Sean Hobson is chief officer and clinical assistant professor at EdPlus at ASU. He is also a longtime collaborator, colleague and friend. Recently, Sean shared with me a new free online tool that he built called DesignedChange.ai. The description of the tool reads,
“This short assessment measures how intentionally your organization has designed its approach to change across five strategic dimensions—and shows exactly where to focus next using the [learning innovation department] framework for transformational change.”
In an effort to both highlight/share DesignedChange.ai and to better understand the motivations and process behind developing this tool, I asked if Sean would answer my three questions.
Q: What is the context in which you developed DesignedChange.ai? Whom do you think it will be most useful for? What sort of conversations or outcomes do you hope that the tool catalyzes?
A: First, thank you, Josh, for the platform and for everything your column does for our community, giving voice and visibility to work that can otherwise feel isolating.
The context is twofold. I care deeply about the importance of learning organizations and their ability to evolve alongside the people and environments that need them. That’s what led me to spend four years on a doctoral dissertation studying how these organizations create transformational change—research that became a book chapter with you and Eddie Maloney for Johns Hopkins University Press and a strategy paper with Jeff Selingo for his culture series. But the framework kept nagging at me. The entire architecture for a diagnostic tool was already there, buried in a PDF that maybe a few dozen people would ever read. I wanted to move it from the dissertation shelf to something practitioners could actually use.
Second, like most of us in this line of work, I’ve spent my career learning by building—prototyping, testing, adjusting. In design, that’s how we were trained. This project was an extension of that instinct, shaped in part by what I’m learning co-designing the Agentic Self course with will.i.am at ASU, where our students engage with AI by building with it, rather than passively consuming it.
I’m imagining the tool can be most useful for anyone leading, building, working within or advocating for innovation infrastructure and processes within their organizations. Within colleges and universities, that could be presidents, chancellors and provosts, but also online learning officers, faculty leaders and individual change makers.
The conversation I hope starts is a simple one: Are we designing our approach to change, or is change just happening to us?
Q: I assume you developed DesignedChange.ai using AI? What tools did you use? How much time did it take? What did you learn in the process of developing the site and how might we apply those lessons to our learning innovation work?
A: Yes and I chose the .ai domain to force myself into going all in. I have very little coding experience and haven’t built anything like this before. I opened Anthropic’s Claude, described what I wanted in plain language and started iterating. My first prompt was something like,
“I want to build a web-based assessment tool grounded in my doctoral research with personalized results, elegant and minimal. Draw on the entire knowledge base of UX principles, survey engagement and game mechanics.”
Nine days later, the site was live with users: 4,500 lines of code, an AI research agent, a database back end, automated email delivery, analytics—all deployed to production for about $200 total, half of that for Claude co-work. The estimated replacement cost if I’d hired a developer, according to Claude: $25,000 to $38,000. Claude’s co-work mode was my primary tool, but it introduced me to an entire ecosystem I’d never touched—Vercel, GitHub, Supabase, Microsoft Clarity, Claude API and Resend.
The experience felt less like software development and more like a design critique and collaborative production. I provided the intent and domain expertise. AI handled the production. The feedback loop was minutes, not several-week sprints. When things broke—and they did and are—I’d describe the problem or share a screenshot and we’d fix it together.
The lessons for our community: The value is only as good as the clarity and imagination going in. AI amplifies domain expertise and can fill gaps if you’re curious and willing to push through. The core skill isn’t coding; it’s describing what you want with precision and knowing when something doesn’t feel right. Those are design skills. The same ones we use to build learning experiences. For anyone curious about the full process, here is a step-by-step build report that I again produced with my new co-work agent. One other lesson is don’t wait for the perfect idea or for your institutions to provide the software. Start building something, anything, right away … and try and make it serious.
Q: At DesignedChange.ai, you define a learning innovation department as “a dedicated enterprise unit within a learning organization that serves a dual function: operating as a reliable service organization delivering core capabilities—while simultaneously functioning as an R&D skunkworks that explores emerging opportunities and drives institutional transformation.”
Can you expand on this definition? What are the professional and scholarly communities of practice that practitioners and researchers in this field participate in? If someone wanted to develop a learning innovation department on their campus, or find colleagues doing similar work at other institutions, where might they turn?
A: The LID is a term I developed to describe the type of organization you and I have helped build within our universities. Nobody had quite grounded this work in the academic literature yet. Centers for teaching and learning have a long history, but they weren’t the right frame for what’s emerging—enterprise-level units with dual functions: a reliable service portfolio delivering instructional design, media production and learning technology, combined with an R&D arm exploring new models and driving institutional transformation. Those two functions require different governance, human resources, funding and success metrics. Running them as one undifferentiated unit is a common mistake I see universities make. You can’t just drop this work onto a library or teaching and learning center and expect them to perform a different function than what they were built for.
To ground this work, I drew on organizational design, management theory and the change literature—Eckel and Kezar, Kotter and Senge. But I also deliberately bridged into business strategy—specifically Lafley and Martin’s cascading choice framework—which became the backbone of the assessment’s five dimensions. That bridge matters. LIDs need to think about competitive positioning, market focus and strategic trade-offs the way enterprises outside the university do and act as the translation layer to the university culture and operating model. Our sector has succeeded for a long time without having to compete in that sense, but the landscape has shifted. The LID needs to help the university stay competitive, even when the business language feels foreign or uncomfortable.
We don’t yet have a formal professional community for this work, though subcommunities exist at OLC, Educause, UPCEA, ASU+GSV and others. Josh, your writing with Eddie Maloney and this column remain one of the best starting points. And I’ll offer one last plug: The AI research agent at designedchange.ai is built for exactly this conversation. If someone is building one of these units or trying to understand how theirs compares, I’d welcome them to take the assessment and engage. Context matters enormously, and I believe these units are mission critical for universities worldwide. I’m excited to see the results come in to further inform the tool and future research by myself or others.
