/what-ev-parking-application-work-taught-me-as-a-designer-and-driver-cbfdaedf7eb9">What EV Parking application work taught me as a designer (and driver)</a> was originally published in <a href="https://medium.com/the-design-loop">The Design Loop</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Copy Across Dimensions: Building a Collaborative UX Writing Practice]]></title>
            <link>https://medium.com/the-design-loop/copy-across-dimensions-building-a-collaborative-ux-writing-practice-1b0bbd6045a1?source=rss----135073075061---4</link>
            <guid isPermaLink="false">https://medium.com/p/1b0bbd6045a1</guid>
            <category><![CDATA[ux-design]]></category>
            <category><![CDATA[ux-microcopy]]></category>
            <category><![CDATA[product-design]]></category>
            <category><![CDATA[ux-copywriting]]></category>
            <category><![CDATA[team-collaboration]]></category>
            <dc:creator><![CDATA[Guthrie Watson]]></dc:creator>
            <pubDate>Mon, 01 Jun 2026 11:31:01 GMT</pubDate>
            <atom:updated>2026-06-01T11:31:01.220Z</atom:updated>
            <content:encoded><![CDATA[<p>UX copy writing sits at the intersection of information architecture, interaction design, and the cultural specifics of language. It has to work across all three at once. When I recognised that, it became clear that copy couldn’t live outside of product design. It had to be part of the workflow itself. So I made a deliberate choice. Copy should be embedded in the design process from the start, worked through collaboratively, bringing our collective expertise to bear on that part of the work in focused, dedicated sessions.</p><p>The idea was straightforward enough. The practice of doing it has shown us something more: that shared ownership of copy changes what becomes possible. This piece reflects on what we’ve learned from taking that approach.</p><h3>Copy as UX, not content</h3><p>Our products sit between sustainability and mobility, offering flexible solutions designed to help users and organisations adapt to a rapidly evolving regulatory and transport landscape. For organisations, they provide the means to measure and report sustainability performance in line with global regulatory requirements. Our work in this context is to create copy as user experience for products. Copy that helps people use them to meet their goals as employees or organisations. We don’t provide copy as content. We provide copy as UX.</p><h3>Creating UX copy without a product copy specialist</h3><figure><img alt="Diagram of UX copy writing, a circular collaborative exercise. Copy creation, review sessions to finalised copy for hand off to the engineering build team." src="https://cdn-images-1.medium.com/max/1024/1*deEQziK9JDPRoeCdScHQVA.png" /></figure><h3>Copy sessions and collaboration</h3><p>Copy sessions have become a weekly discussion. Often about whether copy or UX/UI needs to be addressed. Do you alter part of the UX or adjust the copy? Where does the copy fit in the context of the experience’s information hierarchy? Does the interaction design negate the need for additional explanatory copy? Lastly, we consider the actual words. How will they translate? Are they already in the glossary? Do they need testing?</p><p>Answering these questions together in weekly copy sessions has meant a better UX process and higher performing copy.</p><h3>UX copy, more than just a UI label</h3><p>UX copy is multidimensional. It exists within an experience’s information hierarchy (UI layout), works within a digital application’s interaction space, that non-linear time defined by an app’s task space, as so well described by Brenda Laurel in Computers as Theatre (1991). On top of this, it has to work in a localised translation domain. Idioms, grammar, and culture come into play in order for people “reading” the experience to understand it. Lastly, the copy must be relevant to the language of the service or domain of the business providing it.</p><h3>First dimension: Information architecture</h3><p>If you look at a poster for a film in a bus stop or train station, there is an order in which you typically read the information. Top of the poster is usually the largest name of the film, maybe with a subtitle or a phrase like “in space no one can hear you scream”. Then typically either the director or actors, depending on how well known they are.</p><p>The importance of the information is structured so that even at a glance you grasp the important details first. It’s not always determined by the size of the type. It can be any combination of colour, typographic treatment, style, or position within the bounds of the poster.</p><p>The same pattern appears in screen and web design. Here, button backgrounds use your most striking brand colour, and titles and navigation frame the stage of action. The design of a mobile screen directs the eye to focus on the task at hand.</p><p>Contextual elements, navigational or functional tools, help you complete the task and focus your attention. Which is where we start to move into the next dimension.</p><figure><img alt="Photo by Kevin Luke on Unsplash" src="https://cdn-images-1.medium.com/max/1024/1*FI6NgM4uqnNZdYyxloO2Zw.jpeg" /></figure><h3>Second dimension: Interactions and time</h3><p>Like scenes in a play, our experience of software moves from one action or context to another. Some of these scenes are connected and may or may not be linear. Some have a start, middle, and end; others follow a more controlled, reflexive mode of interaction.</p><p>Context is carried forward with us as we move from one step to the next. Text is a key part of this multidimensional experience. The words we use help to frame and direct us through the experiences we interact with.</p><p>This balance isn’t instinctive at first. Over time the mix becomes more considered, with graphic icons, text, and white space balancing out the designs. The balance itself is a learnt reflex.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*i209DTbe5tsTyHDUgun9zg.png" /></figure><h3>Third dimension: Localisation</h3><p>The next dimension is the translation and localisation space. This space is tightly bound to the business language, where a contextual layer is added from the base language’s UX copy.</p><p>It’s not enough to literally or even correctly translate idioms and phrasing from the base language or culture. We must translate the business’s linguistic context, the meaning of the copy as it relates to the service delivered via the digital application.</p><p>KINTO Join supports more than ten languages now. We maintain one meaning across all languages. The user completing a task in English, our base language, should encounter the same intent and conceptual clarity regardless of whether they’re presented German, Spanish, or any of the other languages we serve.</p><p>Our approach has evolved to handle that complexity. We’ve built glossaries and style guides that carry forward not just words but the business context those words sit within. That material becomes the starting point for AI translation tools, which give us a first pass. But the real work happens when that output goes to human translators working in market. They’re the ones who catch where cultural adaptation risks diverging from the original intent.</p><p>We’ve also learned to write the base copy with translation in mind, keeping it simple and direct enough that adaptation across languages doesn’t require rewriting the meaning. And where we’re uncertain, we test. If something feels like it might not travel well, we validate it across the languages we support before it ships.</p><p>It’s a more rigorous process than we started with, and it demands investment. But it’s the only way to ensure that localisation isn’t just translation. It’s the practice of preserving a single user experience across different linguistic and cultural contexts.</p><h3>How does copy work in a multidimensional interactive space?</h3><p>Where and when should I use words? When is a graphic pictogram or illustration needed? When space is limited, as on a mobile device, how does this change what and how copy is used? If I have more space on a larger handheld or desk-based experience, how does this affect the choice of words? What about other languages, cultures, or needs. How does the text contribute to the accessibility of the app we’ve designed? How does this experience work for those using screen readers or other modes of interaction?</p><p>The answer to all these is very much dependent on the experience you are designing for. One thing is certain: it’s rarely useful to simply add or remove words. Most often it’s about context and picking the right words to suit the situation.