The Story Behind AI Agendas: Making Meetings More Human

At Supernormal, we're all about creating AI tools that make your work life easier and more meaningful, and our process is as thoughtful and people-focused as the features we build. Here's a deeper dive into how we came up with our latest feature, AI Agendas, and what we've learned along the way.

Why We Care About This

We all know how draining it is to walk into a meeting unprepared. It's stressful and can make interactions more reactive and less genuine, which can, as a result, make meetings an unproductive waste of time. At Supernormal, we care about solving this problem because we believe meetings should be a time for meaningful connection and idea-sharing, not just another task on your to-do list.

When we build AI tools to help you prepare, we’re really trying to help you show up more thoughtfully and connect more meaningfully. By handling the tedious details for you, we enable you to focus on having richer, more impactful conversations.

How We Got Started

When we decided to tackle this problem, we kept things flexible and creative. Our approach to product leadership has always been one that leverages the ingenuity of our team. We knew we wanted to help teams prepare for meetings without the usual hassle. With this in mind, we let our team explore different ideas freely.

Once we set the stakes in the ground with a broad vision for AI Agendas and emphasized the importance of making meeting preparation seamless and efficient, the team dove in. With this direction, they started researching how people currently prepare for meetings, gathering insights, identifying key pain points that the new feature could address, and figuring out the mechanics.

To make AI Agendas work, we had to change the underlying infrastructure of Supernormal notes. Previously, everyone received unique notes tailored to them. We needed to ensure everyone in attendance got the same set of meeting notes so that Agendas would work seamlessly. This change was critical to ensure everyone was on the same page (literally!), and it paves the way for more collaborative features in the future. Our Engineering team got to work rebuilding our infrastructure to make it happen, allowing for a shared, consistent experience.

While our Engineering team was updating our infrastructure, the Design team focused on how AI Agendas would show up for users. What started as a simple UI within the Supernormal app that allowed users to add an agenda before a meeting quickly expanded into other surfaces when we realized agendas were just as important for keeping people on track during a meeting and structuring meeting notes afterward.

Meanwhile, our AI team was busy experimenting with different prompts for AI agenda suggestions for recurring meetings. We wanted to ensure our recommendations were not only accurate and helpful but also formatted in a way that would be most effective for our users. This meant testing various styles and structures until we found the perfect fit.

Learning and Growing: Internal Testing

Once we had an MVP, we started by testing AI Agendas internally. At Supernormal, we have meetings too, and we need tools that actually work for us. This internal launch was crucial as it allowed us to gather honest feedback from our team and make necessary tweaks.

Feedback from our team came in various forms. Our designers and developers provided formal feedback during daily standups and weekly product and design reviews. The rest of the company shared their insights more informally, often in a dedicated Slack channel set up for the feature's ongoing development. This channel became a treasure trove of ideas, with teammates from all functions sharing their thoughts on what they wished the feature could do.

The real magic happened when we saw how much AI Agendas improved our own meetings. You don’t truly understand how unprepared you’ve been until you experience the opposite. With AI Agendas, you get a moment to think about what to say, leading to more thoughtful and meaningful interactions. It was a game-changer for us, and we knew it would be for our users too.

Listening to You: The Beta Phase

Throughout this journey, we learned a lot by launching a beta version to a select group of users and listening to their feedback. This phase was essential to ensure our product worked well for everyone, not just us.

During the beta phase, users told us that creating agendas often felt one-sided and they wanted a more collaborative approach. In one interview, we heard comments about how distributing an agenda ahead of time and getting buy-in is especially important for high-stakes meetings with external clients. This was something we didn’t fully consider when testing on our internal meetings because the stakes are lower, but it highlighted how important it is to have functionality that lets everyone contribute ahead of time.

This feedback led us to develop easier ways to assign agenda items and share agendas once complete, ensuring AI Agendas truly met the needs of our users.

The Finished Product: AI Agendas

After many iterations, we finally crafted a version of AI Agendas that we’re truly proud of. It allows for easy collaboration on meeting agendas, offers helpful AI-driven suggestions, and lets you assign agenda items and share agendas effortlessly.

At Supernormal, we believe in taking our time to perfect each feature, ensuring the final product is something we can stand behind. This commitment to supernormal quality is what sets us apart. Our thoughtful approach ensures that we deliver tools that make your work life not just easier but also more human.

Special shoutout to Mika Isayama and Jay Schaul for their help in bringing this update to life and the whole Supernormal team for making the vision for AI Agendas a reality.

Join 300,000+ teams using Supernormal to move their work forward

Sign up for free to discover the magic of Supernormal.