Creating Alignment When No One Agreed on the Future

BEN's agency model was human-capital intensive — every brand deal required people finding creators, negotiating terms, managing campaigns. The team was understaffed and overworked. Leadership didn't believe the agency model alone had a bright future.

TubeBuddy had its own problems. The founders exited with the acquisition — and they hadn't left a vision behind. The product had been built without designers, so there was years of UX debt. Ten million installs and deep YouTube integration, but no direction for where it was going next.

When I joined, neither side had a clear path forward on their own, and together they hadn't figured out how their work connected. No shared roadmap, no unified vision, and a lot of fear.

VisionType I made the direction concrete. Leadership then shifted to VisionType II — a more flexible, Notion-like systems platform. The Head of Product believed "every creator is different" required infinite customization.

Where I Disagreed

I pushed back — with data. I'd audited the flexible, systems-style products — Notion, Mem, Obsidian, ClickUp, Airtable, Coda, Tana, among others — and one pattern kept repeating: outsized investment in onboarding. Guided setup flows to save users from the blank canvas. Step-by-step walkthroughs. Video tutorials embedded directly in the UI. Template galleries everywhere. Standard-sounding features — but invested in at a scale most SaaS never needs. When a product requires that much teaching, it's telling you users can't intuitively see what to do with it.

We were living the same evidence internally: Notion was meant to be our standard for documentation, but much of the team had a hard time adopting it. And I'd seen the pattern years earlier, consulting for a startup with no experienced product person on the team — badly behind schedule because it kept building customization, in part because it hadn't done the research to know what its users actually wanted.

To be clear, I wasn't against flexibility — I was against unexamined flexibility. The same research doubled as a map of what it would take to do a systems platform well if we committed to that path. I believed in common patterns: nail the workflows first, add smart flexibility where it earns its place.

The Head of Product made a different call — and I took on the challenge with real excitement, becoming one of the two main authors of VisionType II alongside him. That stretch produced some of the collaboration I enjoyed most: the two of us thinking deeply about the system together — what capabilities it needed, how each one would work, how they'd interact. Disagreeing with a direction has never meant withholding my best work from it.

That exploration included designing AI agents as collaborators in the future product vision and building early working prototypes before native agent tooling was widely available. Those prototypes were exploratory rather than shipped product features. Later, Nick Johnson and I partnered on some of the first working agents in our codebase. We each created agents and helped refine one another's work, turning the broader idea into a practical way of working.

Where the vision ended

TubeBuddy faced a post-pandemic decline and growing competition while we worked toward the new platform. The platform progressed, but funding was withdrawn before broad release. Investment priorities changed, our runway shortened, and the focus shifted toward near-term profitability.

We never reached the point where we could demonstrate the business outcome the original vision aimed for. My contribution was the research, product direction, and design work that helped the team move forward under those constraints.

I still think specific workflows with the right amount of flexibility beat infinite customization — and AI may well change that equation. But the starting point doesn't: understand what users actually need first.