</p><p>The constraints themselves teach you something about what the copy needs to do. When you’re writing copy that needs to work for someone using a screen reader, you’re forced to think differently about how words carry meaning without visual context. That rigour, writing within the constraint rather than around it, tends to make the copy work better for everyone else too.</p><h3>How do these dimensions affect what we write?</h3><p>Copy in the context of a task sits within the flow of an interaction and the hierarchy of a screen. Action-based copy sits in the present while contextual copy like navigation, headings, descriptions, or user-generated input surrounds it.</p><p>Contextual labelling copy mitigates the need for lengthy descriptive copy. A lot can be understood about an action users can take based on the experience and the contextual copy and elements that support it.</p><p>This brings us to an approach when reviewing and editing UX copy. It’s also why words aren’t just words in our products but an integral part of the designed interactive experience.</p><h3>Simplify, shorten, or delete the copy in context</h3><p>Adding more copy is rarely the answer to fixing an issue with your design but it’s often the one that even experienced designers initially reach for. In trying to shoehorn more explanation into limited space, say on a mobile screen, we can end up making things less clear.</p><p>The following approach has worked well for us. We try to simplify, shorten, or delete, and use context to drive understanding supported by the copy we write.</p><h3>Simplify and shorten</h3><p>Why use “undo last edit” when “undo” works just as well? In more complex settings we may consider “undo size”, “undo colour” and so on. It needs to be clear what can be undone, what action is performed by the affordance we’ve provided in the software we’ve designed.</p><p>“Undo my last action” as a label for an action might seem better because we are fully explaining what the button will do. But does it provide more information? In the context of writing a document, what else would we undo? What value do the additional words add?</p><h3>Delete</h3><p>Sometimes removing text is the solution. The trash can has been the symbol for deleting something in software for decades. It’s both a stand-in for the literal action of throwing something away and the act of removing some element entirely.</p><p>Combined with a warning colour like red, at least in a west European context, it reinforces the idea of removal or destruction. Deleting words either partially or entirely often simplifies an experience. Being direct when space is limited, like in a mobile screen, can improve the comprehension of that experience.</p><p>If you have the possibility of undoing that action presented after the deletion then the experience becomes learnable. Since delete and undo are common patterns in many software experiences we can rely on that familiarity. These kinds of global patterns can be useful. But they raise questions worth sitting with: what words actually need to be in your glossary and style guide? When and where do you use those guides in your workflow? How do they become tools that work whether you’re prompting AI for a first pass, writing copy yourself, or handing context to translators? The guides don’t need to be elaborate, but they do need to exist and function, to be documents that actually shape the decisions you make, not just reference material that sits unused.</p><h3>UX copywriting for the user’s context, not design or developer tools</h3><p>We design and build software with databases, visual design, and code writing tools. One of the most common patterns is to write copy that relates to those tools rather than what people understand their needs and actions to be.</p><p>Vera, an HR system, uses the phrase “Insert record” instead of “add holiday” or “request annual leave”. It comes close to breaking the experience. After exploring the interface as a user you realise they must be referring to adding some leave into the system.</p><p>As a user I want to be assured that my intention, to request annual leave, is communicated with the system that records it. As a software engineer I need the database to insert that record and display it, but display it in a way that I understand in relation to the task I am trying to complete. As a user I think in terms of the task I’m trying to complete, not what an engineer needs to do to satisfy my request.</p><p>It is a surprisingly common pattern for those drafting copy for software. As authors of UX copy we need to switch modes.</p><p>Designers and engineers need to put themselves in the shoes of people using their tools. If I use my engineering or design language to explain the function of an affordance, how can users then understand what options they have to complete the task they started?</p><p>For UX copy we want to develop a common shared language that everyone uses and that has clearly defined meaning in the context we use it. It should relate to the purpose of the software we designed, not the tools we use to create it.</p><p>To be clear, it is rarely a designer’s or engineer’s fault that this happens. Ever shorter deadlines and release cycles drive the business to rush development without doing the essential discovery processes that would provide them and the team with the feedback needed to write better UX copy.</p><h3>Coda</h3><p>Our weekly copy sessions have become something different from what we anticipated. What started as collaborative reviews of labels and microcopy has evolved into design conversations where we’re thinking through the full dimensional complexity of how words work in context. We’ve learned to spot patterns across hierarchy, interaction, and language that we might have missed working in isolation. That shared language has made it possible to explore options we wouldn’t have considered before, and to move past surface-level fixes into understanding how copy and design are inseparable.</p><p>When copy is treated as an integral part of the design process rather than something that comes after, the whole practice shifts. The multidimensional nature of UX copy becomes visible to everyone, and with it, the possibility of moving through that complexity together.</p><h3>References</h3><p>Laurel, B. (1991). Computers as Theatre. Addison-Wesley.</p><h3>Links</h3><p><a href="https://www.oreilly.com/library/view/computers-as-theatre/9780133390889/">https://www.oreilly.com/library/view/computers-as-theatre/9780133390889/</a></p><p><a href="https://www.nngroup.com/articles/content-strategy-vs-ux-writing/#:~:text=UX%20writing%20involves%20creating%20clear,guides%20people%20through%20an%20experience">https://www.nngroup.com/articles/content-strategy-vs-ux-writing/#:~:text=UX%20writing%20involves%20creating%20clear,guides%20people%20through%20an%20experience</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1b0bbd6045a1" width="1" height="1" alt=""><hr><p><a href="https://medium.com/the-design-loop/copy-across-dimensions-building-a-collaborative-ux-writing-practice-1b0bbd6045a1">Copy Across Dimensions: Building a Collaborative UX Writing Practice</a> was originally published in <a href="https://medium.com/the-design-loop">The Design Loop</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Designing Tables for Large Datasets: Turning Clutter into Clarity]]></title>
            <link>https://medium.com/the-design-loop/designing-tables-for-large-datasets-turning-clutter-into-clarity-990c75528457?source=rss----135073075061---4</link>
            <guid isPermaLink="false">https://medium.com/p/990c75528457</guid>
            <category><![CDATA[data-density]]></category>
            <category><![CDATA[data]]></category>
            <category><![CDATA[ui-design]]></category>
            <category><![CDATA[design]]></category>
            <category><![CDATA[ux-design]]></category>
            <dc:creator><![CDATA[Polly Taylor]]></dc:creator>
            <pubDate>Mon, 30 Mar 2026 10:31:02 GMT</pubDate>
            <atom:updated>2026-03-30T10:31:01.902Z</atom:updated>
