User Experience Innovation

Everything we know about online brand discovery is about to break. People aren’t browsing pages anymore. They’re asking AI. And that means the entire digital value chain is being rebuilt. Users are no longer starting with a Google search, clicking on links, and browsing websites. Increasingly, they’re turning to AI tools like ChatGPT, Perplexity, and Gemini to get direct answers, summaries, and recommendations. No links. No websites. No page one rankings. That shift is forcing a rethink of how visibility works online - and where SEO (Search Engine Optimization) fits. Traditionally, SEO has been about helping businesses appear higher in Google results. That meant optimizing websites to match search terms, earn backlinks, load quickly, and convert well once the user arrived. But that model depends on one thing: people clicking on search results. AI tools don’t work that way. When users ask ChatGPT for help - “compare project management tools,” “what’s the best CRM for startups,” “find me a cheaper alternative to X” - they’re not browsing. They’re expecting a direct, summarized answer in the chat itself. That’s where the biggest shift is happening. Data from Profound shows how user intent is evolving in AI environments: 1. Generative intent now leads at 37.5%. These are prompts where users ask AI to create or do something directly: write an email, summarize a document, recommend a product. 2. Informational intent - traditionally the most common in Google - is down to 32%. These are questions looking for facts or explanations. 3. Navigational intent - looking for a specific website - has collapsed from 32% in traditional search to 2% in AI. In chat, people don’t say “take me to X.com.” 4. Transactional intent has jumped 9x (to over 6%). That includes prompts like “buy running shoes,” “find deals on laptops,” or “compare prices.” 12% of prompts are conversational: things like “thanks,” “make it shorter,” or “can you add a joke?” - which play a subtle but growing role in shaping how AI interprets tone, preferences, and even brands. Why does this matter? Because all of this happens before a user visits a website - if they visit at all. In this new model, there’s no clear click path. No landing page. No bounce rate. That makes most current marketing KPIs and tools largely obsolete. A new wave of startups is helping brands adapt - decoding how AI models reference products and content. The focus has shifted from rankings to being included in AI-generated responses. The move from SEO to “AI visibility” is early, but accelerating. The question now isn’t: “How do I get more search traffic?” but “What do AI systems say about the brand - and is it even part of the answer?” Because soon, it won’t just be users asking. It will be AI agents deciding - on their / our behalf. Are you ready? Opinions: my own, Graphic source: Profound 𝐒𝐮𝐛𝐬𝐜𝐫𝐢𝐛𝐞 𝐭𝐨 𝐦𝐲 𝐧𝐞𝐰𝐬𝐥𝐞𝐭𝐭𝐞𝐫: https://lnkd.in/dkqhnxdg

Something feels off in UX right now 🥺 My 2026 UX Design predictions Over the last months, I’ve had the same conversations again and again. With designers. With teams. With leaders. “Should I learn AI?” “Is my role still relevant?” “Am I falling behind… or overreacting?” Here’s what I’m seeing for UX in 2026 👇 1️⃣ Interfaces are dissolving With Generative UI, interfaces are created on demand. No more fixed screens. UX shifts from designing flows to defining systems, constraints, and intent. 2️⃣ UX research is scaling radically AI can analyze thousands of open-ended responses, run deep research, and surface patterns in minutes. The role of designers and researchers is changing: less execution, more sense-making, validation, and decision-making. 3️⃣ Designing for AI is no longer optional AI products learn from users. That makes them powerful and incredibly confusing if UX is missing. Trust, explainability, recovery UX, and mental models are now core design work. “Just type something” is not a UX strategy. 4️⃣ AI agents change how systems behave AI doesn’t just respond anymore. It plans and acts on behalf of users. This breaks traditional UX patterns. The key question becomes: When can a system act on its own and when must it stop? 5️⃣ Vibe coding removes technical barriers Ideas turn into prototypes and products at near-zero cost. The real bottleneck is no longer code, but judgment. UX shifts even further toward problem framing, direction, and quality control. 6️⃣ Roles matter less. Ownership matters more. Job titles lose their edge. Execution becomes cheap. What matters is who takes responsibility, who makes decisions, and who asks the uncomfortable questions. 7️⃣ Personalization is entering uncomfortable territory AI systems are building long-term memory about users. Helpful at first. Creepy the moment people realize what the system knows. UX must make data use visible, adjustable, and understandable. Not more data. More choice. My biggest takeaway: 2026 is not about designing faster. It’s about deciding better. UX is moving away from interface design toward intent, systems thinking, and responsibility. If you work in UX and ignore AI, you won’t be replaced. You’ll slowly become irrelevant. If you stay curious, critical, and intentional about how you use AI, this is one of the biggest opportunities UX has ever had. P.S.: I also recorded a podcast episode (link in the comments) where I go much deeper into these shifts, with concrete examples and reflections from my own work. If you want the longer version beyond a LinkedIn post, feel free to give it a listen.

