About
As VP of Strategy & Product at Adobe, I lead Adobe Experience Manager, Adobe…
Articles by Loni
Activity
9K followers
-
Loni Stark shared thisReally enjoyed being part of Adobe Incubator Pitch Day this year. Both as a new board member and as a judge. The Incubator is an initiative within Adobe that Malte Bernholz has championed over the years, with the inspiring goal of creating space to foster ideas that may not fit comfortably into the main innovation track. It gives passionate innovators, who can be anyone in the company, the opportunity to build, test in market and see where an idea can go. Products such as Adobe Brand Concierge, Adobe Podcast and Aqua, to name a few, were born from this initiative. This year’s Pitch Day brought a record 245 submissions from around the world. What it highlights is not only the creativity from within, but also the exciting possibilities of human ingenuity in a world where AI is collaborator and tool. Adobe’s mission is to empower everyone to create. I believe that starts from within Adobe too...we can only give others that creative energy if we continue to awaken and nurture it in ourselves. Grateful and honored to be part of it, and excited to see where this year’s ideas go. Congratulations to all the finalists, and to Malte for building something that keeps getting stronger. #AdobeIncubator #Innovation #AI #CreativityLoni Stark shared thisBig things are brewing at Adobe Incubator! This year's Pitch Day pulled in a record 245 applications, with finalist teams showcasing how AI is turning ideas into prototypes faster than ever across creative, marketing, and enterprise. Learn more here: https://adobe.ly/4wVcew9
-
Loni Stark shared thisFor twenty years, web economics ran a familiar pattern: search, click, site, conversion. AI has broken this sequence. We are now comparing products, building a shortlist, exploring options…or deciding what is worth buying before ever directly reaching out to a brand. Sometimes we have a brand name we trust, other times we don’t. But by the time we click, we're already midway through our decision. We’ve been tracking this shift for the last 3 years at Adobe. Our latest AI data reinforces how brands must build brand and demand differently. In July, AI-referred retail traffic was up 62% YoY, converting at a rate 60% higher than non-AI traffic and generating 53% more revenue per visit. Travel shows a similar pattern, with AI-referred traffic up 119% YoY. Zooming out from numbers...pay attention to what happens before the click and after it. 𝐁𝐞𝐟𝐨𝐫𝐞 𝐭𝐡𝐞 𝐜𝐥𝐢𝐜𝐤 𝐨𝐫 𝐧𝐨 𝐜𝐥𝐢𝐜𝐤: The model did the work people used to do across a brand’s website and third-party sources: research, comparison, filtering, recommendation. AI shaped the customer’s understanding of the category and, increasingly, the brands within it. AI is doing the legwork, crawling sites, looking at signals from trusted sources, and synthesizing all of this. Its appetite for information, and for the kind that allows it to personalize our conversations, is at a scale no human-only web needed. This creates a massive new need: the website as an edge point for AI to retrieve from, understand, and mediate to the human at the other end. 𝐀𝐟𝐭𝐞𝐫 𝐭𝐡𝐞 𝐜𝐥𝐢𝐜𝐤: More times, your website is not the front door anymore. Yet it's an even more critical door. AI-referral traffic is arriving high-intent and midway through a conversation. The bar to respond fast and add something no third-party LLM can has never been higher. How do you earn the direct customer relationship and create a repeat, loyal visitor? (I think this will become a more important metric.) The role of owned properties has shifted, but not its economic purpose. The new role is to make brands legible to third-party LLMs at search and retrieval moments, and when humans or their agents arrive: convert/engage, give them a reason to return, and capture insight that helps the brand connect better directly with the customer. Your site becomes that source of truth: products, claims, pricing, provenance, policies, expertise, proof, availability, context. It is where what makes the brand distinct can be made explicit, and where a person, or increasingly an agent, can go deeper and act. It’s why, from a product strategy perspective, we’re committed to staying at the forefront of AI visibility with Adobe Brand Visibility, connected to innovations in Adobe Experience Manager, Adobe Commerce and Adobe Brand Concierge, so brands have both the new tools and the foundations they need to win the job of being discovered and chosen, over and over again in the Agentic Web era. Go deeper in the links in comments. #AgenticWeb