            <content:encoded><![CDATA[<h3>Turning Clutter into Clarity: Designing Tables for Large Datasets</h3><p>Tables are one of the most common ways to present structured data, and one of the hardest to get right.</p><p>A well-designed table can help users make sense of complex information in seconds. A poorly designed one can feel like a spreadsheet nightmare: cluttered, sluggish, and impossible to read. Designing tables for large datasets isn’t just about making them look good. It’s about performance, usability, and clarity.</p><p>Let’s explore the main challenges — and some design strategies that might help.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*JvmraO80yBw0wG63JRlCXg.png" /></figure><h3>1. The Problem of Scale</h3><p>A table that looks great with ten rows can collapse under the weight of ten thousand. Large tables introduce two big challenges: visual and technical.</p><ul><li>Visual: Too many rows and columns overwhelm the eye. When users have to scroll endlessly, they lose context and can’t spot patterns.</li><li>Technical: Large datasets can slow down browsers and interfaces. Every scroll, filter, or sort can feel frustrating if the table is laggy.</li></ul><h4>Design solutions</h4><ul><li>Pagination vs. infinite scroll: Break data into manageable chunks. Pagination helps users track their place and maintain context. Infinite scroll works when the goal is exploration, like browsing a product catalog, but not when users need precision.</li><li>Virtualised rendering: Use tools that only render what’s visible on-screen (like React Virtualised or AG Grid). This keeps large tables fast and responsive.</li><li>Progressive loading: Don’t load everything at once. Instead, bring in data gradually as users interact.</li></ul><h4>Questions to think about</h4><ul><li>What information might be most important in the context users are looking at this table?</li><li>What information is just ‘nice to have’?</li><li>Do we understand what tasks users are trying to complete?</li><li>As designers, we tend to shy away from large amounts of data, but is there a context when showing ‘more is more’?</li></ul><h3>2. Too Much Information, Too Little Context</h3><p>One of the most common design issues in large tables is information overload. When every column competes for attention, nothing stands out, and users can’t see what matters most.</p><h4>Design solutions</h4><ul><li>Prioritise columns: Not every column deserves equal space. Highlight key identifiers or metrics. Hide secondary data behind expandable rows or tooltips — but only if you’re confident users won’t need them constantly.</li><li>Use visual hierarchy: Typography, alignment, and subtle colour cues go a long way. Right-align numbers for easy comparison, left-align text for readability, and use soft shading to separate rows or emphasise headers.</li><li>Group and freeze columns: When you have dozens of columns, group related data under shared headers to create structure. Freeze key columns (like names or IDs) so users don’t lose orientation when scrolling.</li><li>Allow resizing: Give users control. Let them resize columns to fit their workflow and emphasise what’s most relevant.</li></ul><h4>Questions to think about</h4><ul><li>Could the UX copy be condensed, or reduced?</li><li>Could the columns be put into groups or sets?</li><li>Is the most important information hidden, or prominent?</li></ul><h3>3. Helping Users Find What Matters</h3><p>Even the most beautifully organised table fails if users can’t locate what they need. Navigation, filtering, and discoverability are just as important as presentation.</p><h4>Design solutions</h4><ul><li>Search and filtering: A clear, responsive search bar is essential. Filters should be intuitive, fast, and combinable; users shouldn’t have to reload or lose progress when refining their view.</li><li>Sorting: Let users sort by multiple columns (for example, by “Region” and then “Revenue”). Use clear indicators; arrows, highlights, tooltips, to show what’s happening when they sort.</li><li>Show/hide columns: Give users the power to tailor what they see. Persist their preferences so that their custom setup remains when they return.</li><li>Saved views: For power users, the ability to save and reuse table configurations (columns, filters, sorting) transforms a static dataset into a personalised workspace.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*U2rr7IZu8lhn_KDizoUtaQ.png" /></figure><h4>Questions to think about</h4><ul><li>What information is useful to have a ‘sort by’ function, and what isn’t?</li><li>How could we utilise a search functionality to bring users the best results?</li><li>Do we want to retain users’ viewing settings, if they navigate away from the page?</li></ul><p>Ultimately, tables should help users find answers faster, not force them to wrestle with the interface.</p><h3>4. Accessibility and Clarity</h3><p>Accessibility often gets overlooked in large-table design, but it’s critical. Good accessibility isn’t extra work, it’s part of making your data understandable for everyone. For users relying on screen readers or keyboard navigation, massive tables can be frustrating without proper structure.</p><h4>Design solutions</h4><ul><li>Use semantic markup: Always build with proper &lt;table&gt;, &lt;th&gt;, and &lt;td&gt; elements. Add captions and ARIA labels for clarity.</li><li>Support keyboard navigation: Users should be able to move through rows and columns using arrow keys and tabs.</li><li>Mind your colour and contrast: Don’t rely solely on colour to convey meaning, and maintain strong contrast ratios for legibility.</li></ul><h3>In Summary</h3><p>Designing tables for large datasets is a balancing act between information density and usability. The best tables don’t try to show everything at once. They help users see patterns, find insights, and take action, without getting lost in the noise.</p><h4>Here are the core principles to keep in mind:</h4><ul><li>Show only essential data at first.</li><li>Support fast, flexible interactions through filtering, sorting, and search.</li><li>Use hierarchy and grouping to guide the eye.</li><li>Design responsively and accessibly.</li></ul><p>Of course, all of these solutions might not be available to you, or even fit in with your particular design problem. We are, at the end of the day, constrained by the device the user is using, the context they are using it in, and by the resources and time available to us.</p><p>We shouldn’t feel that, just because we haven’t implemented a complex UI solution, that we have failed in designing a useful table. It is also true that if someone really wants to find out something, they will spend a little longer and try a little harder to divine information from a complex table.</p><p>Even just taking a few small steps in the right direction can make a big difference. The responsibility then for us as designers is to spend a little longer and try a little harder to make thoughtful table experiences, even if we don’t always quite hit the mark.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=990c75528457" width="1" height="1" alt=""><hr><p><a href="https://medium.com/the-design-loop/designing-tables-for-large-datasets-turning-clutter-into-clarity-990c75528457">Designing Tables for Large Datasets: Turning Clutter into Clarity</a> was originally published in <a href="https://medium.com/the-design-loop">The Design Loop</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Designing a Corporate Parking Dashboard with AI Tools]]></title>
            <link>https://medium.com/the-design-loop/designing-a-corporate-parking-dashboard-with-ai-tools-df2d5f4d2b5f?source=rss----135073075061---4</link>
            <guid isPermaLink="false">https://medium.com/p/df2d5f4d2b5f</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[ux-design]]></category>
            <category><![CDATA[user-experience]]></category>
            <category><![CDATA[design]]></category>
            <category><![CDATA[figma]]></category>
            <dc:creator><![CDATA[Renata Waskiewicz]]></dc:creator>
            <pubDate>Mon, 16 Mar 2026 11:31:01 GMT</pubDate>
            <atom:updated>2026-03-16T11:31:00.899Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/768/1*wnCGOuvZQM89S3EOl1pp2Q.png" /><figcaption>isometric parking illustration</figcaption></figure><h4><strong>How I Used Figma Make and AI Tools to Solve Overcrowding, Violations, Grace Period Chaos and Shift Swaps</strong></h4><p>By a product designer working under real operational pressure.</p><p>Every morning, the same complaints arrived.</p><blockquote>“There’s nowhere to park.”<br> “Executives are angry.”<br> “Employees are getting violations.”</blockquote><p>But the numbers didn’t add up.</p><p>The campus had <strong>850 parking spaces</strong>, yet people insisted parking was impossible. That contradiction is what led me to design an <strong>AI-powered corporate parking dashboard.</strong></p><h3><strong>Why We Started Building This</strong></h3><p>This problem didn’t appear in isolation.</p><p>Our team works on a <strong>suite of fleet and mobility tools</strong> that support companies managing vehicles, employees, and shared transportation infrastructure. Across multiple client conversations, one operational pain point kept appearing again and again:</p><p><strong>Parking management.</strong></p><p>Different companies described the problem differently:</p><blockquote>“Executives can’t find their reserved spaces.”</blockquote><blockquote>“Employees park in EV spots.”</blockquote><blockquote>“Visitors don’t know where to go.”</blockquote><blockquote>“Shift workers arrive before others leave.”</blockquote><p>But underneath those complaints was the same operational reality.</p><p><strong>Parking systems were being managed with static rules inside dynamic workplaces.