Product managers & designers working with AI face a unique challenge: designing a delightful product experience that cannot fully be predicted. Traditionally, product development followed a linear path. A PM defines the problem, a designer draws the solution, and the software teams code the product. The outcome was largely predictable, and the user experience was consistent. However, with AI, the rules have changed. Non-deterministic ML models introduce uncertainty & chaotic behavior. The same question asked four times produces different outputs. Asking the same question in different ways - even just an extra space in the question - elicits different results. How does one design a product experience in the fog of AI? The answer lies in embracing the unpredictable nature of AI and adapting your design approach. Here are a few strategies to consider: 1. Fast feedback loops : Great machine learning products elicit user feedback passively. Just click on the first result of a Google search and come back to the second one. That’s a great signal for Google to know that the first result is not optimal - without tying a word. 2. Evaluation : before products launch, it’s critical to run the machine learning systems through a battery of tests to understand in the most likely use cases, how the LLM will respond. 3. Over-measurement : It’s unclear what will matter in product experiences today, so measuring as much as possible in the user experience, whether it’s session times, conversation topic analysis, sentiment scores, or other numbers. 4. Couple with deterministic systems : Some startups are using large language models to suggest ideas that are evaluated with deterministic or classic machine learning systems. This design pattern can quash some of the chaotic and non-deterministic nature of LLMs. 5. Smaller models : smaller models that are tuned or optimized for use cases will produce narrower output, controlling the experience. The goal is not to eliminate unpredictability altogether but to design a product that can adapt and learn alongside its users. Just as much as the technology has changed products, our design processes must evolve as well.

🧠 “How We Brainstorm And Choose UX Ideas” (+ Miro template) (https://lnkd.in/eN32hH2x), a practical guide by Booking.com on how to run a rapid UX ideation session with silent brainstorming and “How Might We” (HMW) statements — by clustering data points into themes, reframing each theme and then prioritizing impactful ideas. Shared by Evan Karageorgos, Tori Holmes, Alexandre Benitah. 👏🏼👏🏽👏🏾 Booking.com UX Ideation Template (Miro) https://lnkd.in/eipdgPuC (password: bookingcom) 🚫 Ideas shouldn’t come from assumptions but UX research. ✅ Study past research and conduct a new study if needed. ✅ Cluster data in user needs, business goals, competitive insights. ✅ Best ideas emerge at the intersections of these 3 pillars. ✅ Cluster all data points into themes, prioritize with colors. ✅ Reframe each theme as a “How Might We” (HMW) statement. ✅ Start with the problems (or insights) you’ve uncovered. ✅ Focus on the desired outcomes, rather than symptoms. ✅ Collect and group ideas by relevance for every theme. ✅ Prioritize and visualize ideas with visuals and storytelling. Many brainstorming sessions are an avalanche of unstructured ideas, based on hunches and assumptions. Just like in design work we need constraints to be intentional in our decisions, we need at least some structure to mold realistic and viable ideas. I absolutely love the idea of frame the perspective through the lens of ideation clusters: user needs, business problems and insights. Reframing emerging themes as “How-Might-We”-statements is a neat way to help teams focus on a specific problem at hand and a desired outcome. A simple but very helpful approach — without too much rigidity but just enough structure to generate, prioritize and eventually visualize effective ideas with the entire team. Invite non-designers in the sessions as well, and I wouldn’t be surprised how much value a 2h session might deliver. Useful resources: The Rules of Productive Brainstorming, by Slava Shestopalov https://lnkd.in/eyYZjAz3 On “How Might We” Questions, by Maria Rosala, NN/g https://lnkd.in/ejDnmsRr Ideation for Everyday Design Challenges, by Aurora Harley, NN/g https://lnkd.in/emGtnMyy Brainstorming Exercises for Introverts, by Allison Press https://lnkd.in/eta6YsFJ How To Run Successful Product Design Workshops, by Gustavs Cirulis, Cindy Chang https://lnkd.in/eMtX-xwD Useful Miro Templates For UX Designers, by yours truly https://lnkd.in/eQVxM_Nq #ux #design

