When 1Password announced we were having an AI-themed Hackathon, my first thought was that the content design team had to be involved.
Over the last few months, our content design team saw massive changes not only in tech but also in our ways of working. It felt like everything was shifting to AI all at once: how we designed, prototyped, built new features, and collaborated across teams.
I didn’t want our team watching from the sidelines. Content designers often do great work, but in the margins of larger design projects. This felt like the perfect moment to show how words can shape a great UX experience.
And ironically, being the anointed “wordsmiths” of 1Password’s products is just a fraction of what we do. Content designers shape page structure, look at the information architecture, create journey maps, and are systems thinkers at heart. We look at: what users need to know, when they need to know it, how information shows up, and how each action builds towards the best possible outcome for the customer.
AI lets us take those skills and supercharge them. Content designers can use AI to prototype the flow of information in real time. When you’re building with AI, you can look beyond a single screen or interaction to see how a user’s entire journey unfolds. That’s exactly the kind of thinking that content designers do best, and the Hackathon was our chance to prove it.
Participating was the easy part. We have a small but mighty content design team which include the talented Amar Majali and Grace O’Neil, who I happily “voluntold” into the hackathon. Our design leadership was genuinely excited to see what three content designers could do with AI. The hard part was deciding what area of the product we wanted to improve.
We tossed around a few ideas but came across a report that struck our interest: the SaaS Manager Savings tracker. It’s a report that shows reclaimed licenses and calculates cost savings. This report is a goldmine for any B2B admin trying to evaluate the value of their investment in 1Password SaaS Manager.
Let’s take, for example, an imaginary B2B admin. We'll call him Jerry.
Like most admins, Jerry wears a lot of hats. He’s responsible for optimizing SaaS spend, he prepares for renewals and budget reviews, and he needs to communicate the value of tooling to his executives. The Savings tracker can help him come up with that value, but it’s buried inside the product. Even if he was to find the table, sharing it with his team wouldn’t be an automatic home run, since it’s all numbers with little context.
For the three of us content designers, this is where the lightbulb went off. We already had the raw materials from the report. Our challenge was making it easier for Jerry to show how 1Password was saving his company money. If you haven’t guessed it from Jerry’s name, we lovingly called our Hackathon project “Show me the money.” The name was a bit cheeky, but it summed up the design problem nicely: how might we make SaaS Manager’s value impossible to miss and effortless to prove? The team needed to use our UX thinking to answer three core questions:
How does Jerry easily discover the value of his savings? (aka “the money”)
How can he optimize to save more money?
How can he share wins with his team and stakeholders?
We quickly realized that answering these questions required us to think about Jerry’s entire journey, and find opportunities to AI most effectively.
For a hackathon team with zero engineers, creating a workable prototype was a tall order. Our team set out to build a vibe-coded prototype hosted on the Knox playground (a playground created by our amazing Design Systems team). This meant getting set up in GitHub and Cursor, setting up our branch so we could collaborate, and entering commands in terminal while praying that it all worked out. As cliché as it sounds, we were building the plane while flying it: learning how to be pseudo developers while still using content design thinking to build out Jerry’s experience. If I’m being completely honest, it was messy and occasionally terrifying. But somewhere in the chaos, the team stopped worrying about whether we were entering the right Git commands and started learning how to build together.
Here’s what we learned:
Learning 1: Alignment beats chaotic building
With great power comes great responsibility, and using the Cursor agent felt dangerously powerful. With the right prompt, we could spin up any component or design we could think of, and the urge to vibe-code everything all at once was very tempting.
In theory, this is great when working solo, but as a team it gets chaotic fast. It’s easy to lose sight of who’s changing what and where, while trying to understand the UX rationale behind the decisions they’re making. Our team quickly realized that we needed to sit down and align on “what good looks like” for Jerry’s journey. Using a Figjam board, we huddled on key questions to understand where in the journey we wanted to make changes:
What do we want to show? Total savings month-to-month.