</strong></p><p>Most campuses were relying on a mix of:</p><ul><li>static allocation lists</li><li>manual enforcement</li><li>spreadsheets</li><li>disconnected booking tools</li></ul><p>None of them accounted for <strong>real-time behavioural patterns</strong>. That’s where we saw an opportunity.</p><p>Instead of treating parking as a standalone tool, we decided to design <strong>an admin layer for parking management</strong> that could integrate into our broader mobility platform.</p><p>The goal was simple. <strong>Turn parking from a reactive enforcement problem into a real-time operational system.</strong></p><h3>Every complex dashboard starts with frustration.</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/768/1*QoLy8HSWgC0aIP1aGlN-bA.jpeg" /><figcaption>dashboard image generated by AI</figcaption></figure><p>In this case:</p><blockquote>“There are never enough spaces. Executives are frustrated. Employees park illegally. Visitors complain.”</blockquote><p>The campus had:</p><ul><li>1,200 employees</li><li>850 parking spaces</li><li>14% daily overflow</li><li>60+ weekly violations</li><li>reserved executive spaces</li><li>EV allocations</li><li>visitor booking</li><li>rotating shift workers</li></ul><p>At first glance, it looked like a simple capacity issue. It wasn’t. It was a <strong>dynamic allocation problem </strong>inside<strong> a static rule system. </strong>And two hidden variables made it worse; <strong>grace periods</strong> and <strong>shift swaps</strong>.</p><p>That’s where AI helped me think beyond surface metrics.</p><h3><strong>Stage 1 — The Hidden Complexity</strong></h3><p>When I started mapping the system, the KPIs looked straightforward:</p><ul><li>occupancy %</li><li>violations per zone</li><li>reserved utilisation</li><li>peak hours</li></ul><p>Then operations added:</p><blockquote>“We allow a 15-minute grace period before issuing violations.”</blockquote><p>And HR added:</p><blockquote>“Shift workers often swap shifts informally.”</blockquote><p>Now the system wasn’t just spatial. It was <strong>temporal and behavioural.</strong></p><h3><strong>The Grace Period Problem</strong></h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*nF3NqrPoMYJladih335syg.png" /><figcaption>electrical vehicle grace period</figcaption></figure><p>Grace periods are meant to reduce friction. But they introduce ambiguity.</p><p><strong><em>Example</em></strong></p><ul><li>Parking capacity: 300 (General Zone A)</li><li>Occupied: 300</li><li>18 vehicles technically overstaying</li><li>12 of them within the 15-minute grace window</li></ul><p>The dashboard shows:</p><ul><li>100% occupancy</li><li>18 violations pending</li></ul><p><strong>But in reality:</strong></p><ul><li>only 6 true violations</li></ul><p>Without clarity, leadership thinks enforcement is failing. The real issue it turns out was <strong>timing thresholds.</strong></p><h3><strong>Designing Grace Period Visibility</strong></h3><p>Instead of counting violations blindly, I introduced layered states.</p><p><strong>Parking Status Categories</strong></p><ul><li>✅ Valid</li><li>⏳ In Grace Window</li><li>⚠️ Confirmed Violation</li><li>🔄 Awaiting Validation (sensor uncertainty)</li></ul><p>This changed the map logic.</p><p>Instead of <strong>orange = violation</strong>, I split it into:</p><ul><li>light orange = grace</li><li>dark orange = confirmed violation</li></ul><p>This small design decision reduced <strong>perceived enforcement errors</strong>.</p><h3><strong>Prompt I Used for Grace Period Logic</strong></h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*OmadknwkboWBc4Sq426J7Q.png" /><figcaption>Figma Make prompt example</figcaption></figure><p>I’m designing a corporate parking dashboard with a 15-minute grace period before violations are confirmed.<br><strong>Help me:</strong></p><ol><li>Define how the grace period should be visualised</li><li>Avoid misleading violation metrics</li><li>Suggest executive-level summary logic</li></ol><p><strong>AI helped clarify:</strong></p><ul><li>separate <strong>“Violation Risk”</strong> from <strong>“Confirmed Violations”</strong></li><li>show grace volume as an <strong>operational pressure indicator</strong></li><li>add countdown timer logic in the space-level view</li></ul><p>That became a powerful behavioural signal. Instead of punishing users, we <strong>warned them</strong>.</p><h3><strong>The Shift Swap Chaos</strong></h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/799/1*gHWTJyyD0NSrGcDk6TsTSA.png" /></figure><p>This was even more subtle, the corporate campus had:</p><ul><li>Morning shift: 6:00–2:00</li><li>Mid shift: 2:00–10:00</li><li>Late shift: 10:00–6:00</li></ul><p>On paper, it was balanced. In reality employees swapped shifts informally.</p><p>So between <strong>2:00–2:30 PM</strong>:</p><ul><li>morning shift still parked</li><li>late shift already arriving</li></ul><p>Result was, <strong>temporary micro-overflow spikes.</strong></p><p>These were often misinterpreted as permanent capacity failures.</p><h3><strong>Data Snapshot During Shift Overlap</strong></h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*WIUg1fX7qVbVpM2qroojJw.png" /><figcaption>UI example made in Figma Make</figcaption></figure><p>At <strong>1:30 PM</strong>, it looked catastrophic. But it <strong>self-corrected. </strong>This insight reshaped the dashboard logic.</p><h3><strong>Parking demand during shift swaps</strong></h3><p>Instead of static occupancy charts, I introduced:</p><p><strong>1️⃣ Overlap Risk Indicator</strong></p><p>A timeline bar showing:</p><ul><li>scheduled shift changes</li><li>predicted overlap density</li><li>risk level (Low / Moderate / High)</li></ul><p><strong>2️⃣ Temporary Capacity Buffer Visualisation</strong></p><p>On the interactive map, during overlap windows:</p><ul><li>zones get a <strong>soft red pulse</strong></li><li>label: <strong>“Shift Overlap in Progress”</strong></li></ul><p>This reframes overflow as <strong>temporary</strong>.</p><p><strong>3️⃣ Grace + Shift Interaction Layer</strong></p><p>The biggest insight.</p><p><strong>Grace periods + shift overlap amplify perceived violations.</strong></p><p>AI helped me think this through.</p><h3><strong>Prompt Used for Scenario Simulation</strong></h3><pre>Simulate a parking system with:<br>a 15-minute grace period<br>3 overlapping shift windows<br>10% shift-swap unpredictability<br>What temporary anomalies appear in violation data?</pre><p>Key insights returned:</p><ul><li>artificial spike in grace-window occupancy</li><li>false positive violation trends</li><li>increased repeat offender appearance</li><li>executive complaints during overlap</li></ul><p>This allowed me to:</p><ul><li>add a <strong>“Shift Window Filter” toggle</strong></li><li>show violation trends <strong>excluding grace</strong></li><li>provide <strong>contextual messaging</strong></li></ul><h3><strong>Interactive Map Evolution</strong></h3><p>The final map included <strong>four layers</strong>.</p><p><strong>1️⃣ Live Occupancy</strong></p><p>Colour-coded by real capacity.</p><p><strong>2️⃣ Grace Countdown View</strong></p><p>Each at-risk vehicle showed:</p><ul><li>14:32 remaining</li><li>a subtle animated timer ring</li></ul><p><strong>3️⃣ Shift Density Overlay</strong></p><p>Soft gradient showing:</p><ul><li>arrival wave</li><li>departure wave</li></ul><p><strong>4️⃣ Violation Confirmation State</strong></p><p>Separate from grace. This transformed the map from a <strong>static visualisation into a living system.</strong></p><h3><strong>Leadership Layer — Avoiding Misinterpretation</strong></h3><p>Executives don’t want raw complexity.</p><p>They want <strong>clarity, </strong>so the summary included:</p><ul><li>confirmed violations (not grace)</li><li>active grace vehicles</li><li>shift overlap status</li><li>reclaimable reserved spaces</li><li>peak forecast</li></ul><p>Most importantly, a new KPI:</p><p><strong>True Capacity Stress Index</strong></p><p>The formula combined:</p><ul><li>occupancy</li><li>grace %</li><li>shift overlap multiplier</li><li>reserved inefficiency</li></ul><p>This prevented <strong>overreaction during temporary spikes.</strong></p><h3><strong>Using AI to Accelerate the MVP</strong></h3><p>This project also became a real experiment in <strong>AI-assisted product design under pressure. </strong>The timeline was tight. Leadership wanted a <strong>demo-ready concept in under a week.</strong></p><p>Instead of starting from blank canvases and long documentation cycles, I used AI tools — especially <strong>ChatGPT</strong> — to accelerate several parts of the process.</p><p>Not to design the product for me. But to <strong>stress-test the system logic.