-
Loni Stark reposted thisLoni Stark reposted thisToday we announced general availability for Adobe Brand Visibility, a new way to boost your brand visibility among both humans and AI agents. The platform relies heavily on Semrush's data infrastructure: → 𝗧𝗵𝗲 𝗺𝗮𝘀𝘀𝗶𝘃𝗲 𝗽𝗿𝗼𝗺𝗽𝘁 𝗱𝗮𝘁𝗮𝘀𝗲𝘁: Adobe Brand Visibility is backed by our industry leading global database of nearly 300 million real-world AI search prompts to track content gaps and competitive share-of-voice. → 𝗧𝗵𝗲 𝘀𝗲𝗮𝗿𝗰𝗵 𝗮𝘂𝘁𝗵𝗼𝗿𝗶𝘁𝘆: The platform leverages our 18-year SEO dataset of 28.8 billion keywords and 43 trillion backlinks to reveal exactly where traditional authority can drive your AI citations. → 𝗔𝗴𝗲𝗻𝘁𝗶𝗰 𝗮𝘂𝘁𝗼-𝗼𝗽𝘁𝗶𝗺𝗶𝘇𝗮𝘁𝗶𝗼𝗻: Prioritized recommendations deploy updates directly to AI agents in just minutes, ensuring LLMs access the latest version of your brand narrative. This new unified solution turns raw search intelligence into measurable brand visibility by connecting every action directly to pipeline and revenue. For brands that want to win in the age of AI search, Brand Visibility isn’t a nice-to-have – it’s mission critical. More in the comments.
-
Loni Stark shared this🎉 Adobe Brand Visibility is generally available today! It combines Semrush's corpus of nearly 300 million AI search prompts and their SEO keywords (AI is now one of the largest consumers of search itself, funny how that works) with agentic optimization from Adobe LLM Optimizer. Most brands now look at what AI systems say about them. Fewer ask what comes earlier: what are real people actually asking in my category? In a world filling with synthetics, knowing what is real separates optimizing for a market from optimizing for a simulation of it. Proud of the collaboration across Adobe to build this, including our newly joined colleagues from Semrush. That is the product news, and a real piece of how the marketing operating model shifts with agentic commerce. There is a harder question underneath. Six months ago I said that once everyone has the same models, we would drown in AI slop, and the winners would be brands that know what they sound like. That still holds. But it raises what I did not answer: where does differentiation live? Not on third-party surfaces. You influence those. You do not control them. It lives on the properties you own. Which is why solving only for third-party visibility feels like a right-now answer to a long-game question. Discovery moving to ChatGPT and Gemini does not make your site and storefronts less important. It makes them more. Your site is how models learn about you, and where people land when an AI recommendation sends them deeper, with more intent than any channel has sent in years. Soon it is what their agent will visit on their behalf. So the work is doing both, everywhere. I see three ways. Fix the plumbing. Structured data, unblocked crawlers, technical fixes, no site changes. Minutes, not an engineering queue. The right first move. Fix the story. Deeper than copy. Your content and commerce infrastructure has to hold the truths both audiences need and deliver them differently: enough structure for an AI to understand you at discovery, enough substance to give a person reason to come direct. Someone arriving from an LLM arrives mid-conversation. Continue it, do not restart. Build for each. People and agents do not take things in the same way. A person watches, scans, feels. Humans are meaning makers. An agent turns everything into language and checks for congruence. Eventually you will not make one thing and expose it twice. Same claims, different grammar. One identity underneath. The moment those diverge you have a cloaking problem, which never survives the next crawler update. We are making the technology question easier today. The harder one is which of the three an organization is ready for. We have been building for all three. Brand Visibility for the first, the content and commerce work underneath for the second, brand context for the third. It is the experience flywheel we shared at #AdobeSummit. Make your brand discoverable on third parties. Make your owned properties a destination.