Scalable? Yes. Boring? Also yes. Let’s be real: We’re so obsessed with scalability, we’ve forgotten how to surprise users. Design systems. Tokens. UI kits. All optimized for consistency—but at what cost? 1.Every product is starting to feel like a remix of the same Figma file. 2.Same rounded corners. Same muted palette. Same safe CTAs. Is every product supposed to look the same? Don’t get me wrong—scalability matters. But sameness isn’t strategy. If your brand blends in with everyone else’s, you’re not scalable. You’re forgettable. Design systems are vital. They’re the scaffolding that holds the experience together across teams, touchpoints, and time. They bring consistency, reduce chaos, and speed up collaboration. But here’s the catch: your design system should never mute your brand.The answer isn’t to abandon structure—it’s to evolve it. A well-crafted design system doesn’t erase your identity—it enhances it. It should include: >>UI pattern libraries tailored to your brand’s personality >>Micro-interactions that feel uniquely “you” >>Visuals, tone, and motion that speak your story >>And most importantly, components built with a deep respect for your product’s “why.” Because good design systems scale consistency. Great ones scale character. So no—every product doesn’t have to look the same. But every great product does need a system that helps it stay true to what makes it meaningful—and memorable. Design systems shouldn’t limit creativity—they should amplify identity.

From GenAI to GenUI We’re witnessing a shift as significant as the leap from MS-DOS to graphical user interfaces. The AI era marks our latest upgrade in how we interact with technology. For decades, we designed for workflows and specific actions. Everything was deterministic. Behind every interface sat a flowchart, with logic carefully coded. The backend made decisions, and the frontend rendered them. This model worked because we could predict every path a user might take. With agents, this paradigm breaks down. Text alone isn’t sufficient anymore. Chat works for conversation, but interaction demands something more. We need to engage with agents, not just talk to them. Reasoning state and intent become critical factors in the exchange. LLMs can now generate UI, and this capability feels like natural progression. Model Context Protocol enables mini-apps to emerge on the fly, no longer bound by deterministic rules. This opens the door to genuine hyper-personalization. We’ve moved from designing screens to designing for outcomes. Agents now dynamically assemble workflows based on intent, available data, and accessible tools. The fact that agents can create interfaces without traditional designers and developers is revolutionary. We can finally shift from UI-centric thinking to truly user-centered experience design. This fundamentally transforms the designer’s role. We’re no longer pixel pushers or interface assemblers. The work of arranging buttons, spacing elements, and crafting individual screens can now be handled by agents. Instead, designers become architects of experience, defining the principles, guardrails, and intent that shape how agents respond. We set the boundaries of possibility, orchestrate the logic of interaction, and ensure coherence across dynamic, personalized experiences. Our canvas expands from static screens to adaptive systems. We design the intelligence behind the interface, the relationships between user needs and agent capabilities, the quality standards that govern generated UIs. We curate outcomes rather than outputs. The ability to adapt, reorganize, and respond to both user intent and application context is transformative. With reasoning and action combined, agents can generate dynamic artifacts that enable interaction, not merely conversation. What a time to be alive as a designer! 🫶 #ai

Your research is only as good as your ability to get people to listen to it.  Here's 7 tips for making sure your insights actually land. 📂 Start with conclusions, not methodology Think of your debrief as a landing page–you need to hook people immediately. Put your key takeaways front and center. No one has time to wade through your research methods before getting to the good stuff. Everyone is busy with their jobs already. 📂 One finding per slide Don't overwhelm your audience with multiple insights at once. Share one finding per slide, support it with data (mix qualitative and quantitative), and include a clear recommendation. Yes, recommendations! Don't just drop insights and run. Your job isn't done until you've suggested what to do next. 📂 Connect to business goals Your organization cares about metrics and outcomes, not research for its own sake. Frame your insights in terms of business impact. For example, this finding will help us reduce the 30% churn we're seeing in week 1. 📂 Use real user voices Nothing makes research stick like hearing it directly from users. Include direct quotes and, if possible, short video clips. The more human connection you create between stakeholders and users, the more memorable your insights become. 📂 Ditch the UX jargon Simplify everything and speak in terms business stakeholders understand. 📂 Address stakeholder fears When executives push back, it's usually fear-driven. Find out what they're afraid of missing, losing, or failing at—then frame your insights as solutions to those fears. 📂 Save methodology for last Your professional expertise should be trusted. Keep the "how we made the sausage" details for the end. What's your best tip for making research insights stick?

