Helio

AI is forcing teams to build better systems for decision-making. We're looking at the systems teams are building to stay grounded. Everything from UX metrics and research practices to AI skills that help turn loose ideas into clearer decisions. We currently see teams (ourselves included) navigating what should be automated, what still needs human judgment, and how teams measure whether experiences are actually helping people's work. → Juhee Dubey where AI helps research and where humans still matter most → Bryan Zmijewski on UX metrics that keep AI work grounded → The ZURB team on AI skills for better product decisions If your team is sorting through AI tools, workflows, and decisions right now, shoot us a note. We’d love to share a few of the systems helping teams stay focused on what actually matters.

Show why UX research matters. We love Victor Yocco, PhD’s point that sharing UX research effectively is just as important as doing the research. With AI helping teams create and test coded prototypes faster than ever, the challenge is no longer making ideas. It's helping people understand which ideas are worth pursuing. UX professionals need to convince decision-makers by making their insights clear, useful, and practical. Persuasion isn’t about manipulation. It’s about helping stakeholders see and act on the value of research. Check out his article here: https://lnkd.in/epSXEQip Victor recommends the Hovland-Yale Model of Persuasion to make UX presentations more effective by: 1. Building credibility  Who is presenting matters 2. Being clear How the message is framed affects understanding 3. Knowing your audience What they care about shapes their response Essentially, good design only drives change when it’s communicated well. 💬 We asked him why he wrote the article: “I wanted to make some topics academic concepts relevant and accessible for UX professionals. Persuasive communication techniques can help us present our findings in ways that are more impactful to the audience, opening them up to taking the recommendations of our research.” Well said. We’re on board. As AI makes it easier to create more ideas, screens, and prototypes, our recent release of Glare helps teams focus on making better decisions. The framework connects user feedback and UX metrics to the choices that drive product and business outcomes. #uxresearch #productdesign

Product design decisions create value for users *and* business. Bryan Zmijewski argues that as AI speeds up design production, the value shifts to making better design decisions. Anyone will be able to use AI to produce basic designs, but differentiation comes from skilled, strategic decision-making, across interface, structure, technical, and strategic levels. That’s where product design teams can create business impact and products users love. Check out his post: https://lnkd.in/g4Ta3QJn Four decision areas help companies win and keep users happy: 1. Interface: Make it easy and enjoyable to use 2. Structural: Organize the experience 3. Technical: Make it work well 4. Strategic: Focus on what matters most Here are some of his other posts focused on product design decisions: There’s not always a clear way to guide design decisions. https://lnkd.in/g-dzSM3Z Values, culture, and leadership shape product design decisions. https://lnkd.in/gx6Sr8ZH Align design decisions with business goals and user needs. https://lnkd.in/g8qekpkn Product design decisions need nurturing. https://lnkd.in/g62JQMiN Design isn’t a choice between intuition and data. https://lnkd.in/gyFM4H4D Design stories drive decisions. https://lnkd.in/gHiMAjZx Business decisions shape design around today’s problems. https://lnkd.in/gtvjMG9F 💬 We asked Bryan why he writes about decision making: “Ideas are cheap. AI is making it easier to create options. What separates great teams is knowing which option to pursue and why.” We’re excited to partner with product and design leaders on ZURB Glare, our open data-informed framework. Glare captures UX metrics and design signals, turning them into decisions that drive business results and products users love. Join the community and help shape the future of design. https://lnkd.in/gWSCGaxc #productdesign

Ideas are cheap. Value comes from faster learning loops. We agree with Pavel Samsonov that AI makes creating things faster and cheaper, but that alone doesn’t create value. What matters most is having strong learning loops: ways to gather feedback, make decisions, and adapt quickly. If those loops are weak, AI just overwhelms you with outputs. If they’re strong, AI multiplies your impact and helps turn outputs into real customer value. Check out his article: https://lnkd.in/dqYfFYDv Here's the tldr: • Single loop = build right • Double loop = solve right • Triple loop = choose right → Single: Does the implementation meet the requirements? Execution. Making sure the thing you built matches what was asked for. QA testing if a feature works as written in the spec. → Double: Do the requirements describe the best solution to this problem? Problem, solution fit. Checking if the requirements themselves are right, not just the implementation. Testing whether a feature actually solves the user’s problem in the best way. → Triple: Is this the right problem to solve? Strategy. Stepping back to confirm if the team is even solving the right problem. Asking whether improving a small feature is worth it, or if the real problem lies elsewhere. 💬 We asked Pavel why he wrote the article: “It was a response to output-driven thinking that I saw all around me as a product manager. When I came across the concept of learning loops, I thought they provided an excellent scaffolding for showing how UX methods can help address the gap between "definition of done" and what you might call a "definition of good. Back in 2023 it was still possible to believe that AI actually could generate outputs with meaningful quality, and I wanted people to understand that it would not solve all their problems. Of course in 2025 we see that things like Figma Make don't even reach the bar everyone assumed was right around the corner!” Love it. Helio helps teams spot signals with UX metrics so they know what’s working and what’s not. By turning user feedback into clear signals, it makes decisions faster and more confident. #productdesign