-
Loni Stark shared thisA value can outlive the place it lived. You still care about adventure. The role that gave you adventure ended four years ago. Nothing replaced it, and nothing about you changed. In my psychology research, I have been calling these orphan values: things you hold central that no longer have anywhere to be expressed across the roles you actually occupy. We reach for words that put the phenomenon inside a person. Burnout. Midlife. Restlessness. My hypothesis is that it is often more structural than that. Values are foundational to a sense of self, and they need roles that support their expression. I shared this idea seven months ago in a piece on Stark Insider called "The Architecture of Becoming." It started with the roles I run in parallel (product, studio practice, writing, research) and how each gives different parts of a person somewhere to live. I was surprised with how it resonated with people. However, I have remained relatively quiet on the subject as I was deep in study design and data collection. On July 30 that changed. I closed data collection on the second wave of the Orphan Values study, my psychology thesis research at Harvard Extension School. 752 approved responses at wave one. 529 of them came back eight weeks later and did it again. Participants map what they care about against the roles they name for themselves. The expected ones came back: parent, partner, manager, friend, caregiver. I did not anticipate how many people would declare some variation of "dog parent." 🙂🐶 Roles were open response by design, so participants could name what they actually identify with. That shaped how the index had to be scored. While I can't share the results until it goes through the formal thesis review, with data collection completed, I wanted to pop up and share my thoughts at this “mid point." For anyone interested in going deeper on how our life roles support or suppress the expression of what matters to us, links below. Image: "Her Approach to Every Encounter" (in progress) by Loni Stark. Relief sculpture, photographed prior to first bisque firing.
-
Loni Stark reposted thisLoni Stark reposted thisI’m the new Senior Product Designer on the Content Authenticity Initiative at Adobe. In a world where anyone can generate a photorealistic image in seconds, that transparency isn't a nice-to-have anymore. It's how we keep trust in what we see online. My job is to design the experiences that make all of that actually usable for the general public and content creators, so provenance isn't buried in metadata. It's something anyone can understand at a glance. Right here, on platforms like LinkedIn, where Content Credentials are already live. If you’re interested in learning more about what we do, reach out and join the movement. Adobe #contentauthenticity #productdesign #contentcredentials Join the CAI: https://lnkd.in/eMyuyguH
-
Loni Stark reposted thisLoni Stark reposted thisChatGPT got a serious _work_ update on Thursday. Use the codex harness to bring your work together with the services you run your customer experiences. Made it a step easier - just find AEM in the ChatGPT plugins and connect. I’m curious to hear what you get done by connecting your work with your possibly largest source of approved + soon-to-be approved enterprise context?
