From skeptic to advocate: How experimenting with AI led to transforming work at Asana
The Spark of Realization
Almost two years ago, my journey with AI began with skepticism. My first interaction with an LLM-based chatbot (GPT-3.5) left me unimpressed – it could barely compose a coherent email. Little did I know that this underwhelming start would lead to one of the most transformative periods in my career.
The real turning point came with GPT-4’s launch. On release day, I decided to test it with the same email task that had previously failed. The result? Excellence. In mere months, AI had evolved from producing subpar content to crafting hyper-personalized emails better than I could. This moment of clarity hit me like a thunderbolt: we were witnessing the dawn of a new era in knowledge work automation.
I went home that night and started telling my friends and family that “AI is all that matters”. Turns out that intuitive feeling was right, and it would lead me down a path that mirrors what MIT’s Eric von Hippel calls the “User Innovation Revolution.”
The Learning Phase: Becoming a Lead User
The months following this realization looked a little manic. My brain got obsessed with learning everything I could about this new technology. Three questions kept driving me:
- “What use cases are primed for today?”
- “How does this technology fundamentally work?”
- “What techniques can be used to increase the usefulness of the technology?”
While many rushed to monetize their AI knowledge, I took the approach that von Hippel’s research suggests is most valuable: becoming what he terms a “lead user” – someone at the leading edge of a market trend who benefits significantly from innovations in that space. I devoted myself to two parallel tracks:
- Consuming every bit of AI content I could find
- Sending as many messages into the AI system as possible to understand its capabilities firsthand
From Enthusiasm to Action at Asana
This realization sparked an obsession with AI’s potential. While others in the industry often overcomplicate AI to inflate their expertise, I discovered what von Hippel’s research had shown across multiple industries: innovation often starts with users who develop solutions to serve their own needs, which manufacturers can then discover, polish, and scale.
At Asana, we translated this understanding into a three-part strategy that embraces the user innovation paradigm:
- Specialized AI Chats: We began by deploying purpose-built AI assistants for specific tasks, following von Hippel’s principle that users often create the first functioning prototypes. These aren’t generic chatbots – they’re focused tools designed to excel at particular jobs. When we launched these in February, some chats quickly reached thousands of daily interactions, proving that focused AI solutions deliver real value.
- Specialized AI Workflows: Building on von Hippel’s three-phase innovation model, we moved from individual user innovations to creating AI-powered workflows with Asana’s AI Studio. This wasn’t just about automating tasks – it was about creating what von Hippel calls “user innovation toolkits” that allow teams to turn successful AI experiments into sustainable, scalable workflows.
- User Experience Optimization: Our approach mirrors von Hippel’s research showing that user communities often improve and iterate on initial innovations. As adoption grew, we created interfaces that feel natural, providing a scalable system for AI to take action while maintaining human oversight, avoiding the common pitfall of overwhelming users with notifications.
The Power of Practical Implementation: User Innovation in Action
Rather than following the common path of expensive, large-scale AI projects, we embraced von Hippel’s “User Innovation Revolution” concept. We took AI knowledge around prompt engineering and specialized chatbot creation to every employee at Asana, knowing that as the AI literacy of these local teams increased, they would bring us the best use cases for each of their specialized functions.
The results validate von Hippel’s research on the power of user innovation:
- The Asana Security team revolutionized their Security Alert Escalation
- Our GTM team in Australia reduced time to triage from 3 days to 3 minutes
- These successes emerged organically from users who understood their own needs best
Our process follows von Hippel’s three-phase innovation model:
- Users identify and prototype solutions (5-10 minute chatbot evaluations)
- Successful innovations spread through user communities (internal validation by SMEs)
- Manufacturers (in this case, our IT/ET teams) polish and scale the solutions
This approach has proven that, as von Hippel’s research suggests, the most valuable innovations often come from users who are solving their own problems, rather than from top-down corporate initiatives.
Key Learnings for the Asana Community
Our journey, supported by von Hippel’s research on user innovation, has revealed several essential insights for organizations looking to embrace AI:
- Start Small, Think Big:
- Begin with specific, well-defined tasks rather than trying to solve everything at once
- Follow von Hippel’s three-phase innovation model: user prototypes → community validation → scaled implementation
- Remember that every major innovation, from skateboards to heart-lung machines, started with a single user solving a specific problem
- Context is King:
- Provide AI with clear instructions and context – just as you would for a new team member
- My favorite prompting style mirrors von Hippel’s user toolkit approach: “Your job is to XYZ. Here is the context you need to do XYZ well: context”
- This structured approach ensures AI has the same contextual understanding that made early user innovations successful
- Empower Your Lead Users:
- Identify and support what von Hippel calls “lead users” – those experiencing needs ahead of the market
- Create spaces for these innovators to experiment and share their discoveries
- Our most successful AI implementations came from teams who were closest to their problems
- Test Quickly, Learn Continuously:
- Evaluate potential use cases rapidly with minimal resources
- Use the “5-10 minute chatbot test” to validate ideas before investing heavily
- Let user communities naturally select the most valuable innovations through adoption
- Integrate Naturally:
- Build AI capabilities where work already happens, not as separate tools
- Remember that human change management is 90% of the battle
- Create what von Hippel calls “binding surfaces” that draw users naturally into the innovation process
Looking Forward
The future of work is being reshaped by AI, much like the industrial revolution transformed manufacturing. But this transformation follows a pattern that von Hippel’s research predicted: the most impactful innovations come from users solving real problems, not from top-down technological initiatives.
At Asana, we’re not just observing this change – we’re creating what von Hippel describes as a “systematic way to do what used to be done by happenstance.” We’re building tools that allow users to innovate within their own contexts while providing the infrastructure to scale successful innovations across the organization.
Your Role in the AI Revolution
As you begin your own AI journey, remember:
- You don’t need to be a technical expert to innovate with AI
- Your deep understanding of your work challenges is more valuable than technical knowledge
- The best innovations often come from users who are solving their own problems
- Success comes from creating systems that capture and scale these user innovations
This journey from AI skeptic to advocate has taught me that the key to successful AI implementation isn’t about having the most advanced technology – it’s about understanding how to apply it meaningfully to real work challenges. Through Asana’s AI tools, we’re making this potential accessible to everyone, regardless of their technical background, creating what von Hippel calls a “user innovation revolution” in enterprise AI.
Now, we’d love to hear from you!
- Have you found creative ways to use AI to tackle Asana workflows challenges?
- What’s one problem you’re excited to solve with Asana AI?
- Or, if you’re just starting out, what’s one thing you’d love to learn about Asana AI?
Share your thoughts in the comments below, and let’s inspire each other on this journey!