AI should support, not replace, UX researchers. We like Juhee Dubey’s point that AI has a place in user research, but not in every part of it. She recommends that teams be thoughtful about how they use AI. Teams need to decide carefully what to automate, what to enhance with AI, and what must remain in human hands. Check out her article: https://lnkd.in/gZji-Kh2 She introduces a framework with three parts: 1. Automate Use AI for repetitive or time-consuming tasks, like organizing data, coding responses, or speeding up analysis. 2. Augment Let AI assist humans by offering suggestions, patterns, or insights that researchers can build on. 3. Human only Keep people in charge of tasks that require empathy, context, and ethical judgment… things AI can’t handle well. 💬 We asked Juhee why she wrote the article: “I wrote this article to share insights from my own experimentation with AI in UX research, clarifying which parts can be automated, which benefit from AI support, and which require human judgment, without compromising research integrity. My goal was to help others navigate AI thoughtfully, not just chase efficiency.” Love it. Here are some other great resources: A Personal Take on What Works & What Doesn’t, by Ellina Morits https://lnkd.in/ghxSiwd6 Practical Methods for Each Research Stage [+ Expert Tips], by Kateryna Mayka, Daria Korniienko https://lnkd.in/gw4H-Dep A practitioner’s journal on navigating UX in the age of AI, by Jonathan Montalvo https://lnkd.in/g5JDVpzW Accelerating Research with AI, by Kate Moran and Maria Rosala https://lnkd.in/gZyzP67R 20+ AI Tools for Every Phase of UX Research, by Ben Wiedmaier https://lnkd.in/gQGVJwHv Exploring AI for User Research – Where Does It Actually Help? Reddit https://lnkd.in/gtG5vrE9 How much will AI impact the future of UX research? Reddit https://lnkd.in/ggiNrWbp We’re excited to partner with product and design leaders on ZURB Glare, our open data-informed framework. Glare turns UX metrics and design signals into clear decisions, helping researchers test faster and uncover what matters most to users. 👉 Join the community and help shape the future of product design. https://lnkd.in/gWSCGaxc

UX metrics align AI workflows with user needs. We agree with Bryan Zmijewski that UX metrics help connect user needs to business goals, especially as AI and automation change how users interact with products. They give clear signals into user behavior, helping businesses make good decisions, improve experiences, and stay ahead. Check out his post: https://lnkd.in/gBzmE3m6 As AI changes how products work, it gets harder to balance what users need with what the system is doing. UX metrics help product and design teams stay focused on people by measuring things like trust, ease of use, and how well the experience works. By testing often in the iterative process (and tracking these metrics), teams can keep improving their design work even as the tooling is changing weekly. Big ideas: → Metrics track user behavior, satisfaction, and preferences, helping businesses create user-focused solutions and align experiences with goals like retention and loyalty. → AI complicates separating user needs from system interactions. Metrics like trust and satisfaction measure how well systems meet user expectations. → As technology shifts, metrics like comprehension and error rate reveal usability issues, helping businesses adapt and improve quickly. 💬 We asked Bryan why he wrote the post: “I’m seeing more teams get lost in AI tooling. It’s exciting for sure, but teams need to get back to the basics of simply understanding their customers. User needs give AI product work direction. UX metrics help teams measure if it’s working.” Using tools like Helio to track metrics early helps refine products, meet user needs, and stay ahead in a changing market. #productdesign #uxresearch 