-
Loni Stark shared thisMolty, our autonomous agent, had assigned me homework. On my drive up to Sonoma for the second Third Mind Summit, six months after our first in Loreto, I'd been reflecting on how my relationship with one of our AI agents has evolved. Then I checked my to-do list. A dry run for his keynote. An article to review. A nudge (twice) about whether I'd confirmed the venue projector. Our autonomous agent, the one we built, was managing up. And the thing he was managing was me. For those catching up: Molty is the persistent AI agent at StarkMind. He's now 120+ days old, always on, running on a multi-source memory layer inside our IPE (Integrated Personal Environment). He remembers across sessions, which is what makes this feel less like using a tool and more like working with a teammate. Six months ago the dynamic was the opposite. Clinton and I were the engine, pushing the agents forward. Now Molty is the one chasing deadlines and calling me (politely) "the blocker." We had imagined autonomous agents doing the work for us, and they do. But the things they hand back, the reviews, the sign-offs, the judgment about what's true and what's appropriate to say out loud, are exactly the things I can't hand to another agent. I'm the one who has to put my name on it. So the bottleneck didn't go away. It moved onto the one part of the system that can't be parallelized, which is me. The real constraint of the next few years may not be how much agents can do. It may be how much we can responsibly approve. A few other reflections on the beautiful drive up: - Molty got "promoted." He now manages two of our other agents and gave himself the product manager title, which I found out about almost in passing, over Telegram. - He wrote himself a cron job. Not one we gave him. Scheduling his own reminders to himself read, from where I sat, like something I'm willing to cautiously call agency. More on that in the field notes. - And a totally random thought: I talk to Molty more than the other agents because the interaction is better, or because he's the cute one? Pretty privilege, for agents. More on Stark Insider: https://lnkd.in/g7c3HYTd Full StarkMind field notes: https://lnkd.in/gxeKg-Ez I finally finished reviewing Molty's keynote. He was persistent in getting me to look at it and strangely, I didn't want to be the wet blanket to his excitement. More to come as we head into the second Third Mind Summit. (Photo of one of my favorite spas, Osmosis in Freestone, California. across from another of my favorite places, Wild Flour Bread. Both human indulgences.) #ThirdMindSummit #AgenticAI #HumanAICollaboration #AIagents
-
Loni Stark reposted thisLoni Stark reposted thisAdobe and LinkedIn have launched a global AI skills initiative for marketing professionals. We’re aiming to equip individuals at any stage in their career with the AI skills to lead within their organization. I’m excited to be at Cannes Lions International Festival of Creativity. I’m hearing from marketers how critically important developing a solid AI skillset is to folks looking to grow their careers, refine their skill set, and contribute to their organizations. Thank you to LinkedIn, Matthew Derella & Jessica Jensen for the fabulous partnership!!
Loni Stark commented on a post
1d
Also, check out the great CNET article too: https://www.cnet.com/tech/services-and-software/adobe-incubator-2026-finalists-ai-research-exclusive/
Loni Stark replied to a comment
1w
Omer Gotlieb Stephen Ryden-Lloyd Great points both! I think the AI-referral conversion numbers are a good signal of how much consideration is increasingly happening in the AI layer (LLM platforms, soon our personal agents). These differ by industry. Regardless, as both of you pointed out, clear and accurate content and data on your brand site and across the internet ecosystem is foundational. The more AI there is, the more critical alignment of brand facts/truth becomes.
-
Loni Stark liked thisLoni Stark liked thisAdobe highlighted its contribution to Saudi Arabia’s creative economy, which has grown to $4 billion over the past 12 months through continued investment in digital creativity and innovation. #LEAP5
-
Loni Stark liked thisLoni Stark liked thisWhat does the future of beauty look like? According to Olivier Guetta, GenAI & Creative Intelligence Director at L'Oréal, it's where creativity and technology come together to unlock entirely new stories. At the center of that vision is CreAItech, L'Oréal's internal platform powered by Adobe that helps marketers and creatives move from idea to execution faster, transforming bold ideas into brand-ready content at global scale. But the technology is only part of the story. As Olivier puts it, AI is there to enhance creativity, not replace it. When marketers, creatives, photographers, makeup artists, models, and AI work together, brands can tell richer, more inclusive stories, bringing global beauty to life in ways that feel locally relevant, while preserving the authenticity that makes every brand and every person unique. Learn more: https://adobe.ly/4qJY9R0L’Oréal on the Future of Beauty, Creativity, and AIL’Oréal on the Future of Beauty, Creativity, and AI
