Repositioning Zoom's platform for the agentic AI era

Background

Zoom had outgrown its own homepage

I joined Zoom in 2021 as the web platform was still optimized for a meetings company. By 2024, Zoom offered 19 products across communication, productivity, and collaboration. The homepage still told the old story.

Shaped by years of incremental additions without structural realignment, the experience had become fragmented. Users could find individual products, but couldn't understand how they connected or why they mattered together.

We were given a broad directive to fix it. No defined point of view. No strategic brief. Just "update the homepage."

Building the foundation from scratch

Rather than designing to a brief that didn't exist, I initiated four parallel research tracks to build one.

How do you run four research tracks simultaneously without losing coherence?

Three concepts, one shared language

My team synthesized the research into a strategic brief and developed three homepage concepts, each emphasizing a different priority: platform clarity, user outcomes, and AI positioning. These weren't presentations. They were a shared vocabulary for making decisions.

Then the story changed.

Pivot

More than a new strategy: a new company

Midway through the project, we received a company-wide directive: reposition Zoom as an agentic AI platform. Public announcement at Enterprise Connect. March 2025. 10 weeks away.

This was the start of the "agentic" wave. Leadership was intent on positioning Zoom at the forefront of it. But beyond the label itself, there was almost no guidance on what that positioning should mean or how it should be communicated.

Because the website is the most visible external expression of the Zoom brand, it became the primary vehicle for articulating the shift. The homepage wasn't just supporting a new company position; it was the position.

There was no narrative to execute. We had to write one.

What does it actually look like to write a positioning narrative as a design team?

Scope doubled. Timeline shrank.

What began as a homepage redesign became a full web ecosystem effort: global navigation, the All Products experience, the AI Companion page, core product and marketing pages. The timeline that had 13 weeks remaining just became 10, even as the surface area roughly doubled.

Insights & Strategy

Translating "agentic" into something users would actually believe

Leadership handed us a buzzword. Our research told us why that was a problem.

The qualitative study had surfaced a consistent pattern: users were curious about AI, but deeply skeptical of AI hype. Abstract capability claims landed poorly. Concrete, outcome-oriented framing landed well. Users didn't want to hear that Zoom was an "agentic AI platform." They wanted to understand what that meant for how they work.

So we filtered the directive through the research. Rather than leading with the label, we anchored the positioning in tangible outcomes, building a narrative that made the AI story credible by grounding it in things users actually care about.

Process Reinvention

Redesigning the way we worked

With a fixed 10-week timeline and no defined narrative, our standard linear process (content, design, develop, test, launch) was no longer viable. I was responsible for coordinating work across design, engineering, Product Marketing, content, partner web teams, localization, and testing, each with their own dependencies and review cycles.

We didn't manage the timeline. We rebuilt the process to fit it.

What did you cut when scope doubled and time shrank?

1. 01

   Parallel execution with modular pods

   Organized work into small, cross-functional pods, each responsible for a defined set of components or sections. Content, design, and engineering progressed together within each pod, in parallel. Bottlenecks dropped, and ownership lived at the section level.

2. 02

   Centralized alignment

   Formed a cross-functional task force of team leads that met daily. Not to report status, but to resolve blockers and make decisions. Fast feedback loop, decisions without delay.

3. 03

   Building for iteration

   Designed the page modularly so different narrative approaches could be tested and swapped with minimal rework. This reduced risk and set us up to validate with real data post-launch.

4. 04

   Deciding quickly, simplifying aggressively

   With incomplete information and limited time, we prioritized speed of decision-making over certainty. Cut scope, reduced edge cases, made calls early. Momentum over perfection.

Execution

A unified AI-first platform experience

The final experience unified three things the old homepage had failed to connect: a credible AI narrative, a clear product ecosystem, and pathways built around what users are trying to accomplish.

Outcome

A reimagined web experience in 10 weeks

In 10 weeks, we launched a cohesive web experience that positioned Zoom as an AI-first, agentic platform, despite an evolving narrative, a scope that roughly doubled midway through, and complex cross-functional dependencies across more than half a dozen teams.

We didn't wait for a fully defined direction. We created the structure that allowed the work to move forward: using research as a filter on strategy, design as a tool for alignment, and process reinvention to make an impossible timeline workable.

The result established a clear, unified model for how Zoom's platform is presented: AI, products, and user outcomes connected into a coherent experience that could scale across the broader web ecosystem.

Three takeaways