</strong></p><p>I used prompts to:</p><ul><li>simulate parking scenarios</li><li>identify edge cases</li><li>clarify metric definitions</li><li>structure dashboard logic</li><li>clean up ambiguous assumptions</li></ul><p>A typical workflow looked like this:</p><ol><li>Write rough operational assumptions</li><li>Use AI to challenge them</li><li>Refine the logic</li><li>Translate it directly into interface structures</li></ol><p>Example prompt I used early on:</p><blockquote>“Simulate a parking system with:</blockquote><blockquote>a 15-minute grace period</blockquote><blockquote>3 overlapping shift windows</blockquote><blockquote>10% shift-swap unpredictability</blockquote><blockquote>What temporary anomalies appear in violation data?”</blockquote><p>The responses helped surface patterns that were easy to overlook:</p><ul><li>temporary capacity spikes</li><li>false violation signals</li><li>misleading occupancy percentages</li><li>enforcement appearing inconsistent</li></ul><p>These insights informed the <strong>data model before I even opened Figma.</strong></p><h3><strong>AI + Figma Make for Rapid Prototyping</strong></h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Q6LTH2THJSKEQZqV3vfNqA.png" /><figcaption>chatGPT + Figma Make image</figcaption></figure><p>Once the logic was clearer, the next step was moving quickly from <strong>system thinking to visual structure. </strong>AI helped here too.</p><p>Instead of manually building every UI variation, I used <strong>prompt-driven design workflows</strong> to generate rapid layout explorations and component structures, which I then refined in <strong>Figma using Figma Make.</strong></p><p>Typical workflow:</p><p>1️⃣ Define system states in ChatGPT<br>2️⃣ Generate structured UI descriptions<br>3️⃣ Convert them into layout scaffolds<br>4️⃣ Push them into Figma for refinement<br>5️⃣ Iterate interaction logic and hierarchy</p><p>This approach accelerated:</p><ul><li>dashboard information architecture</li><li>map layer logic</li><li>status state definitions</li><li>edge-case handling</li><li>demo storytelling</li></ul><p>Instead of spending days exploring structure, I could <strong>test multiple system interpretations within hours.</strong></p><p>AI didn’t replace product thinking. But it significantly reduced the time between:</p><p><strong>idea → scenario test → interface structure → prototype</strong></p><p>And under real operational pressure, that speed matters.</p><h3><strong>Reflection</strong></h3><h4>Corporate parking isn’t just about spaces.</h4><p>It’s about <strong>time.</strong></p><ul><li>arrival waves</li><li>behavioural grace windows</li><li>informal shift swaps</li><li>human unpredictability</li></ul><p>If we ignore temporal complexity, we design <strong>misleading dashboards.</strong></p><p>AI helped me zoom out. It helped me design not just for <strong>space allocation</strong> — but for <strong>time-based system behaviour.</strong></p><p>And when <strong>40 employees arrive early for swapped shifts while 30 others are still in a grace period</strong>, you don’t need more asphalt.</p><p>You need <strong>smarter visibility. </strong>That’s what intelligent dashboard design — amplified by AI — made possible.</p><h3><strong>Final Thought</strong></h3><p>Corporate parking isn’t just a <strong>facilities problem.</strong></p><p>It’s a <strong>systems problem.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/799/1*cn_CyLfmHuQyNaUdCyET6w.png" /></figure><p>It sits at the intersection of:</p><ul><li>workforce behaviour</li><li>time-based movement</li><li>limited infrastructure</li><li>operational policy</li></ul><p>When those dynamics are invisible, frustration grows quickly. But when they’re visualised properly, something interesting happens:</p><p>The problem stops being <strong>“we need more parking.”</strong></p><p>It becomes <strong>“we need better visibility.”</strong></p><p>That’s the kind of operational clarity we’re working towards.</p><p>If you’re interested in how companies can manage parking, fleets, and employee mobility more intelligently, take a look at what we’re building at <a href="https://www.kintojoin.io/"><strong>Kinto Join</strong></a> — where we’re designing tools to make workplace transportation systems <strong>simpler, smarter, and more connected.</strong></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=df2d5f4d2b5f" width="1" height="1" alt=""><hr><p><a href="https://medium.com/the-design-loop/designing-a-corporate-parking-dashboard-with-ai-tools-df2d5f4d2b5f">Designing a Corporate Parking Dashboard with AI Tools</a> was originally published in <a href="https://medium.com/the-design-loop">The Design Loop</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Navigating Complexity: UX Research and Usability Testing of a Taxonomy-Based Reporting Tool]]></title>
            <link>https://medium.com/the-design-loop/navigating-complexity-ux-research-and-usability-testing-of-a-taxonomy-based-reporting-tool-8cd5c92ea98a?source=rss----135073075061---4</link>
            <guid isPermaLink="false">https://medium.com/p/8cd5c92ea98a</guid>
            <category><![CDATA[user-experience]]></category>
            <category><![CDATA[ux-research]]></category>
            <category><![CDATA[taxonomy]]></category>
            <category><![CDATA[usability-testing]]></category>
            <category><![CDATA[user-research]]></category>
            <dc:creator><![CDATA[Anupam Sharma]]></dc:creator>
            <pubDate>Fri, 13 Feb 2026 11:31:01 GMT</pubDate>
            <atom:updated>2026-04-02T09:05:33.176Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="Illustration of user centredness. A post it note with the word user is pinned to a wooden board with a collage of coloured post-its stuck to the wall representing potential feedback from a research activity." src="https://cdn-images-1.medium.com/max/1024/1*YWPyWydWsBi_Lg-dFh3PZg.jpeg" /><figcaption>AI Image generated from Adobe Firefly</figcaption></figure><h3>Context and Problem</h3><h3>Working Across Domains, One Constant Remains</h3><p>Over the years, my work in UX research and usability testing has spanned multiple industries. From pharmaceuticals and retail to automotive and, most recently, sustainability. Each domain brings its own constraints, regulations, and user expectations. Yet one thing remains consistent: the need to deeply understand users and validate solutions through testing.</p><h3>Why Research Still Matters</h3><p>Regardless of industry, meaningful design begins with listening.</p><ul><li>UX research helps move beyond assumptions</li><li>Usability testing validates whether a product truly supports user behaviour</li><li>Insights help align user needs with the product vision</li></ul><p>What I’ve learned is that successful products sit at the intersection of user needs and product vision. Research helps map this intersection, translating insights into clear requirements, aligning stakeholders, and ensuring that what we build is not only functional, but relevant and usable. Whether it’s a highly regulated pharmaceutical tool or a sustainability platform aimed at long-term impact, the process of understanding users and testing early and often remains the foundation of good UX.</p><h3>Why It Mattered</h3><h3>Internal Reviews Aren’t Enough</h3><p>While internal reviews with designers, product owners, and stakeholders are valuable, they have limitations. Internal teams are often influenced by:</p><ul><li>Familiarity with the product</li><li>Organisational context</li><li>Unconscious assumptions</li></ul><p>The real validation begins when real users interact with the product.</p><h3>What Usability Testing Reveals</h3><p>A structured usability testing approach allows researchers to systematically evaluate key usability parameters such as effectiveness, efficiency, and satisfaction. By combining quantitative metrics (like task success rates, time on task, and error frequency) with qualitative insights (such as user feedback, observations, and behavioural patterns), teams gain a well-rounded understanding of the user experience.</p><p>This balance of data not only strengthens design decisions but also helps communicate findings more effectively to stakeholders.</p><p>Ultimately, usability testing bridges the gap between design intent and real-world use. It ensures that products are not just well-designed in theory, but intuitive, usable, and meaningful in practice.</p><h3>Approach and Methodology</h3><h3>Designing a Focused Testing Strategy</h3><p>Our usability testing approach is intentionally structured and goal-driven.</p><p>We aim to:</p><ul><li>Identify key user roles</li><li>Map roles to core workflows</li><li>Define realistic, task-based scenarios</li><li>Evaluate results using consistent UX metrics</li></ul><h3>Testing a Taxonomy-Based Reporting Tool</h3><p>This study focused on <strong>ZERO by KINTO</strong> a platform built on the <a href="https://www.efrag.org/en/sustainability-reporting/esrs-workstreams"><strong>European Sustainability Reporting Standards</strong></a><strong> (ESRS)</strong> under the <a href="https://www.efrag.org/en/sustainability-reporting/about-sustainability-reporting"><strong>Corporate Sustainability Reporting Directive</strong></a><strong> (CSRD)</strong>.