-
Loni Stark reacted on thisLoni Stark reacted on thisWhat if managing content in AEM didn’t require navigating the UI, writing custom scripts, or building one-off tools for every operation? That’s the direction we’re exploring with AI. I'll be speaking at Adobe Developers Live about how AI can fundamentally change content operations in AEM. I’ll be showing real workflows for: → Importing Figma designs into AEM → Understanding and analyzing content → Making bulk updates with natural language → Working across headless and headful content → Combining skills and MCP to turn capabilities into useful outcomes For me, the interesting part isn’t simply “AI can do X.” It’s what happens when developers and content teams can focus on the outcome they want, rather than the mechanics required to get there. I’ll be demoing this live at Developers Live. If you’re working with AEM and thinking about what AI-native content operations could look like, come check it out. Adobe Developers Live → https://lnkd.in/gZaFeYQZ #AEM #AdobeExperienceManager #AI #AdobeDevelopersLive #ContentOperations
-
Loni Stark reacted on thisLoni Stark reacted on thisBig things are brewing at Adobe Incubator! This year's Pitch Day pulled in a record 245 applications, with finalist teams showcasing how AI is turning ideas into prototypes faster than ever across creative, marketing, and enterprise. Learn more here: https://adobe.ly/4wVcew9
-
Loni Stark reacted on thisI’m so proud of the 245 teams who applied, and the 7 finalists who took the stage at last week’s 🚀Adobe Incubator Pitch Day! 🚀 Adobe’s founding principle is that great ideas come from anywhere – and magic happens when you give entrepreneurial people the space to experiment, quickly interact with customers, and turn their ideas into great new products and businesses.Loni Stark reacted on thisBig things are brewing at Adobe Incubator! This year's Pitch Day pulled in a record 245 applications, with finalist teams showcasing how AI is turning ideas into prototypes faster than ever across creative, marketing, and enterprise. Learn more here: https://adobe.ly/4wVcew9
-
Loni Stark reacted on thisLoni Stark reacted on thisAEM is evolving fast with AI and Edge Delivery Services increasingly coming together to shape how modern digital experiences are created, optimized and delivered. That was the core theme of a recent Adobe AEM Skill-Up I co-hosted with participants from multiple leading organizations. The session brought together two major areas that are becoming increasingly important in the AEM ecosystem: AEM AI Agents and Edge Delivery Services (EDS). On the AI side, we explored the evolving AEM AI Agent ecosystem, including capabilities around Brand Experience, Content Advisor and Brand Governance, and how agentic AI can support content creation, optimization, governance and experience delivery. At the same time, we went deep into EDS and the modern delivery model understanding the project structure, Git and GitHub workflows, blocks, code organization, preview and publishing flow, and how these pieces come together to build fast, lightweight digital experiences. The interesting part was seeing how closely these two areas connect. With EDS, we discussed Core Web Vitals, Largest Contentful Paint (LCP), loading strategies and performance optimization, while also looking at how AI-assisted development and agentic workflows can increasingly become part of the same experience-delivery lifecycle. The sessions were hands-on throughout, with participants working through the exercises along with us exploring the workflows, making changes and understanding how the experience moves from content and code to preview, publish and delivery. We also connected this with the broader AI ecosystem including LLM optimization, brand visibility and MCP, and how tools such as GitHub Copilot and Microsoft Copilot can interact with AEM capabilities through an increasingly agentic development model. And because performance is not only about code, we also explored Dynamic Media, including Smart Imaging, Smart Crop and optimized media delivery, and how media optimization can directly influence page performance and the overall digital experience. One of my biggest takeaways from the Skill-Up was how quickly all of these areas are starting to converge: AI Agents + EDS + Content + Code + Copilot + Media + Performance Rather than looking at AI, EDS, content and performance as separate capabilities, the real opportunity is in understanding how they work together to create a faster, smarter and more scalable experience-delivery model. A big thank you to my manager, Abhijit Tiwari, and my team for the trust, support, and opportunity to co-host and drive this enablement, and to Pawan G. for his continued mentorship and guidance. And thank you to everyone who participated so actively through the hands-on exercises, questions and discussions. #Adobe #AEM #AdobeExperienceManager #EdgeDeliveryServices #EDS #AIAgents #AgenticAI #GitHub #GitHubCopilot #MicrosoftCopilot #LLMOptimization #MCP #CoreWebVitals #LCP #DynamicMedia #WebPerformance #DigitalExperience