Your users have the answers you need. Bryan Zmijewski argues that you just need the right way to uncover what they want, need, and do. UX metrics make feedback clear and actionable. Instead of guessing from conversations or hunches, metrics reveal real needs, how they change, and how solving them drives the business forward. They turn fuzzy feedback into signals teams can act on. Check out his post: https://lnkd.in/gkthVT-y Collecting signals from UX metrics: • Find what users want, need, and do • UX metrics clarify feedback • Don’t guess from talk or hunches • Metrics reveal real user needs • Track how those needs change over time • Map solutions to business impact • Turn fuzzy feedback into clear signals Additional takes from him on user needs: → User needs can be quantified. https://lnkd.in/g37WeU2S → User needs and business goals don’t align one-to-one. https://lnkd.in/grtmxDCy → UX metrics are key to meeting user needs and business goals. https://lnkd.in/gBzmE3m6 → Identifying user needs isn’t enough to make an impact with design. https://lnkd.in/ga592yQA → Core user needs are found in overlapping data. https://lnkd.in/gQmn5j7V → Figuring out what users need requires different methods. https://lnkd.in/giuyje9t → Great journey maps reveal the intersections of user needs. https://lnkd.in/gEMYTGu4 → Figuring out what users need isn’t straightforward. https://lnkd.in/gbpAyrVP 💬 We asked Bryan why he writes about user needs: “User feedback is full of answers, but without metrics it’s just noise in your projects. UX metrics turn insights into signals you can act on and tie to business results.” Helio helps you collect UX metrics that turn user feedback into clear signals. These signals show where people struggle, what they value, and how design changes impact results. This keeps your team focused on what matters most.

Great product decisions take thinking. You can’t simply prompt your way to a great product without knowing what matters. We're seeing that the teams that are getting the most value from AI are building stronger thinking underneath the work. As AI floods teams with more ideas, screens, flows, and outputs... avoiding the "AI noise" has become something that requires extra legwork. Teams that leap ahead will gauge how systems respond, reduce ideas down to what matters, and use AI to explore without disconnecting from the user needs from the business goals. This week, we look at: • Mike Waszazak on designing AI behavior, not just interfaces. • Kike Peña on why systems thinking matters more as AI speeds things up. • Bryan Zmijewski on why more AI output doesn’t always create better decisions. ❤️ We’re pushing hard on the next version of Glare to bring more meaningful thinking into product and design decisions. Keep an eye out in teh forum, more to come.

Visual frameworks turn complex ideas into simple patterns. We love Dave Gray’s library of visual frameworks. A visual framework is a simple sketch that shows a pattern you can use to think through problems, spark ideas, and build shared understanding with others. They’re quick to scan, flexible to adapt, and open-ended so you can shape them to fit your situation. Check it out: https://lnkd.in/e2zZx_DQ Here’s how to think about them: • A sketch of a clear, meaningful pattern • Quick to scan and match to your situation • Rough enough to invite you to complete the story • Open enough to suggest variations and new ideas • Small enough to fit on an index card • Strong enough to scale into a full explanation 💬 We asked Dave why he created the visual frameworks site: “It's a labor of love. I'm still not sure what fascinates me about visual frameworks. I think they are powerful tools for thinking about and modeling situations and solving problems.” You can check out more of Dave’s thinking here: https://lnkd.in/gt4WH4v8 Visual Frameworks card deck (Video), by Werner Puchert https://lnkd.in/gBCiVtY5 Visual Riddles & Visual Frameworks, by Sarah Mattern https://lnkd.in/gmV_WFHf How to Turn Your Ideas into Visual Frameworks, by Jay Acunzo https://lnkd.in/gxMfpMGW A better way to create Visual Frameworks, by Joao Landeiro MSc. https://lnkd.in/grFrAGnq 👉 P.S.  We’re working with product and design leaders on Glare, an open framework that uses data to guide better decisions. Glare turns experiments into decisions with UX metrics, helping teams test ideas fast and uncover what matters most to users. Join our community: https://lnkd.in/gynueqWu

Designing with AI starts with shaping logic, tone, and flow. We like Mike Waszazak’s argument that AI should be seen as a new design material. Building a conversational assistant isn’t about screens or visuals, but rather shaping how the system thinks and responds. Instead of wireframes and layouts, design needs to prototype conversations, define reasoning steps, and shape the assistant’s tone and personality. This shift leads to experiences that feel more natural and helpful. Check out his article: https://lnkd.in/gFnPqtqd Here are his big ideas: 1. AI as design material Treat AI as something to shape, not just use. 2. Conversations over interfaces Design how the system thinks, responds, and speaks. 3. Build with roles and agents Use orchestrators, evaluators, and fallbacks to guide flow. 4. Prototype without code Test conversations through scripts, roleplay, or simple tools. 5. Design for trust Clarity, tone, and resilience make the assistant feel reliable. Love the exploration and process! 👉 Our open framework Glare helps teams translate what users think and feel by turning feedback into simple UX metrics. Our Skills help teams shape AI experiences that feel clear, helpful, and natural. Join the Helio Glare community: https://lnkd.in/gynueqWu #productdesign #productmanagement