</p><figure><img alt="Screen shot of KINTO Zeros imported XBRL report template with a ready to fill out report." src="https://cdn-images-1.medium.com/max/1024/1*34tIo7GjMWQGyHjT8fCW8A.png" /><figcaption>Screen shot from ‘ZERO by KINTO’ showing an XBRL based reporting cycle template</figcaption></figure><p>Key characteristics of the product:</p><ul><li>Built for ESG reporting requirements</li><li>Structured to the <a href="https://www.xbrl.org/the-standard/what/key-concepts-in-xbrl/taxonomies/"><strong>XBRL taxonomy</strong></a></li><li>Designed for accuracy, consistency, and compliance</li></ul><p>This raised an important question: how does usability testing a taxonomy-based reporting tool differ from testing a tool built for more common, consumer-facing use cases? The answer lies in the nature of the product itself. This tool is grounded in the <strong>European Sustainability Reporting Standards (ESRS)</strong>, developed under the <strong>Corporate Sustainability Reporting Directive (CSRD)</strong>, which represents a significant shift in how organisations report on non-financial performance.</p><p>However, the goal of our testing was not to assess users’ knowledge of ESRS, CSRD, or ESG regulations. Instead, our focus was firmly on usability, measuring how effectively and efficiently users could complete tasks, and how satisfied they felt while using the tool.</p><p>Separating domain expertise from product usability allowed us to evaluate the experience on its own merits and ensure the tool supports users, rather than relying on their regulatory knowledge to compensate for design gaps.</p><h3>What We Did Differently</h3><figure><img alt="Illustration showing a light globe plugged in the wrong way round but lit up to suggest how thinking different can look." src="https://cdn-images-1.medium.com/max/1024/1*TB7LSTn_n2EBLx1OwSdd4A.png" /><figcaption>AI Image generated from Adobe Firefly</figcaption></figure><h3>Avoiding Cognitive Overload</h3><p>Early in scenario design, we realised that tasks closely tied to XBRL taxonomy could overwhelm participants — especially in an <strong>unmoderated testing environment</strong>, where:</p><ul><li>No facilitator is available for clarification and</li><li>First-time users rely entirely on the interface</li></ul><p>Our challenge was to ensure users could progress confidently without being blocked by unfamiliar regulatory language.</p><h3>Designing Neutral, Familiar Scenarios</h3><p>To address this, we redesigned our test scenarios to feel familiar and relatable.</p><p>Instead of ESRS-driven tasks, participants were asked to:</p><ul><li>Build a form for others to fill out</li><li>Configure fields for a specific reporting cycle</li><li>Enter and structured data using common form controls</li></ul><p>This approach allowed us to focus on what we truly wanted to evaluate: whether users could effectively create a form using the interface elements provided, such as text, numeric, and date input fields, text formatting options, drop-downs, radio buttons, and other controls.</p><p>Our assumption was that by keeping tasks generic and intuitive, users would concentrate on completing the task itself rather than struggling with unfamiliar ESG terminology. As a result, we were able to gather clearer insights into the tool’s usability without domain complexity skewing the findings.</p><h3>Key Insights</h3><p>The usability testing largely confirmed our initial assumptions and hypotheses about how users would engage with the platform. Participants naturally gravitated toward the interface itself, prioritising interaction and task completion over understanding ESRS-specific terminology.</p><p>This behaviour reinforced our belief that the UI should absorb most of the cognitive load, allowing users to work efficiently without needing deep domain knowledge.</p><h3>Discoverability Issues</h3><p>A key challenge was the discoverability of certain input methods.</p><ul><li>Some controls were harder to find than expected</li><li>Users hesitated or relied on trial and error</li><li>Placement and visual affordances needed improvement</li></ul><p>Another recurring expectation was around <strong>real-time feedback</strong>.</p><ul><li>Users assumed configuration changes would update instantly</li><li>Delayed or absent feedback caused uncertainty</li><li>This revealed gaps in the interaction model</li></ul><h3>Outcome / Impact</h3><p>The decision to keep tasks free from ESRS-specific terminology proved effective. All participants were able to successfully complete their assigned tasks, while also offering valuable feedback throughout the process. By shifting the focus away from domain-heavy language, users were able to concentrate on the interface and task execution rather than deciphering terminology.</p><h3>Where Friction Appeared</h3><p>More complex interactions still caused challenges, especially:</p><ul><li>Conditional logic (e.g., “If date of birth is before X, require attachment”)</li><li>Multi-step configurations</li></ul><p>These moments highlighted opportunities to improve:</p><ul><li>System logic clarity</li><li>Feedback mechanisms</li><li>Progressive guidance</li></ul><h3>Role of Step-by-Step Instructions</h3><p>Providing structured instructions helped by:</p><ul><li>Guiding users through complex tasks</li><li>Preventing deviation from core objectives</li><li>Producing more focused and actionable feedback</li></ul><h3>Learnings &amp; takeaways</h3><figure><img alt="Illustration: A person looking at an explosion of ideas represented by a wall of multicoloured post-it notes fanning out from the center to the edges of the image." src="https://cdn-images-1.medium.com/max/1024/1*zvS0Dsh8S3NxYk-86YUW5A.jpeg" /><figcaption>AI Image generated from Adobe Firefly</figcaption></figure><p>One of the key learnings from testing a complex taxonomy-driven platform was that limited resources do not have to limit the quality or usefulness of usability insights.</p><p>Constraints such as the absence of domain experts with deep knowledge of reporting structures, or the inability to run fully moderated usability sessions, can feel like significant blockers at the outset. However, our experience showed that these limitations do not prevent teams from gathering meaningful evidence about what is and isn’t working.</p><h3>Neutral Testing Unlocks Better Insights</h3><p>By removing domain dependency and minimising prior explanations, we were able to evaluate the product without influencing user behaviour or expectations. This neutral approach allowed participants to interact with the interface based solely on what they saw and understood in the moment.</p><p>As a result, we were able to:</p><ul><li><strong>Test the interface on its own merits<br></strong> Users relied entirely on the app’s visual cues, labels, and flow to complete tasks. This helped us assess whether the interface was intuitive, self-explanatory, and usable without external guidance, revealing gaps in clarity, navigation, and affordances.</li><li><strong>Observe natural user interpretation<br></strong> Participants interpreted features, icons, and actions based on their own mental models rather than domain knowledge. This surfaced authentic user assumptions, misunderstandings, and expectations, providing valuable insight into how first-time or non-expert users experience the product.</li></ul><p>Ultimately, the goal of usability testing is not to validate domain expertise, but to understand how real users make sense of a system. When tests are designed around this principle, even a diverse group of non-expert users can generate enough qualitative and behavioural data to meet core research objectives.</p><p>This approach not only makes usability testing more accessible under tight constraints, but also ensures that the product is understandable and usable by a broader audience.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8cd5c92ea98a" width="1" height="1" alt=""><hr><p><a href="https://medium.com/the-design-loop/navigating-complexity-ux-research-and-usability-testing-of-a-taxonomy-based-reporting-tool-8cd5c92ea98a">Navigating Complexity: UX Research and Usability Testing of a Taxonomy-Based Reporting Tool</a> was originally published in <a href="https://medium.com/the-design-loop">The Design Loop</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[From Compliance to Commitment: Our Journey Toward Accessibility by Default]]></title>
            <link>https://medium.com/the-design-loop/from-compliance-to-commitment-our-journey-toward-accessibility-by-default-c0abdc51388a?source=rss----135073075061---4</link>
            <guid isPermaLink="false">https://medium.com/p/c0abdc51388a</guid>
            <category><![CDATA[accessibility]]></category>
            <category><![CDATA[mobile-accessibility]]></category>