Start with people. Invest in real end user problems, not in fleeting ideas. That is the role of designers in the innovation process, and it is pivotal. This video shows why. Because more than most, designers empathize with end users. They put themselves in the shoes of an 8-year-old patient who will soon have their first MRI exam. And ask: what do they really need?   The answer: they want reassurance. That is why solutions like Pediatric Coaching matter so deeply. To help children prepare for their MRI exam in an engaging way. So they can get confident, finish the exam, and say: “It wasn’t scary at all!”

Are you building products for users who look nothing like you? Then focus on accelerating your “pathway to empathy”. This week on the Product Thinking podcast, I had Tony Brancato on and we talked about customer research and empathy as a critical step in building successful products. One thing he said really stuck with me: "Find a pathway to empathy as quickly as possible. Anytime I did that as a young PM, I always built better products." The key word here is 𝘲𝘶𝘪𝘤𝘬𝘭𝘺. Too many product managers spend weeks or months trying to understand their users through surveys and data analysis. But if you're building financial products for seniors and you're 28, or creating tools for busy parents when you don't have kids, you need faster ways to bridge that gap. Shadow customer service calls. Visit where your users actually are. Conduct quick, focused interviews. The goal isn't perfect research methodology but rapid understanding that prevents you from building the wrong thing. I've seen PMs transform their product decisions within days of getting real exposure to their users. Having all the data matters less than developing genuine empathy for problems you've never experienced yourself. What's the fastest way you've built empathy with users who are different from you?

Innovation is unlikely to be achieved through consistent, conventional thinking. Most teams unknowingly favour 𝗼𝗻𝗲 𝘁𝘆𝗽𝗲 𝗼𝗳 𝘁𝗵𝗶𝗻𝗸𝗶𝗻𝗴—and it’s limiting their potential. Ever been in a meeting where big, bold ideas get shut down too soon? Or one where endless brainstorming leads to zero action? That’s the clash of Divergent vs. Convergent Thinking—and most workplaces get the balance wrong. Convergent thinkers love logic, structure, and clear answers. Divergent thinkers thrive on possibilities, creativity, and unconventional ideas. The real challenge? Most workplaces reward convergence and overlook divergence. 💡 If you’ve ever felt like your ideas weren’t landing, this might be why. (Chances are, you already use both thinking styles—just not in the right sequence.) Here’s how to make both work for you, not against you: 1) Don’t Judge Ideas Too Soon ↳ Separate Idea Generation from Decision-Making ⎌ Innovation dies when every idea is scrutinized immediately. ✔︎ First, expand possibilities—then refine. 2) Create a Safe Space for Bold Ideas ↳ Creativity flourishes when ideas evolve, not when they’re dismissed. ⎌ Innovation dies in judgment-heavy environments. ✔︎ Encourage “Yes, and…” instead of “No, but…” to keep ideas flowing. 3) Pair Opposites for Problem-Solving ↳ Convergent thinkers help refine wild ideas. ↳ Divergent thinkers help break rigid thinking patterns. ⎌ Mixing the two? That’s where teams get stuck. 4) Pair Thinkers Strategically ↳ Visionaries need detail-oriented partners to bring ideas to life. ↳ Give each role equal importance. ✔︎ If an idea feels too safe, ask, “What’s a bolder alternative?” ✔︎ If it’s too abstract, ask, “How do we make this actionable?” 5) Create Space for Both Thinking Modes ⎌ People won’t share unconventional ideas if they fear judgment. ✔︎ Encourage curiosity over criticism. ↳ Schedule separate sessions for idea generation vs. decision-making. ✔︎ You’ll get better ideas and faster execution. 💡 The best teams don’t just have great ideas—they know how to shape them into reality. Which thinking style do you lean toward? Comment below! ------------------- I’m Jayant Ghosh. Follow me in raising awareness for mental health that inspires growth and well-being.