**Where do we want to show it? **Front and centre on the admin dashboard (not buried in a report).
**How do we want it to show up? **A number that shows real-time savings with a graphic.
**How do we want Jerry to feel? **We want to spark delight. Let’s do something to celebrate milestones for big savings.
**Why does it matter? **Let’s make sure Jerry can easily share the savings with context to his team or executives. If he wants to dig into the numbers, let’s direct him to where he can find the Savings Tracker.
Answering these questions helped us align on what we needed to build upfront to move fast later. It also helped us envision what sort of story we want to tell as Jerry sees his savings grow.
Learning 2: Don’t eat the elephant all at once
Once we knew what parts of the journey we wanted to change, we set out to divide and conquer. Each of us owned a piece of Jerry’s journey:
Amar built the shareable report experience to show Jerry’s wins to stakeholders.
Grace built the celebratory moment for when he hit a savings milestone.
I built the scaffolding, starting with the dashboard card and then the seat optimization flow.
For each of us, we learned the hard way that big prompts, like ones asking the agent to build out an entire report or dashboard card, led to ambiguous results. Instead, it was easier to work in small, iterative prompts. For example, I asked the agent to duplicate an existing card, explained the new card's purpose (tracking savings), and then asked the agent to design a visual to accompany the card. Each small iteration allowed us to see how our UX rationale held up or where we could make improvements. For the dashboard card, my team member Grace was able to iterate on the visual to make a graph with a slider that shows savings month-by-month. The original static visual would’ve worked, but seeing it move in real time made it obvious how much more valuable a dynamic version would be to Jerry’s journey.
Learning 3: AI demands more collaboration, not less
Contrary to the popular assumption, working with AI made us more dependent on collaboration. Alignment gave us direction, iteration kept us moving, but collaboration is what held the two together.
Since the three of us were working on one branch, we had to huddle constantly on Slack or Zoom to discuss new changes or pivot when parts of the plan didn’t work out. When one of us pushed their changes, we had to let the others know so they could pull the latest designs. When something broke, it was easier to figure it out together than trying to use Claude to troubleshoot on our own.
On demo day, we had two minutes to share a video of our workable prototype. The team knew the value of the feature we were building and why it’d be helpful for B2B admins like Jerry. We just needed to bring our audience along the same journey. In truth, we needed to make them feel Jerry’s dilemma as much as we did.
So instead of walking through the feature, we incorporated the actual “Show me the money” clip from Jerry Maguire. Yes we were trying to be funny, but also it was the exact thing real admins would hear from their stakeholders: show us how 1Password is saving us money.
We introduced Jerry from the clip and then showed how our prototype helped him:
Surface savings instantly
Understand where and how to cut costs
Celebrate when he hit a big milestone
Share that story with his stakeholders to land a promotion
We didn’t know if our unorthodox demo would resonate with an audience that was mostly technical, but we did what content designers know how to do best: tell a good story.
In the end, the gamble we took with the movie clip paid off. Our “Show Me the Money” project won the Customer Impact Award and was runner-up for Bit’s Choice. We’re also working to get a first iteration of the dashboard card and graph into SaaS Manager this year.
Before that hackathon, I was worried that the content design team might get left behind or stuck on the sidelines of the AI shift. The hackathon put some of that worry to rest. While it proved we could vibe-code solutions, our team also realized that our skills as storytellers offered a new way of thinking and building.
In the end, while AI helped us build faster and more independently than we could have before, what made this project a success are the human qualities that AI can’t supercharge: deep UX experience and instincts, nuanced context about our customers and our business, and the ability to reference ‘90s pop culture.
The funny thing is that the UX superpower of content designers was always there. We’ve always helped shape the product vision: asking the right questions during user research, creating lo-fi sketches, and even defining product terminology and names. But now we can take it further. We can design content-first, and bring it all the way to the build.
For any content designer wondering whether this space is for them... it is. Your instincts about clarity and story are exactly what AI-assisted builds are missing.