            <category><![CDATA[web-accessibility]]></category>
            <category><![CDATA[inclusive-design]]></category>
            <category><![CDATA[web-accessibility-a11y]]></category>
            <dc:creator><![CDATA[Polly Taylor]]></dc:creator>
            <pubDate>Thu, 22 Jan 2026 12:32:59 GMT</pubDate>
            <atom:updated>2026-01-22T12:32:59.537Z</atom:updated>
            <content:encoded><![CDATA[<p>Our journey toward becoming an <em>accessible by default</em> product company began with pragmatism.</p><p>We started our road towards accessibility from a sales perspective — we had requests for proposals (RFPs) requiring accessible products. What initially looked like a box to tick for commercial reasons quickly became something more meaningful. What started with a financial goal in mind soon turned into ‘let’s do the right thing.’</p><p>Like many organisations, our early accessibility efforts focused on compliance. We achieved <a href="https://www.w3.org/WAI/standards-guidelines/wcag/">WCAG 2.0 certification</a> for two products and began introducing features aimed at inclusivity. One example was within <a href="https://www.kintojoin.io/kinto-ride/"><strong>KINTO Ride</strong></a>, our ride-booking app, where users can specify additional needs — such as requesting a Wheelchair Accessible Vehicle — directly as part of the booking flow.</p><figure><img alt="A screenshot of a white iphone with a clear blue and white interface with a list of options such as ‘car seat’ and ‘wheelchair accessible vehicle’ — all options are selected. There’s a option to add a note also." src="https://cdn-images-1.medium.com/max/1024/1*3pob8k3lTUu7XtU9_E6HXg.png" /></figure><p>These steps towards progress were great, but they also exposed a limitation.</p><p>We realised that retrofitting accessibility after development was not enough for the future we wanted to build. Accessibility treated as a final checkpoint inevitably leads to compromises, rework, and missed opportunities. Certificates can be reassuring, but they can also be misleading — suggesting that a product is “done” when, in reality, accessibility is never truly finished.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*1aHHOXKAVlJQeIInsqz5Sw.jpeg" /><figcaption>Photo by <a href="https://unsplash.com/@stphnwlkr?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Stephen Walker</a> on <a href="https://unsplash.com/photos/a-close-up-of-a-cell-phone-on-a-table-Qk1hahQXmm4?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a></figcaption></figure><h3>Shifting the Mindset</h3><p>Becoming accessible by default requires more than meeting checklists; it requires a fundamental shift in how teams think and work. Our goal evolved into embedding accessibility from the very start of product development — so inclusion is considered from day one, not discovered through testing weeks or months later.</p><p>This shift is not without challenges. Accessibility work rarely delivers immediate, visible results, which can make any upfront investment in time and effort feel hard to justify. To address this, we formed an <strong>accessibility council</strong>, with early support from our CEO, <a href="https://www.linkedin.com/in/patrizia-niehaus-5584031b7/">Patrizia Niehaus</a>.</p><p>That leadership endorsement is helping to set expectations across the organisation and reinforce that accessibility is not optional or ‘nice to have,’ but part of how we build products. The council has a clear purpose: to explore how accessibility could be integrated into our processes from the outset, and to anticipate resistance and practical challenges before they became blockers.</p><p>Just as importantly, the council helps establish a more nuanced view of accessibility. We will be careful to avoid a rigid, one-size-fits-all interpretation of Web Content Accessibility Guidelines (WCAG). Not every checkpoint is relevant to every product. For example, in the <a href="https://www.kintojoin.io/kinto-ride/"><strong>KINTO Ride driver app</strong></a>, drivers use their phones mounted on dashboards — making landscape mode support unnecessary. Recognising these contextual realities can allow teams to focus on meaningful accessibility outcomes, rather than simply chasing compliance.</p><h3>Learning Through Practice</h3><p>For engineering and design teams, the learning curve was real.</p><p>Sasa, our Android Tech Lead, recalls confronting the first accessibility audit. “The results of our first application accessibility testing were not great,” he says. “The team was overwhelmed with the number of issues raised and didn’t always understand their importance.”</p><p>That initial shock, however, became a turning point. As the team became more familiar with accessibility principles — and, crucially, with how improvements directly help people with access needs — something changed. “The dev and design teams really stepped up and put a lot of effort into resolving what was reported,” Sasa explains. “The most important outcome was the change in mindset. Accessibility is now an integral part of our refinements and tickets.”</p><p>Some challenges were entirely new. “Navigating through the app using an external keyboard was a big one,” he notes. “We hadn’t used that at all before, so we had to adjust both design and code to resolve it properly.” What once felt like edge cases are now part of everyday development thinking.</p><p>On the iOS team, Filip experienced a similar evolution. “When you have no prior experience, accessibility can sound scary,” he admits. “But once you dig into it, it’s not that hard — though it can be time-consuming.”</p><p>He also highlights the complexity of accessibility standards in practice. “Sometimes our testing provider reported bugs that are actually how accessibility is designed to work,” he says. In those cases, he found himself demonstrating how Apple’s native apps behave or explaining how accessibility should be tested. “There are standards, but there’s also a lot of subjectivity. Understanding the platform conventions is just as important as following the rules.”</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*mIadsMu1cqsi1h7bIgwKsg.jpeg" /><figcaption>Photo by <a href="https://unsplash.com/@sigmund?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Sigmund</a> on <a href="https://unsplash.com/photos/person-holding-black-plastic-case-HE1zW44Zm_Y?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a></figcaption></figure><h3>Designing for Real Use</h3><p>Accessibility also reshaped how we think about design systems.</p><p>Daniele, UX lead for <a href="https://kintozero.com/">KINTO Zero</a>, describes how colour accessibility moved away from abstract labels and toward real usage. “Colour accessibility is defined by the contrast ratio between a foreground and a background,” he explains, “so it only applies when those roles are clearly defined.”</p><p>Early attempts to mark individual colours as “accessible” or “not accessible” proved unreliable. Light tints will always fail on white backgrounds, and dark shades will fail on black. Instead, the team focused on how colours are actually used. Primary colours for text, brand, and core UI elements were tested against both white and black backgrounds in light and dark modes, with accessibility validated at the component level.</p><p>Clear rules emerged from this approach. Components with light backgrounds never use light tints for foreground elements; dark backgrounds never use dark shades. Buttons are a clear example — the fill acts as the background, the label as the foreground, and the contrast is tested accordingly.</p><p>At the same time, the team acknowledged nuance. “Not every state is required to be accessible,” Daniele notes. Disabled components are intentionally low contrast to signal non-interactivity, and hover states are transitional by nature. Accessibility, here, is about clarity and intent — not rigid perfection.</p><h3>An Ongoing Journey</h3><p>Today, accessibility is no longer an afterthought or a compliance exercise. It is an ongoing journey — one that requires continuous learning, thoughtful application of guidelines, and constant reflection.</p><p>We know that a certificate does not mean the work is done, and a clean statement does not tell the full story of effort behind it. What matters most is the mindset shift: building products that consider diverse needs from the very beginning.</p><p>We’re excited about the steps we’ve taken so far — and aware that there is still much to learn.</p><p><em>What challenges is your organisation facing on the road to accessibility by default, and how are you working to overcome them?</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c0abdc51388a" width="1" height="1" alt=""><hr><p><a href="https://medium.com/the-design-loop/from-compliance-to-commitment-our-journey-toward-accessibility-by-default-c0abdc51388a">From Compliance to Commitment: Our Journey Toward Accessibility by Default</a> was originally published in <a href="https://medium.com/the-design-loop">The Design Loop</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Designing the UX of an Automated Data Consolidation Feature for ESG Reporting]]></title>