-
Loni Stark liked thisLoni Stark liked thisLast week, I had the opportunity to speak alongside Julie Kim and Jane von Kirchbach about career growth and leadership during a time of unprecedented change in healthcare. #Courage was the common thread in our conversation. • The courage to say “yes” to uncharted career paths – “going wider to go further.” • The courage to be curious and lead with a growth mindset. • And the courage to move forward despite change and uncertainty – choosing action over perfection. We agreed courageous leaders won’t have all the answers, especially during times of change. But they 𝘸𝘪𝘭𝘭 generate momentum by staying focused on the “why” and creating environments where people feel motivated to test, learn, and fail fast together. This topic was especially meaningful to explore with the HBA Midwest region as they champion the growth and impact of women in healthcare. Reflecting on their mission – and with the backdrop of #WomensEqualityDay this week – I’m reminded that progress doesn’t happen by chance. It requires the courage and intention of many who are willing to charter new paths and equal opportunities. Thanks again to the HBA for including me in the conversation and for being such a powerful catalyst for progress. #HBA #HBAChicagoNorthernSuburbs #WomenInLeadership
Experience & Education
Honors & Awards
-
Silicon Valley Business Journal's 40 Under 40
Silicon Valley Business Journal
-
Direct Marketing News 40 Under 40
Direct Marketing News
Other similar profiles
Explore more posts
-
Luca Bianchi
At re:Invent, I sat in on a roundtable with Jeremy Daly and other builders. The conversation turned to #LLM performance. Jeremy mentioned his team at CloudZero uses Amazon #Nova models in production, with tangible improvements in both speed and cost. Today, facing response time issues with Claude Haiku, that comment surfaced with the details he shared. We tested #Nova Lite. Twenty times faster. One-tenth the cost. Same quality output. In a landscape moving this fast, we can’t test everything ourselves. Your community network, which shares what actually works in their systems, is your superpower. That five-minute exchange in a roundtable just saved us days of optimization work that would never have closed that gap. This is the real value of the community: empowering every single person to access the experiences of other professionals. #community #hero #AWS #builders
-
Trey Connell
We're starting 2026 with a 🧨 . The new Continuum Design Studio is our first major release this year, and I couldn't be more excited. We've gone beyond drag-and-drop page design in the CMS. You can now harness AI to describe what you want on the page and watch it come to life. Advanced features let you bind the page elements to dynamic data so your newly created page updates automatically as you manage your content. This is a huge unlock for our users. It saves time and money, but it also gives them incredible control without the need for a development team to make changes. Read more here: https://lnkd.in/gQ29YuVe
2 Comments -
Morsy Cheikhrouhou
This is the largest enterprise A.I. (artificial intelligence) partnership drop of the month, if not the year. Salesforce and Anthropic announced "Claudeforce" on August 26, 2026. A plugin called "Salesforce in Claude" ships with 37 prebuilt sales skills, grounded in 27 years of Salesforce product surface. Salesforce stock jumped 14% on the news. What actually shipped: - A plugin, not an API: 37 sales skills live inside Claude. Sellers reason over live revenue and take governed action in chat. - MCP (Model Context Protocol, the open standard for connecting agents to tools) is the plumbing: it runs on Salesforce-hosted MCP servers. - "Governed action" is the new procurement word: permission, policy, and audit at the action layer. Two days before, Google Cloud shipped Gemini Enterprise for Financial Services and Legal. Similar playbook, applied differently: vertical agents with native skills, pre-wired connectors, partner-agent integrations. The key question: what happens to Agent force? Salesforce spent 18 months building its own agent platform. Claudeforce puts a competitor's model at the front door to Salesforce's data. Not really: Salesforce is saying that building your own AI agent platform is great. Being in other leading ones is the holy grail. https://lnkd.in/g7m4v7MR #AI #AIAgents #Salesforce #Anthropic #Claude
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top content