            <link>https://medium.com/the-design-loop/designing-the-ux-of-an-automated-data-consolidation-feature-for-esg-reporting-efb79bff1ae8?source=rss----135073075061---4</link>
            <guid isPermaLink="false">https://medium.com/p/efb79bff1ae8</guid>
            <category><![CDATA[esg-reporting]]></category>
            <category><![CDATA[product]]></category>
            <category><![CDATA[user-experience]]></category>
            <category><![CDATA[design]]></category>
            <category><![CDATA[data-consolidation]]></category>
            <dc:creator><![CDATA[Daniele Quartu]]></dc:creator>
            <pubDate>Wed, 07 Jan 2026 12:32:47 GMT</pubDate>
            <atom:updated>2026-01-07T12:32:46.434Z</atom:updated>
            <content:encoded><![CDATA[<h3>Designing an Automated Data Consolidation Feature for ESG Reporting</h3><p>When I first started working on a reporting tool for ESG data, I assumed the main challenge would be usability: clearer screens, more intuitive flows, better dashboards. That turned out to be only part of the story.</p><p>What became clear pretty quickly is that good design in reporting tools isn’t just about making things easy to navigate. It’s also about deciding what work the system should take on, and what work users should never have to do in the first place — especially when that work is repetitive, mentally draining, and prone to mistakes.</p><p>This became especially obvious once we started digging into data consolidation.</p><h3>Reporting Isn’t Hard Because Users Don’t Understand It</h3><p>Most people involved in ESG reporting are not confused about the data itself. They understand emissions, headcount, energy use, and financial metrics. The challenge usually isn’t knowledge — it’s the conditions people are working under.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-r70xUKG2Lz2EL7wtBgcSA.jpeg" /><figcaption>Photo by<a href="https://www.istockphoto.com/portfolio/AEIOUO?mediatype=photography"> Chaiwat Nookleang</a> on <a href="https://www.istockphoto.com/photo/esg-concept-digital-interface-on-laptop-with-environmental-social-and-governance-gm2237662356-652782734">iStock</a></figcaption></figure><p>For organisations with multiple subsidiaries, ESG data comes from many different places:</p><ul><li>Different countries</li><li>Different currencies</li><li>Different units of measurement</li><li>Different levels of reporting maturity</li></ul><p>All of that information eventually needs to come together in one consistent report.</p><p>Traditionally, this consolidation happens in spreadsheets. Data gets copied, pasted, converted, summed, adjusted, and checked — often more than once. Most of the time, this works well enough.</p><p>What stood out early on is that manual consolidation places a lot of responsibility on the user, and that’s where many issues tend to creep in.</p><h3>User-Friendly Isn’t Enough If Users Still Do the Heavy Lifting</h3><p>As designers, we often focus on making interfaces intuitive and pleasant to use. But even a well-designed interface can fall short if users still have to:</p><ul><li>Convert currencies manually</li><li>Normalise metric and imperial units</li><li>Apply formulas correctly</li><li>Track what has already been converted</li><li>Recalculate everything when one number changes</li></ul><p>At that point, the product may look clean, but the underlying workflow is still fragile.</p><p>This led me to a broader realisation: designing a reporting tool also means deciding where responsibility lies.</p><ul><li>What is the workflow?</li><li>Who does the work?</li><li>Who carries the risk?</li><li>Who is ultimately accountable for accuracy?</li></ul><p>If most of that burden sits with the user, the system isn’t really helping as much as it could.</p><h3>Why Data Consolidation Became the Perfect Test Case</h3><p>Data consolidation turned out to be a clear example of how design, automation, and accuracy intersect.</p><p>In ESG reporting, consolidation isn’t optional. Parent companies are expected to aggregate data from subsidiaries and present it in a standardised format, often aligned with regulatory requirements.</p><p>At the same time, consolidation increases risks:</p><ul><li>One incorrect conversion affects the whole report</li><li>One broken formula can cascade across metrics</li><li>One misunderstood unit can invalidate comparisons</li></ul><p>Working on this feature forced a simple but useful question: What would consolidation look like if users didn’t have to calculate anything themselves?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*OOix46cqEJlaoNQsxtW48Q.jpeg" /><figcaption>Photo by <a href="https://www.istockphoto.com/portfolio/tadamichi?mediatype=photography">tadamichi</a> on <a href="https://www.istockphoto.com/photo/ai-agent-and-generative-artificial-intelligence-robotic-processes-automation-and-gm2185337747-604108405#">iStock</a></figcaption></figure><h3>Letting the System Do the Math (and Some of the Thinking)</h3><p>One of the most important decisions we made was to remove calculations from the users’ responsibilities as much as possible.</p><p>Instead of asking users to:</p><ul><li>Enter pre-calculated values</li><li>Decide when and how to convert units</li><li>Keep track of exchange rates</li></ul><p>…we designed the system to handle those steps automatically.</p><p>Subsidiaries can report data in their local currencies and units. Consolidated results are then displayed using the parent company’s units — for example, euros instead of pounds, or metric instead of imperial — without manual intervention.</p><p>This change alone removed a whole class of potential errors.</p><h3>Automation Only Works If Users Can Still Trust It</h3><p>Of course, automation comes with its own challenge: trust. Users don’t just want a final number. They want to understand how that number was produced. This is especially valuable for the audit process. If the calculation process isn’t visible, automation can feel like a black box.</p><p>One lesson that became clear early on was this: Automation shouldn’t hide information — it should structure it.</p><p>Using the principle of progressive disclosure, users can always see original values, understand how conversions were applied, and trace aggregated numbers back to their sources. That transparency turned out to be just as important as the automation itself.</p><h3>The Real Goal: Reducing Errors by Reducing Effort</h3><p>One of the more interesting things I noticed is how closely effort and error are connected.</p><p>Every manual step — converting, recalculating, double-checking — introduces another opportunity for mistakes. By removing repetitive tasks, we didn’t just save time. We reduced risk in a very tangible way.</p><p>In that sense, designing automated data consolidation wasn’t about creating a “smart” feature. It was about acknowledging a simple principle:</p><p>If a task can be automated reliably, asking users to do it manually is rarely a good design choice.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3KVUThRLbXbCbwrxpZmZKg.jpeg" /><figcaption>Photo by <a href="https://www.istockphoto.com/portfolio/surprisestock?mediatype=photography">kanruthai khamthet</a> on <a href="https://www.istockphoto.com/photo/magnifying-glass-on-spreadsheet-paper-gm2212584269-629511952">iStock</a></figcaption></figure><h3>Looking Back</h3><p>Building an automated data consolidation feature changed how I think about the user experience design of reporting tools. User-friendly interfaces still matter.</p><p>But great design also:</p><ul><li>Takes into account the context and environment of the entire workflow</li><li>Reduces opportunities for human error</li><li>Absorbs complexity instead of pushing it onto users</li><li>Automates repetitive, high-risk tasks</li><li>Maintains transparency</li><li>Helps users feel confident in their data</li></ul><p>Data consolidation was a particularly good place to learn these lessons, because it sits right at the intersection of usability, trust, and accuracy. Once you start looking at design through that lens, it’s hard not to apply it everywhere else too.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=efb79bff1ae8" width="1" height="1" alt=""><hr><p><a href="https://medium.com/the-design-loop/designing-the-ux-of-an-automated-data-consolidation-feature-for-esg-reporting-efb79bff1ae8">Designing the UX of an Automated Data Consolidation Feature for ESG Reporting</a> was originally published in <a href="https://medium.com/the-design-loop">The Design Loop</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>